The Entire Sentient Trading Society Public Material (text only):

Strategy Engineering Volume 1

The STS Community Handbook Sentient Trading Society Expanded Edition Welcome to The Sentient Trading Society (STS) First of all, I need to stress that the STS methodology is not complex. We are just extremely precise for optimal understanding. Some FAQs are below. BE PATIENT and do not be intimidated! It takes a hell of a lot longer to write it than to read it. So read and benefit! Several people, from beginners to advanced traders, have said these materials helped them refine their approach. Know that, regardless of where you are at in your trading journey, Sentient Trading Society is an excellent place to study. We are a community that prioritises research and facts first. All strategies and techniques we teach are backed by our trading performance on regu≠lated platforms only. Full trade history and platform footage is publicly available and shown on our YouTube channel (click to view). Click to play footage of us on a regulated trading platform. Everything we teach is backed up by peer-reviewed submissions to respectable academic journals rather than anecdotes. Everything will click after reading the documents and watching the videos, and once it does, it is a powerful catalyst for a profitable trading career. The goal is that, after work≠ing through the documents and examples provided, you will have a personal framework for building and testing your own strategies. 1 We recommend revisiting parts of the documents every 2-4 weeks to reinforce your understanding and make learning more efficient. If you have a busy schedule, setting reminders can help you stay on track with≠out feeling overwhelmed. The documents we provide are interactive PDFs with playable voice notes! These voice notes are worth listening to, as they include embedded voice notes from us (Ron and Ali). If you open them in a compatible PDF reader, you can listen to the explanations alongside the text. This makes the material more interactive and much easier to learn. To get access to these features, you must install Adobe Acrobat on your PC or MAC and you must ìTrustî or ìEnable mediaî (it is free). Remember, it is optional. Our materials take less time to complete than a video course and are much more precise. For the best learning experience, we recommend reading at least one document per day and using reminders to stay on track. To make this even easier, Text-to-speech software, such as NaturalReader or Speechify, can read PDFs aloud, allowing you to use spare time more effectively and helping to support reading when busy. Latest changes Learning quality improvements, one by one Over 100 hours committed to improvement. 1. Clauses have been shifted, and hundreds of sentences have been reworded for clarity, including adjustments to help bilingual readers understand semantics better. 2. Conversational tone has been stripped in areas where seriousness is required but sprin≠kled in places to reduce tension and help key parts register. 3. Over a dozen figures have been made so each concept is easier to understand, and some are automated (they autoplay on Acrobat PC and Mac). 4. The material has been compressed and reordered. Noisy parts have been removed; =40 pages have been removed entirely. Unnecessary voice notes have also been removed to reduce reading time. 5. Several additional sections have been added in every section. Key pages will be named below. Additional references have been added to strengthen the rigour of the material. 6. Intraday trading period intricacies are now covered, so it feels easier to pick logical trading hours for your specific strategies. Instead of generic ranges. 7. A post-reading guide is now available in Section 5. 8. Interactive Easter Eggs are now a part of the material; some are clickable links, and some are videos in plain sight. Some of it is STS lore, and the rest are jokes. It is your job to find them on your second read ;). STSí Strategy Engineering Papers: 2 Easter eggs, Mental Frameworks & Discipline: 1 Easter egg, Execution & Venue Mechanics: 2 Easter eggs. 1 Significant Changes Feel free to use CTRL+F to search for document title names. If you are new, take it slow and download one PDF at a time; if you have read the material previously, re-download everything. There are now 5 Volumes + Academic References. Aim to read at least two PDFs per day. If reading becomes difficult, text-to-speech software may be helpful. Do not give up. Changes: 1. Strategy Engineering: Volumes 1 and 2 Volume 1: The Grounded Approach to Trading Strategy Design. An underfitting, overfitting and good fit section explaining them with visual aids and original analogies. Volume 2: Out-of-sample testing has been added îOut-of-Sample Dataî (CTRL+ F) îEssential Clarifications & Insights Following Readingî has been enhanced. Note: ï Trading costs have been separated from volume for readability. ï Core FAQs has been renamed Key Insights and has been reworked. 2. Market Microstructure Principles & Warnings Strategy Engineering Volume 3: Logic Optimisation & Intraday Timing: More depth on îIntraday Trading Session Optimisationî and îSTS Strategy Typesî Note: The tone has been adjusted in some areas, and noisy parts have been removed entirely. Over one hundred individual sentences have been reworded. 3. Mental Frameworks & Discipline The Logical Fallacy Handbook Cognitive Conquest Real Market Psychology Multiple sentences have been reworded for clarity. 4. Mechanism Validation & LLMs îAI x Trading: Prompt Layeringî has been enhanced with valuable information about AI that most people ignore when researching. An additional section on AI reasoning flaws has been added, paired with peer-reviewed sources, references and figures. 5. Execution & Venue Mechanics: Brokers, CFDs and Prop Firms Prop Firm Mechanics Additional points added surrounding industry conflicts of interest with examples of prop firm manipulations. Broker Choice, Venues, and Regulation. The FX aggregation figure has been improved, and the section has been enhanced with a reworded explanation. What to Do After You Finish the Material (New and important) Volume Changes The materials have been numbered and spread over 3 volumes for optimal absorption and enhanced referencing. More precise guidance on AI and information hygiene has been issued throughout multiple documents. The delivery (tone) and statements (wording) have been adjusted to complement the nature of our work. STS Strategy Engineering We have reframed the entire strategy design material as an unravelling of a decision tree, derived from our reasoning and the marketís first principles. We have set the expectation explicitly and reinforced it with our knowledge throughout volumes 1-3 with dozens of added figures, added interactive media and in-depth, precise explanation. We have tightened the link between psychology, averaging, variance and execution. Parts have been reworded for clarity and reorganised for improved impact. More actionable statements and enumerated paths have been added and explained, so readers can understand the foundations that hold us up in our reasoning phases. Any parts where we previously assumed the reader would know X, we have reworded and explained rigorously to lift the pedagogical quality of the materials. The strategy design parts now house custom clean microstructure examples and explanations delivered cleanly for things to hit quickly. We have written to make îConstraints firstî more persistent throughout the material through wording and real examples of applying STS-style reasoning to trading situations. We now make it super clear that rule-building happens inside concrete life constraints (sessions, holding overnight, maximum risk) and directly link ignoring constraints to noisy results. We also provided Raw STS Implied Backtesting Cost-Style Calculations within Volume 1. I have also talked about how industry-grade media psychological operations take place, how retail traders suffer, but not just how to avoid it, how to benefit from it as well, with real examples, visuals, and evidence rather than unanchored conspiracy. Being contrarian is not an edge. Evidence is the edge Exposing Industry Nonsense The auction / microstructure-based material now includes the ìcrowd-induced slippage eventî example with multiple custom Depth Of Market tables and an explanation of ad≠verse selection and how the mechanism(s) can look. Broker Choice and CFD Microstructure I lay out what CFDs actually are and how their markets differ from exchange products. I have actively tied this to a clear explanation of market makers, A-book versus B-book, and why internalisation exists, including the fact that most brokers use mixed models. îMarket makingî is not automatically malicious; it is a risk-management function. But not only that, I explain why everything is done so you can understand the FX and CFD industry (as a trader) down to the molecule, with minimal additional effort. Once you know the product(s) youíre trading and how to see through the vague legal kung≠fu they force us through. The understanding is powerful and will be reflected in your P&L through saved costs of working with a good firm (if you choose to operate with CFDs). We donít just talk about the product itself, but why every single part of our setup, including platforms, is crucial for low-cost interaction with CFDs for UK citizens, Europeans, and Asians, without shilling any broker affiliate links (we are not affiliated with any). Contents 1 Significant Changes 3 2 The list of actionable topics covered in STSí materials 6 3 How to work through these documents 7 3.1 Section 6: Academic References and Supplementary Material . . . . . . . . . 7 4.1 Why dowe publish? ............................... 4 Our Principles 7 4.1.1 Theconnections.............................. 7 4.1.2 We Enjoy Teaching ............................ 8 4.1.3 Our approach leans heavily on research and academic literature rather thanexperience. .............................. 8 4.1.4 When you adopt a process like this, you can avoid many common traps. 8 4.1.5 It isObvious toMany, But... ...................... 9 4.2 AreWe ProfessionalTraders orRetail?..................... 9 4.2.1 DidWe Have aMentor? ......................... 10 4.2.2 We Are Not Here to Force Your Hand; We Are Here to Guide It . . . 10 5 Disclaimers: 19 2 The list of actionable topics covered in STSí materials 1. Logical fallacies, critical thinking and how to improve your reasoning abilities. 2. How to filter out noise in trading circles. 3. How to avoid social strain and psychological pain when pursuing trading. 4. How to reframe your psychology to have the mindset of a market practitioner. 5. How markets really work. 6. How market makers actually work with extra nuances. 7. How Futures, FX, Derivatives and CFDs work with extra nuances. 8. We issue guidelines on how to make a mechanical rule based trading strategy with examples on how to develop a real market edge. 9. Exposing how the retail trading industry operates and why strategy sharing does not make sense (with academic citations). 10. Market Microstructure basics and the reality for modern institutions. 11. We compare retail prop firms versus real institutional prop desks. 12. We provide guidelines on how to beat standard retail prop firm evaluations. 13. Finish with clear insights on how to grow with retail prop and scale responsibly like we have. 3 How to work through these documents We recommend reading the materials in the default order provided in Discord. Once you have completed them, move everything into a folder so it stays organised in the exact order we published it. Section 6 is the only optional part. 3.1 Section 6: Academic References and Supplementary Material These document is reserved for those who are curious to see the proof and backing behind our process and thinking (from reputable economists and others in the industry) or for those who desire additional deep trading knowledge and relevant academic citations to pair with the STS strategy design modeL. This is for those that need reassurance. Leave the additional reading until the very end. Remember, it is optional. An academic book list is also included. An STS member contribution from a Psychology Major is also both readable and listenable. 4 Our Principles STS focuses on teaching people strategy development, testing, developing skills, building discipline, and providing research-backed insights. These principles guide not just how we trade but also how we learn. To us, a framework is only as strong as the people who apply it, which is why we focus so much on logic and reasoning, as well as collaboration and connections with one another. 4.1 Why do we publish? 4.1.1 The connections The point is to have a place where traders who take the work seriously can talk through ideas with others dealing with similar situations. Members can benefit from exchanging perspectives with other traders who are focused on relatable goals. Some members have met in person when it has been practical, and conversations often spill over into related areas such as business and career decisions. The main aim is to have a place where effort and clear thinking are normal. Beyond the social aspect, mentoring in our experience has reinforced our knowledge. Teach≠ing has given us greater exposure to nuance and to questions we had not asked ourselves. It also helps us better understand the needs of the average trader. 4.1.2 We Enjoy Teaching It is satisfying seeing students win, and reaffirming improves our abilities and pushes us to research even more. We still read peer-reviewed papers to this day. There are always new things to learn. The commitment to research and constant improvements has defined our work, both private and public, from the start. 4.1.3 Our approach leans heavily on research and academic literature rather than experience. Everything we teach is backed by research and academic-grade literature instead of anec≠dotes. Rigour: the evidence, the research, the numbers. This is what comes first. It is tough for a trader to have faith in a system without proof of efficiency. We are by no means someone who never gets anything wrong; we test, we fix, we move on. 4.1.4 When you adopt a process like this, you can avoid many common traps. We make solid reasoning central on purpose because it is essential for rational decision-making, rejection of poor information, and recognising high-quality insights. The main point and benefit of STS is for readers to have a robust filter against poor infor≠mation so they can grow and improve decision-making with what they know and have access to. Each traderís work is their own. The aim is to give you tools and filters so you can build and assess your own edges. That is why we always remind members that statistics outweigh anecdotes and experience-based opinions. We know this very well. Knowing where liquidity is and being able to exploit it for alpha are two different things. I am talking about a genuine market edge and not just short-term profitability. I (Ron) have been doing this for eight years now (together, much longer), and our first profitable year was in 2021. Before then, it was just trial and error paired with research, trying to figure out how to create systems that produce real edges without overfitting and, after that, how to manage risk and profitability optimally. We realised what was missing: designing strategies to take advantage of market behaviour by anticipating it, with every part of the strategy being logical. Then everything else on top of that is optimised to maximise the small edge that has been extracted from the market. Why before what. We were focused on data before logic when it should have been the other way around. When we were struggling in 2020, our shift in mindset forced us to re-examine even the simplest truths and common narratives on why markets move. It took us almost three years to realise that modern markets are not ruled solely by the buy/sell pressure that YouTubers waffled about. Three years. Think about it. 4.1.5 It is Obvious to Many, But. . . The only way to make a profit from buying is if people buy after you do, and the only way to make money shorting is if there is sell volume after you. Many know this but do not take it into account in their strategies. We teach people to anticipate the mean reversion and align with the trend when favourable. There are many ways to succeed in trading. The focus here is on methods that can be defined, tested and repeated. 4.2 Are We Professional Traders or Retail? A lot of people confuse retail for professional: A retail trader is an individual who buys and sells financial assets with their own personal accounts and with their own capital. Technically, we are ìretailî because we are not financial advisors or licensed practitioners (we only trade our own capital). Professional traders are licensed practitioners who manage other peopleís capital, often for a performance fee. ìProfessionalî broker status in the UK is different and it is based on experience and capital. Review FCA Handbook COBS 3.7.1 If a trader meets at least two of the following three conditions, they can ap≠ply to be classified as a ìprofessionalî trader in the UK or Europe. 1. Investment portfolio: You are required to have a ìfinancial instrument portfolioî (defined as cash deposits and financial instruments) of Ä500,000 or more. Acceptable instruments include cash, stock portfolios, stocks and shares ISAs, trading accounts, mutual funds, and SIPPs. Managed company pensions, non-tradeable assets, property, luxury cars, or physical gold are not acceptable for applications. 2. Trading experience: Traders must have placed 40 trades of significant size within the last year. Significant size is £10,000 notional for equity trades and £50,000 for everything else. 3. Professional experience: You must have worked in the financial sector in a profes≠sional position (which requires knowledge of CFDs or spread betting) for at least one year. We have spoken to a few individuals in the industry, but we have never worked for any of them. All of our conclusions are based on personal discoveries and research. When people realise this, we are usually asked. . . 4.2.1 Did We Have a Mentor? We did not have mentor(s), and we have never had interest in conforming to typical retail narratives or reviewing retail trading education to align with them. I know this might be perceived as out of touch, but we do try our best. We do not want to commit time to reviewing false ë îeducatorsî. 4.2.2 We Are Not Here to Force Your Hand; We Are Here to Guide It For example, in my anti-discretionary posts, I give a constructive guide on how to extract value (if any) from common methods taught online through YouTube, TikTok, courses, and so on. Filter and optimise; there is no need to abandon completely. The main point and benefit of STS is for our members to have a robust filter against poor information so they can grow and improve decision-making with what they know and have access to. We exist to bring clarity to trading through reason and research, built for traders who demand both clarity and precision. Across retail CFD trading (a common form of leveraged retail trading), regulators consis≠tently report that the majority of retail accounts lose money (around 80% depending on jurisdiction). Most trading successes are blocked by a weak process: unclear rules, poor cost modelling, low samples paired with weak testing, and no operating discipline. We know because we were there over five years ago, and the most unfortunate thing is that most people who have the potential never figure it out due to retail distractions. This material exists to strip that away so itís possible to succeed. Turning Market Logic into Tradeable Systems The Sentient Trading Society Framework Why Many Traders Remain Unprofitable for Years and How We Help People Break the Cycle The Real Reason It is because many people tell themselves everything is îrisk managementî and îpsychologyî instead of developing a proven system. People are quick to say ígamblingí without knowing the full extent of a losing traderís story. What worsens the probability of success is the reinforcement of poor practices through media, gurus, and articles. Most traders will read and watch anything but a financial markets book from accredited sources; STSí literature is only a derivative with concise delivery. What most traders ignore If a strategyís setup does not return a positive value on average, no matter what the trader does, after costs, they will lose money. You need a positive expectancy to generate positive returns. 1% Risk per trade isnít going to produce any financial return worth talking about as retail. Risk isolation creates wealth; we have leverage. We use it on smaller capital instead of lump sums. A risk-averse management style can help traders churn for longer; thus, it makes market makers, brokers, educators and exchanges more money the longer you trade, not profit. Retail guidelines frequently function as industry incentives rather than objective operational advice. The more volume we trade, the more spreads and commissions are earned by our counterparties and the venues involved (conflicts of interest). When a strategy works as intended, we compound it; if it does not perform, we move on. When we have multiple edges, we run them separately to isolate the gains. A lot of traders say... îYou can make anything work.î Reality: You can be profitable purposefully with logic based on research backing up your trades or reach profitability coincidentally with hope and theory in ways that cannot be replicated. We would rather be profitable on purpose instead of relying on luck, If there is something baseless that you have to îmake workî, it is an illusion, as it does not take genuine market mechanics into account. As counterintuitive as it sounds, you should make ideas that should happen to work when applied; logic comes first before the data. The best way to use data is to show that strong, unbiased ideas are profitable, instead of focusing on patterns that may not repeat. Sure, the strategy appeared to work last year, but is there a logical reason the success should persist that is not a conclusion drawn from the backtest or forward test itself? If you canít explain why it would succeed without testing data, the edge can easily be weak or non-existent. The wrong thought process is: I have an idea; let me test it to see if it works on this market. I have seen this price action setup work a few times; let me test that. Iíve seen this indicator, and I thought of this idea. Let me see... This flawed inductive reasoning weighs a lot of traders with potential down with this common process; any success is guaranteed to be coincidence without direction and rigour. This leaves traders scrambling aimlessly for years; to succeed, we need order. The correct thought process is: These behaviour(s), backed by research or first-party testing, happen in [THIS MARKET, E.G., Nasdaq] or [ASSET CLASS, E.G., Soft Commodities] during these hours because of this [EVIDENCE HERE, Light or heavy]. What evidence? Evidence that the supposed undercurrents which support your edge are real: Proof that the cause is real and not theoretical with paired knowledge of how price moves and how it looks when applied to charts is how you get there. It is up to the trader to define ideas mechanically afterwards. What is unfalsifiable is useless to us. The market does not care about anecdotal successes behind a method. We run on empirical evidence. Logic before data. Why would it work in a way that can be proven plausible with an un≠derstanding of how markets actually operate? Robust test results are great, but we need to know about the mechanism(s) responsible for every supposed edge. Why would it provide any advantage if deployed in the way thought or taught? Do price and datasets behave in the way we think, or is it an illusion? Circular reasoning never measures up for anything, especially for íevidenceí. Heuristics can be helpful, but they must have an underlying basis. We trade alongside well-established market behaviour The establishment itself does not establish edge; the derivative of the establishment produces profitability. It is a baseline itself for understanding that gives you the leverage. The correct mindset when building: When you design your own strategy, you earn and own a unique execution pattern for dealing with it, which gives you the edge. The point is it exists. It has a basis, proof. We do not think this behaviour exists; we know it does. The advantage: îThis strategy I have made to take advantage of this move should return good results. If the data agrees with the strategy rules and hours, I will optimise and consider running after rigorous tests without overfitting.î This deductive reasoning is oxygen for serious traders. You do not even need to view charts to create market-beating edges with this reasoning. When you get good at it, profitable strategies transition into something you are dying for, into shower thoughts. When there are abundances of ideas unique to you, it minimises the struggle into îjust another strategyî. Behaviours in markets do not just happen to happen. As a serious trader, your focus needs to be isolated on building a strategy that is likely to work, instead of throwing stuff at a wall until something sticks. Just because something sticks on the wall doesnít mean it stays. With evidence, you rig the game of edge discovery. Why before what. For example, I understand how a market I approach trading operates; for example, if it tends to mean-revert intraday, such as AAPL stock, YM Futures/US30 or 6E Futures/EURUSD, I will look to anticipate and fade the trend. I aim to position myself to benefit when that happens. That mean-reverting behaviour is well established and will persist, so the strategy should remain effective for a while. It is a way of exploiting what genuinely exists; something that sticks. Having an edge is about acting before others do. There is evidence that from 8:30 to 10:30 EDT, there is a strong mean reversion tendency in the Nasdaq as well; with that, you can create a strategy to predict and fade the short-term intraday range. There is also peer-reviewed work showing when specific instruments tend to trend. Strategies can be made to exploit this as retail; for institutions, it is more competitive as they move the price. As retail, you do not have to worry about this for a while, but since we have experience dealing with high size on thin order books, we have processes on how to deal with that as well. The Comfort and Structure Within STSí Strategy Design Process Comfort and Pre-defined Execution Patterns Institutional literacy reduces paranoia; ignorance is one of the biggest reasons we experience psychological issues when trading. Humans want to feel in control, as it is an innate desire within us. îI donít knowî does not suffice when money and pain are on the line. It makes us spiral; it is uncomfortable. A clear understanding of market structure and costs irons this out by bringing what needs to be known to the table. Understanding market participants and their roles, market price mechanics, and simplified but essential statistics helps you avoid retail myths and keeps you operating at practitioner-level efficiency instead of on fear, this makes execution calmer. When everything is defined, even during high-stress periods such as large drawdowns, you experience little to no decision fatigue because you have not made a single decision since the strategyís deployment. You see a setup, you place it, wait, and... nothing. 1. The hard work should be designing your edge rather than executing it. 2. The use of predefined rules and brackets makes relaxed trading possible and scalable. 3. Trading in a way that legitimately takes away from your peace is not worth participating in, especially in the beginning. 4. If in the beginning you struggle, we suggest traders opt in for higher timeframe setups, for example, =hourly charts with or without overnight holds (depending on the asset class and leverage constraints). STSí Strategy design is comfort first by design to match real-life constraints and has liquidity comfort at the core of the operation. All entries and exits are pre-planned; execution is mostly limits with predefined exit points, so there is little room for deviations in the heat of the moment. When a strategy is 100% mechanical in nature psychological control becomes easier and these benefits surface. 1. You donít react; you observe and execute. 2. You wait and pocket the expansion, never chasing it. 3. Your strategies run only on hours you can control. No constant screen watching, no fear of missing out. 4. Sessions are consistent, planned, and repeatable. Whether the session is long or short is up to the designerís goals, risk tolerance and execution feasibility. You canít be doing 5-minute candles in an office job, but you can check your phone once an hour... 5. You operate on instruments and hours where reasonable market depth is available, making fills orderly and slippage tolerable. You are not forced into markets with poor conditions. Psychological comfort You feel calm before, during, and after trades because proof replaces hope. Reviews are short and factual instead of emotional. Strategies built this way feel more comfortable to execute in real time, are led by evidence, and are built for robust, quiet consistency. We know, at least subconsciously, that real trading is not flashy; it is the experience of trading with the clarity provided by rules you can trust instead of relying on faith and anecdotes. You need to engineer certainty with data, even delusion. But in a way that benefits your wellbeing Confidence does not come from pep talks and hope. We are human; we know subconsciously that the only thing that will ease anxiety is certainty, or at least the feeling of it, it is how we are wired biologically. This opinion was initially deduced from our experience in the field and a phenomenon affirmed in institutional work regarding psychology, which we have studied and cited on your behalf to ensure balanced, precise delivery. Trading for a ísteady incomeí is an expectation planted by marketing Most traders who are great get large cash flows from trading and leverage that elsewhere to become wealthy. That is the optimal, replicable way. If you want to do it solely through trading, it is much harder but possible. You can make a LOT from trading if approached seriously, but expecting a monthly salary-type consistency is delusional; there will be los≠ing months or even losing quarters, and I have had them. 20k in a month, followed by a 10k drawdown, followed by 30k profit the next month is an example of what an established traderís ëincomeí pattern could look like. You walk with more confidence after each recovery. We talk about the importance of withdrawals and other aspects later in our material. Withdrawing is a part of the game, not a betrayal of it. You can either be the person who lets the chickens youíve raised roam free without any guards or guardrails for the illusion of extra profit, standing to lose everything when it is fox season. Or you can take a profit. Most traders who experience profitability fail to take anything off the table, while winning tactical withdrawals that balance growth and self-preservation are planned in advance by us, not improvised in panic at the point of ruin. As clich¥e as it sounds: work smart, not hard. Blind effort and return are not correlated. Be lazy but tactical; make your strategy so easy to execute in real time that it feels effortless. No looking at charts without alerts, that is how far Ali has taken it. What is the point of ìfreedomî if youíre forced to look at charts for 4+ hours a day? ***** that. Ping me, and I am ready to place that order. Success in Business Is Not That Different from Trading Creating a profitable business is similar in principle to trading. For a successful business, you have to identify inefficiencies, and you must fill them. To sell a business is like setting a take profit; if you do not want to run it perpetually, you turn the ítradeí into an investment. The marketís pricing is highly efficient, though not completely, and each position pushes the price towards where value should be. By trading efficiently, you are rewarded for participat≠ing in this process. For example, you could fill an inefficiency at a market low on the S&P 500 and hold it for 50 years. Or you could sell it in 5 years and do another business at a different inefficiency, leveraging the realised profit to compound your wealth into a larger number, just like a sequence of trades instead of relying on one large position. Using íequityí from unrealised P&L to take more risk is like using a profitable businessí assets as collateral to fund a new business; more risk equals higher profit potential but higher risks of getting liquidated. Bankruptcy. A margin call notification is like a creditor in business issuing a warning to you about loss of valuation; if you do not restructure or add funds, you get liquidated. Manually trailing your stop loss is like selling a percentage of your business. Involving an experienced associate is like scaling into your trade. It can work against you, but if lower risk values are used (share assignment) and limits are set (stop loss/legal paperwork), it can increase potential upside by double or more while risking less than 50% more compared to the initial position. Same game, different wrapper. We operate alongside well-established market behaviour Well-established phenomena in markets do not give participants an edge; the strategy derived from the establishment is what produces P&L. With that, you have a unique execution pattern for dealing with it, which gives you the edge. The point is that it exists. It is not a theory; it is real. And at the right angle, with the correct positioning, you can benefit. For example, you can open a business that sells desserts. Everyone sells desserts, but they do not prepare them in the same way you do. The way you prepare and offer the product is what gets you that elevated business. 1. People like waffles and milkshakes, and that interest persists (a well-established market behaviour). 2. People in your area only sell sugar waffles and nothing fresh; waffle irons are hard to get (imbalance). Regardless, people want fresh waffles (market inefficiency). 3. Strategy: You source a waffle iron seller and study the mechanics of waffle creation, ingredients, and food business marketing (strategy design). 4. Deployment and preparation: You pick a suitable venue to execute your strategy, for example, leasing a space in a busy mall and hiring staff, the same way you choose and fund a futures, stock or a CFD broker that meets your needs. Choosing to operate in a destitute mall is like depositing in a broker with poor execution or regulation. 5. Customers/market takers come to the mall/exchange and buy from you, the one who makes a market (like a limit order), taking the liquidity (the waffles) from you. And there it is, all demystified; you have brought efficiency to the marketplace briefly by rebalancing the market inefficiency. The cash flow/return on those trades with customers persists until the strategy loses effectiveness. When other people use the same strategy at the same venue your edge decays. By sharing your exact strategy rules, e.g., your waffle iron supplier or the specific recipes or ingredients, your strategy loses its market edge. It is the same thing institutional traders call alpha decay. Insiders, like employees or business activity trackers, can leak information under the table, which can erode your advantage. This is why ítrade secretsí matter. When you decide to ditch the granular perspective on everyday markets briefly and zoom out, you will realise how close all these systems really are to each other whilst others pretend they are wildly different. Even in physics, inefficiencies behave like vacuums: gaps attract pressure, imbalances invite a response, and eventually, things tend to drift towards balance. That is natural order. Just as our estimates about how nature behaves are imperfect, success in trading is never guaranteed, just as it is not guaranteed in any other business. Time and Opportunity: You need to adopt this as your mindset: Efficient trading can be leveraged for wealth creation or as an assist to get large capital for business or asset ownership, such as real estate, even with lower starting amounts. íSmallerí traders with account sizes under 20 million USD do not need to worry about influencing the price on liquid markets so we have flexibility in areas they do not. Trading is not a get-rich-quick scheme, unlike what some common educators may lead you to believe. If you were listening to someone about business and they said it would take years to get your first sales, you wouldnít listen to them, right? So why would you for trading? You need to create effective systems now, not in three years, now. It is easy to underestimate how much structured work you can get through in a year if you are focused. With structured testing and a realistic work ethic, many traders can develop and validate a strategy in less time than the typical ìit will take you yearsî narrative suggests. The material here is intended to shorten the learning curve by collecting research, examples, and process ideas in one place. It does not remove the need for your own testing and decision-making, but it can reduce the amount of trial and error. Our material will help you make that work more focused. You need urgency to get notable success in anything, unless you are born into it. We compiled a structured set of examples and references to reduce trial and error, with the aim of shortening the required time commitment by many months at minimum. We showcase our findings and philosophy both publicly and, in more detail, privately. Many members also bring experience from STEM fields and business, which shapes the discussions. The secret is being intentional with who you talk to about trading. People who treat trading as entertainment usually drift away. The ones who stay tend to use discussion to refine their work. Those who donít understand financial markets can only add noise to your journey. Most traders love the fantasy, not the work. Consider our work as a composite of the logic needed to succeed as a retail market participant. How it feels designing mechanical trading strategies. Multiple traders have said that once their rules are clear, the backtesting itself feels straight≠forward, even though to avoid experiencing decision fatigue during real-time deployment. When you achieve this: It will be like this permanently. Once you have your own collection of decent models, a lot of the work can be isolated to building upon what you already have and your own understanding. Testing becomes significantly simpler when noise and flaws are reduced. Your ability to deduce ideas will speed up over time as well. The only intuitive aspect is the line of thought for system design instead of individual trades. The drawbacks are removed by knowing what is logical, what to look for, and what to push back against. Any intuition in the process now saves you from decision fatigue and potential mistakes under stress later. You will make most of your decisions in the design phase, so you have the luxury of making close to zero decisions during real-time execution. This is what streamlines our process while keeping our psyche in check. If you feel something, examine it. Strategies with holes will not hold up, and the gaps often show in your P&L first. Remember this. When you read in silence, no one else is watching, but the work you do here will shape what follows. Beating the market is not a casual task. So you must create standards for yourself and be set on keeping them. We have distilled what took us years of learning into clear reads with visuals to save you time. Each sentence is important. When your trading career is not established, comfort and hesitation are the handcuffs you must unlock, and excuses are the invisible chains you need to break. We aim to expose what may bind you and show you how to use that freedom to your advantage. References [1] Discussion Paper DP25/3 Expanding Consumer Access to Investments. Cites the FCAís finding (2022) that approximately 80% of customers lose money when trading CFDs. [2] The European Securities and Markets Authority (ESMA) cites analyses showing 74-89% of retail investor accounts typically lose money when trading CFDs across EU jurisdic≠ tions. Note for USA readers: The drivers of retail trading losses are not unique to CFDs; leverage, costs, and weak pro≠cesses compound across retail trading environments. In the United States, products such as options and futures can introduce even greater leverage, amplifying downside risk. It has taken us over 500 hours to make this material. And seeing STSí students apply it and win is a real reward. Their progress makes the work, including updates, worthwhile. Creating the material only needs to be done once, and whoever takes it seriously will realistically draw benefits. 5 Disclaimers: Document Legal Use and Sharing All Sentient Trading Society (ìSTSî) written materials, branding, and likeness are protected by copyright. These works are released only under the applicable Creative Commons licence stated in the document and may be shared for personal, non-commercial use, provided they remain unaltered and full attribution to STS is preserved. This work is licensed by Sentient Trading Society, under CC BY-NC-ND 4.0. To view a copy of this licence, visit https://creativecommons.org/licenses/by-nc-nd/4.0/ Any unauthorised use of STS material or likeness, including but not limited to the cre≠ation of derivative works, translation, rebranding, resale, or other adaptations outside the scope of the stated licence, is strictly prohibited. Plagiarism, misrepresentation of author≠ship, or any attempt to pass off STS content as your own will be treated as a violation of our intellectual property rights. Any unauthorised use, derivative work, or plagiarism will result in takedown requests and, where necessary, formal legal action to enforce our rights and seek appropriate remedies. We Are Not Financial Advisors or Licensed íPractitionersí All Sentient Trading Society (ìSTSî) materials are provided for educational and informa≠tional purposes only and do not constitute financial, investment, trading, or legal advice. Nothing in these documents, discussions, or examples should be interpreted as a recommen≠dation to buy, sell, or hold any financial instrument, nor as a solicitation to engage in any specific trading strategy. Trading financial markets involves substantial risk and is not suitable for all investors. You can lose some, all, or more than your initial capital. Past performance, hypothetical results, or backtested examples do not guarantee future outcomes. You are solely responsible for your own trading decisions, risk management, and for complying with all applicable laws and regulations. By using STS materials, you acknowledge that you do so entirely at your own risk and that STS, its authors, and contributors accept no liability for any loss or damage arising from your use of this information. A CFTC Risk Disclaimer on Derivatives is available to view here: CFTC Disclaimer Sentient Trading Societyô The Grounded Approach to Trading Strategy Design Strategy Engineering Volume 1 STSí Building Principles Ron -Sentient Trading Society How to approach this document A few practical tips: ï Read one section at a time: start with the foundations here and progress. Do not skip the queue. Make sure to stop and think at each important point. If you just read for the sake of reading, you will not get far with expanding your knowledge. ï Rereading is required for full comprehension. ï Attempt to sit down and create your own ideas, which are candlestick ideas based on what you have learnt. Here, adaptability is key. Scroll slowly. This collectionís aim is to help move you away from retail storytelling to fair test ex≠periments. Your reliance on traditional retail strategies should fall significantly. Try to be humble when things fail, and be precise when things work. Document everything, test continuously, and resist the urge to stitch lots of untested tricks together. These insights can be used and can help you to make systems to be used for confluence, trend bias, and much more in accordance with the two other Strategy Engineering volumes. For notes, platforms like Notion and Google Docs can be used to organise your workflow. The STS framework is simple once you understand the basics behind the research and the statistics that back it up, and learn the same language from the research and its fundamental meaning. You are reading the updated version of Strategy Engineering, so there is an extensive glossary included at the end for post-reading. Once you understand the meaning, you will understand why these industry terms are important. We do not rely on complex tools to trade or test. We focus on the first principles of modern financial markets (market microstructure) and apply them rigorously to OHLC data. Throughout this document, we will walk you through our reasoning step by step, giving you a base understanding of how this approach influences our strategy design process. Each volume goes more in-depth with carefully crafted analogies to make framework click. Let us begin. 1 Introduction Most active traders do not fail simply because they are lazy. They fail because they overadjust, build strategies backwards and/or never collect enough back test data. I have been there. I have chased systems and setups which did not make entirely logical sense, maybe intuitive, but not logical to earn the title of being systematic. They also were not suitable to my schedule either so I had difficulty trying to keep up with my trading. Eventually, I stopped following noise and started designing and building my strate≠gies from bare bones. Right from the beginning. The following document will concisely break down step by step (not just rules) regarding what should be done from little trading experience. For a trader with the will and discipline to design a strategy that can take advan≠tage of existing market edges, this is how they should go about designing their strategies. In this document, we walk through the decision tree from the very first node. The goal is to help you build a clear way of thinking about decisions, so you can develop an approach that stays adaptable as market conditions and life responsibilities change. What We Avoid The STS framework rejects traditional retail methodologies, including: ï Indicator-based entries ï Retail auction market theory interpretations ï Retail price theories ï Real time intuition or saturated strategies Each avoidance will be explained rigorously throughout multiple volumes. Our materials are about revealing the structures required to succeed without forcing a single method. Many of the principles we lay down also apply to other demanding fields. You also stand to learn and actively avoid the logical fallacies (flawed thinking) that held us and many other traders back for years, such as circular reasoning. The purpose of each of the explanations is to expose how randomness can fool us com≠pletely by momentarily looking good, and what the market does not acknowledge. For other useful ideas, we discuss the effort versus potential reward in a balanced manner to help you make your own decision(s). Everything changes the moment you move from following strategies to building your own. You have to take control. What We Do We have developed this framework (unique to STS) to mechanically integrate insights we have learned from market microstructure and continuous auction behaviour (well≠documented and researched first principles) into objective, repeatable entry and position-management techniques, allowing us to benefit indirectly from already existing market tendencies instead of relying on overfitted or saturated trading strategies. Every consistent trading signal is based on some kind of repeatable pattern, whether it is expressed through candlesticks, order flow, or predictive models. Guidelines on filtering what is acceptable have been attached. Some people use other entry methods too. As long as the base framework is understood, it can be expressed through other tools, including order flow. This is a case of ordinary chart trading versus trading grounded in mechanism, context, and testing. We do not trust entries unless they are logically grounded, mechanically defined, and tested under realistic constraints and costs. Here is a shortlist of distinct practices private STS traders apply: 1. We have risk isolation and withdrawal techniques to retain returns. 2. We avoid overfitting with fixed guardrails to avoid it (this will be defined later). 3. We reject bad ideas with specific processes instead of relying on feel. 4. We produce mechanical definitions for consistency and reproducibility (price aspect). 5. We have procedures in place to manage an entryís randomness relative to price (time aspect). 6. We produce specific reasoning paths within the mechanical definitions to decrease path dependence (to stop a past tradeís outcome interfering with a future tradeís outcome). 7. We accept the reality of edge decay acceptance and prepare for it in advance. 8. We collect out-of-sample data (most traders do not bother). The public STS material is a demonstration of how we reason and make trading work for us. Each person who applies these principles will share similar reasoning while operating with their own 1-of-1, unique, and private strategies, allowing us to teach without saturating our trading. Our work requires more effort due to its universal applicability, but if the way we were teaching could saturate our trading, we would not create the society. By the end of Volume 3, you will know exactly what we are talking about. Now that is covered, let us get started. 1 Feel and Adjust Constraints First We must figure out our initial constraints. Doing this will remove a lot of noise from your trading and subsequently, will make your life easier. So, choose: ï Time of day you can realistically trade. Be very realistic, avoid idealising. ï Knowing in advance if you need to sleep or work through certain sessions and what that means for your trading execution. ï Whether you want to hold trades overnight and whether that is compatible with your system. This is a yes or no, and is on a strategy-per-strategy basis. ï How much capital you will trade with. Starting now and also forecasting into the future. These are chosen as all rule-building happens within constraints. If you work a day job and trade five-minute charts, you are probably not able to trade the New York session. If you only trade during the London session, you do not build rules around the Asian session. It really depends on time zones and other factors. Higher time frames like hourly allow for higher versatility. For example, most could realistically execute once per hour if busy, but not every 5 minutes during high-volume hours. Ignoring constraints is why a lot of retail traders go nowhere, they copy others without aligning their system with their actual life. If you are "trading here and there", then it is adding noise to your results. The more variance in consistency, the worse it is for your bottom line. But there is something we must cover before we go further... 1.1 An Essential Data Science Concept To Learn Underfitting, Good Fit, and Overfitting. Figure 1: Three separate panels with different strategies applied on the same data. See underfitting as a strategy not being loosely in sync marketís mechanics or not at all. On figure 1, this is shown by the strategy (line) being too simple to align with key data. Sure, it is close enough coincidentally, but that is not enough to extract an edge after costs. Underfitting is a key reason why classic one simple setup strategies, with a lack of filtering, tend to return close to breakeven or lose money in backtests or during real-time trading. Underfitting is like going for a random walk without any sensations and hoping to find yourself at a date planned at 8pm but you find yourself at the supermarket instead. See having a good fit as surfing the marketís wave, not knowing exactly when it starts, but the average timing when it shows up is good enough to catch it on average due to the strategyís efficiency. Strategies with true good fits are desirable because they have an increased tendency to work in real time. Having a good fit is like having a date, time and route for a plan in advance but thereís traffic and other natural hurdles (drawdowns) on the way. On Figure 1 is shown by the strategy being in line with the data efficiently but not matching it 1:1 See overfitting as forcing the system, historically, to catch most individual waves exactly when they start. Every single wave is different, and differences in future wave formation will destroy the supposed historical efficiency that the strategyís data claims. Overfitting is like having a specific route to walk from the supermarket to your house in Spain, but then teleporting to a random place in Italy and being forced to take the same path. You will not arrive at your house. Overfitting is the reason why most backtested retail trading strategies fail to perform effectively in real time. This is simple, but without knowledge of this, the probability of a traderís success is significantly reduced. Both underfitted and overfitted strategies will suffer in real time, and they hold back many traders who fail to acknowledge it. The worst part is that overfitting your strategy is what human intuition guides us to do. Traders who use discretion do this in real time through journaling and are influenced by recent trading results, while systematic traders tend to make the same mistake when designing strategy rules, aiming for the best results instead of aiming for the best logic. Even AI and machine learning operators face this problem, it is a heavily documented institutional problem which most traders do not know about. But now you do. Let us move on to the next section. 2 Selecting One Market and Timeframe (At the start) You cannot experiment with everything. Pick one instrument and one timeframe. For instance, you may choose Dow Jones and the hourly chart. This is because different markets behave differently. Attempting to make a system that works on, e.g., The Euro, Gold, Oil and the Nasdaq at once is usually unwise as fitting a strategy to past data rather than the marketís principles reduces real-time effectiveness and can invalidate it entirely. Now, linking back to the previous section, executing a single system effectively is demand≠ing; attempting multi-system deployment across multiple markets at different times (per strategy) without establishment increases the risk of operational failure significantly. One market. One behaviour set/trade setup. But when you decided to run multi≠ple instruments or systems, split the risk amongst them. Note that each one should be good enough such that if you were to isolate the risk, then each would perform well enough on their own. There is no space for mediocrity. Next you need to understand how your chosen market behaves, see [Note 3 and Reading 5] post-reading. Is it mean reverting, close to a random walk, or trending. These following examples must be refined and understood by yourself. This forces you to research and learn. Plenty of articles and books cover this. These examples are not absolute, they serve as a guide. Here they are, Intraday examples: ï Mean reverting markets: Apple/AAPL, Dow Jones/YM, EURUSD ï Near random walk (alternating): S&P 500/ES (close to a random walk with drift) ï Trending: Gold/GC These are not the only markets you should be looking at, as there are many more to consider. In Figure 8, we present extremes for heuristic purposes only. A key mistake traders make when attempting to build something profitable is rely≠ing solely on market data (price, order flow, indicators, or any other derivative of real-time price) as the sole contributor to inference. Those systems are descriptive with no material predictive value in isolation. The difference is that our strategies are built on objective, market-specific mechanisms that indirectly benefit from price movement rather than explicit price prediction, so we consistently rely on processing skills rather than coincidental successes. Simple anchors are provided in this document, alongside more nuanced examples and protocols for interaction, to keep the guidelines universally applicable whilst retaining objectivity. We do not accept circular explanations for supposed strategy effectiveness.îBecause the data agreesî, îthis book saysî, or îthis person saysî, as these have never been sufficient for robustness. The why, what, and how need to be clear before deployment. Do not worry if you feel like you do not understand everything at first. By the time you finish the STS material, most questions you have now will be answered. 3 Start Building with Logic Instead Of Results Beginners may initially study charts to understand candlestick geometry, but must transition to logic-first strategy design to avoid confirmation bias. You should first write an idea down and then test it. Never the other way around. Think about why your idea would work before you know what the full strategy is. Reverse engineer why it would work in the market and condition(s) and cross-examine the logic. Figure 2: Simplified Visuals for STS Deductive Reasoning STS System Class 1 How we deal with price in three sentences: 1. We trade based on OHLC based micro auctions and time series inefficiencies. (explained later within the material) 2. Our edge comes from favourable market regimes and predefined filters. 3. We tactically provide liquidity with limit orders only when doing so has positive expectancy on average. While order flow software provides advantages in ultra-short-term trading, it introduces unnecessary subjectivity for multi-hour to multi-day horizons. For trading strategies with horizons ranging from several minutes to hours or multi≠ple days, the potential benefits of such tools are interpreted as highly context-dependent. They can introduce subjectivity due to the dynamic nature of liquidity, while regular OHLC price data is delivered in a more simple, static, repeatable way. For most peo≠ple, these tools often introduce additional complexity and over-optimisation risk (for systematic traders) or decision fatigue (for intuitive traders) without materially enhancing decision-making. Later in the material, we will explore this in more depth, with examples both for and against the softwareís use. Nonetheless, we have succeeded without it. We leverage the understanding of the market itself (aggressive, passive, efficiency, and rebalancing) to build effective systems, order flow is one of many ways to apply our framework. Proper interpretation of OHLC data can provide comparable insights and trading opportunities. It is also worth noting that institutional exclusivity in a product, software, or service is often what makes it îinstitutionalî. Prime brokers are institutional, standard retail-facing order flow is not. The trading industry is quite misleading on this point, but we revisit this in the market microstructure section with some numerical examples for you to weigh before reaching your decision(s), for now, we must focus on the foundations, let us resume. We prioritise methodologies that produce clearer and more objective generalisations with a low number of parameters (preferably) for consistency. These tools are not com≠pletely off the table, and when our trading sizes start to impact or influence the market(s) we trade, we plan to revisit it. 3.1 Rely on logic and statistics The basics: OHLC refers to open, high, low, close data, which is typically represented with candlestick bars. Market Regime is referring to marketís state, trending higher, trending lower or ranging. A Micro Auction is a small event where price overextends, then reverts quickly to a more balanced level before continuation or reversal. We aim to get filled on the correction, obtaining a premium price to either go long or short. In institutional microstructure related work, this is often described as a short term "liquidity shock" and recovery. The "shock" in market microstructure is what is inefficient, and the recovery is efficient. We anticipate that inefficient prices will become efficient within the sequence, and we refer to this as rebalancing, as price temporarily finds balance once again, providing superior prices and costs when compared to market orders on bar closes. We aim to take advantage of the established mechanics of modern price discovery and statistical laws regarding stationarity with clear, effective, and uniquely designed strategies. We have created over a dozen unique entry techniques using our framework. Time series inefficiency: A repeatable, statistically testable, non-random pattern that should not persist without further interaction after formation in a perfectly efficient market or an opportunity revealed by price to absorb aggressive order flow to provide an edge. 3.2 Economists do not trade, why listen? You need to understand the causes behind market moves in an objective way so you can build tests you can reproduce. Peer reviewed institutional work from economists will not magically hand you an edge as a trader, but it can give you the insight you need to design your edge(s) or improve the one you already have. Even the abstracts from solid pieces are golden. We leverage institutional research to perform a completely different job: trading efficiently 3.3 Do not change your rules as you go along. And most importantly! Avoid go searching through charts trying to find ideas to test. Start at the drawing board instead of candlesticks. Forget indicators. Forget entries. First you need structure. The following sections address what to make rules about. 3.4 Trade Time Window (Tied to Constraints) You must define which hours are valid for entering trades, based on when your chosen market has high volume. For Example, 8am to 4pm NY time for US indices. Why? Because you need volatility to reach targets and you need volume at your entries for price to trend in your favour regardless of your system style (reversals, mean reversion or trend trading). Rule example: ìI only take trades between 3 pm and 9 pm UK time.î This could be the time you could realistically execute trades so it is the time period you should be exclusively testing. You can mark this with a sessions indicator (e.g., ìSessions on Chartî indicator on TradingView with the 10:00 to 16:00 setting). Simple, objective ways based on research to adjust the hours you choose to trade without overfitting will be provided later in the material along with references. 1. A peer reviewed paper published in The Review of Financial Studies (Context & Behaviours) combined with 2. Federal Reserve Bank data publication based on 22 years of data 1998-2020 (raw market volatility data and more.). 4 Risk Management Decide what you are risking per trade, as a fixed percentage of account equity (e.g., 3%). In a live environment, this value needs to fit your risk tolerance and goals. Your risk must be planned ahead and adhered to. It may be static or dynamic. There are advanced methods for this, but for now focus on simplicity. For prop firms, calculate your risk to comply with maximum drawdown rules. Normal example: if a system can suffer ten consecutive losses (this would be classed as -10R, where R stands for risk. 10R = 10 ◊ risk in percentage) and the prop firm allows up to 10% drawdown, you might trade (as a random example) 0.8% per trade to allow space for peak-to-trough drawdown plus a buffer (around 20% extra for instance. This is extra space for slippage, human error and general strategy instability). Again, much more advanced methods exist for these calculations. risk when holding on profitable running positions. For instance, when entering another position on another rejection (scaling in), having pre-defined plans to increase risk during winning, or losing periods in live environments depending on their risk tolerance and goals. Decide your risk-to-reward ratio (RRR) before testing (e.g., 1:2, 1:5, etc.). Do not adjust it to chase better performance. It must be based on logic. You must also be aware of your trading costs etc, we explore this further in Strategy Design Volume 2. Rule example: ìI aim for a 5 RRR on reversal trades" or ìI aim for a 3 RRR on continuation trades". If the system does not work, I throw it out. Added annotation for clar≠ity, see [Note 1] post-reading. 5 Entry Style (Define Setup Type) Bar replay back-test only. Never scroll backward to ìcheckî the setup again. Pick something linear and logical. Mean reversion? Reversals? Continuations? Breakouts? Then ask: ï What does that look like? ï Do I want price to hit a level and reject (reversal)? ï Do I want price to push through and pull back (breakout/continuation)? ï Why would it work? ï What does my setup signify via the priceís mechanics? îLiquidityî or îOrder flowî is not a system or strategy like educators teach. It is the basics of how markets move on a tick-by-tick basis (Microstructure). Basic example explanation: If there is a buyer at $10,000.25 who wants 100 units and only 80 are available, then price moves up one tick to $10,000.50 to fill the rest. As an example, consider the following: Ask price Volume available $10,000.50 50 $10,000.25 80 A simplified order book example. A buyer submits a market order for 100 units. 80 units fill at $10,000.25 and 20 units (the rest) fill at $10,000.50. Volume-weighted average fill price: 80 20 10000.25 ◊ ( ) + 10000.50 ◊ ( ) = 10000.30. 100 100 Hence the average fill is $10,000.30 and the last traded price now stands at $10,000.50. This is liquidity. ìLiquidityî is not a strategy; liquidity is how easily an asset can be bought or sold without causing a meaningful move in its price. In terms of individual orders, liquidity is the amount of buy or sell interest available at or near a given price. The only reason price moves is that there is an imbalance between the buy and sell volume relative to what is available in real time. Nothing else. Note that a tick is the minimum price movement on an instrument. That is why markets have a highly random nature, see Fig. (6). For example purposes only, see Fig. (3): 3-wick reversal, ìFor this setup I place limit orders at the beginning wick of a 2-wick consecutive rejection if it forms and closes during my valid trading hours.î On wick 3 -Sell limit filled, limit order pulled/expired if no fill on bar 3. The visual is displayed below. Figure 3: STSí microstructure principles applied with a three-wick set-up as an example. Definitions: Price delta (what we prioritise): net change in price. If price has increased by 100, the price delta is +100, and vice versa. Volume delta: the net difference between buy volume and sell volume within a spe≠cific time slot (as classified by the data feed, e.g., trades executed at the ask versus the bid). For example, if a candlestick has a volume delta of -1000, sell volume exceeded buy volume by 1000 for that single bar, meaning sellers dominated the volume in that time slot. There can be negative volume delta while price still rises, making it less important for our style. A short example applying market microstructureís knowledge to price (Mechanics of price). A wick high in a candle is rejected by the next candle, and it closes away from the highs. In this time slot (a single bar) on a liquid market, it is more likely than not that there was more sell liquidity present at that wick than the number of people willing to buy at market (more liquidity was provided by sellers than taken by buyers), showing buyer absorption and/or follow-through selling, multiple times. This is evidence that higher prices were not accepted within that time slot (timeframe) on the chart. It does not mean there was more sell volume relative to buy volume (negative volume delta). It means that more liquidity was provided than buyers could take, so prices failed to move higher, and selling is what followed (Negative price delta -What we profit from when short in this scenario). To simplify, regardless of how the ìorder flowî took place, in the short term selling pressure was in excess relative to what buyers were willing to take, or the liquidity provided to buyers via limit orders was in excess. Think of this like a wall that buyers failed to breach (absorption). Price tried to trade higher, but it failed to as seen in Figure 3. If price revisits that price or higher and fails again, closing. I want to sell at that price while anticipating a third rejection with the aim to absorb aggressive buyers. Long-biased examples of how one of our custom entry techniques (3 wicks) can provide a short-term edge: Why does it work and What is required? All it takes is for a wall of liquidity (sell limits and ghost liquidity) to hold it when buyers attempt to lift price, which can create a short-term reversal. If those buyers lose momentum and stop pressing higher, it is typically reflected in price .. By positioning ahead of time with a limit order near the suspected absorption zone, we can attempt to fade the move if the rejection plays out. It can also be used as confirmation (first) in conjunction with another setup. e.g., 3 wicks (success) followed by a different setup when price has traded lower (different entry technique or a different time-frame for mechanical execution). Defining ghost liquidity: Ghost liquidity is what we name private instructions to buy or sell in advance, set locally within trading software or algorithms (not on the exchange). These methods are used by larger market participants to hide their intent to buy or sell in advance by placing pending market orders and pending limit orders to fill large trades without placing limit orders on the public limit order book for everyone to see in advance (because that may influence other peopleís decisions or reactions). Example 1: Standard Practice (basic example) Price is at 20000 a larger participant want to sell 50 millon worth of an instrument at 20020 instead of placing a sell limit worth 100 million for everyone to see he may place a pending limit order so when the price reaches 20020 (best ask) he can automatically start placing multiple rounds of 5 million USD order sizes to get filled refreshing each time aggressive buyers sell into it offloading his position with lower market impact. This is called iceberging. Why does he do this? [1] It reduces slippage improving his cost basis. In plain terms it makes the price he gets out at more favorable reducing the costs of him trading. This is called iceberging. Example 2: Aggressive Price is set at 15000 and a large market participant wants to sell modest but high size let us say 10 million at 15050 he places a pending market order sell (emulates a sell limit without it being visible on the order book) this pending trigger allows him to start closing out his position as soon as the ask price hits 15050. Why does he do this? [2] His main concern is not the type of execution he gets it is that he gets out of the trade at a point set in advance without anyone knowing until the order is triggered. Consider this section as the surface of the Why part of our reasoning in this example. Model Variations 3WCT (3 Wick Counter Trend) In a market that is overextended (defined with other filters, price structure based (price action) or indicator (math) based), 3WCT can show an imbalance that may signal buyer exhaustion, which can lead to a shift in direction if buyers fail to present themselves. This provides a small edge when combined with a good filter, market, timeframe, and regime. Each individual component adds to the edge. 3WTT (3 Wick Trend Trading) In a long-biased scenario, in a trending market, when a small rebalancing event occurs in price (a pullback), if three wicks form on the low, it can show that active attempts by sellers to trade lower have failed. This not only offers a discounted buying opportunity if you position yourself at the potential point of absorption with a buy limit, but it also supports the underlying strength, showing there are still willing buyers within the auction. Consider this section as the surface of the what part of our reasoning. Isolated Application Explained (short positioning) Sell limit order fill, Bracketed with SL and TP (values known before the close), vice versa for long setups. Trading can be this simple. The focus should be on genuine market mechanics (microstructure), there is no room for fiction. Basis is everything. Most people who over-complicate with ìsmart money" or ìinstitutional" talk are waffling. ìIf you are using charts to execute, you are not smart money, but you do not have to be dumb money either.î Dismiss educator narratives on why their methods supposedly work and use critical thinking by applying market microstructure knowledge to accept or dismiss trading entry ideas manually. Any trader claiming to follow or mimic "smart money" is posturing; they do not have the equipment or connections to replicate their behaviour(s) furthermore, any unique actionable value is locked behind firm confidentiality agreements. Breaking it for a YouTube video or course is not worth the lawsuit and arbitration that would follow. Think about why price moves on a tick by tick basis and what the candlesticks you are basing your entry off actually indicate. Markets are not driven by patterns, they are ruled by imbalances between liquidity offered (passive) and liquidity taken (aggressive). Without it, price cannot move. If a setup does not have logic like this backing up why it would succeed enough for it to be profitable besides having random luck, you are wasting your time. If your only answer to ìwhy does it work?î is ìmy back-test says soî, then you are doomed. I have asked a trader why he believes his system works besides his data and silence followed for minutes whilst he tried thinking of what to say. I shown him random OHLC candlesticks with his strategy applied and he thrown in the towel. Do not be like this. Examples of what not to base your system on: ï Price cycle theories such as Wyckoff or Dow Theory (Refuted or true effectiveness cannot be accurately measured as it cannot be mechanically applied i.e unfalsifiable) ï Pivot points ï Anything that claims to directly follow Institutions or íSmart Moneyí. Even with complex order flow tools (which we do not use), it is usually an educated guess at best. The truth is large market participants are too good at obscuring their intent in real time to be followed around by retail, and by the time it becomes clear, it is frequently after the fact (often too late to benefit from). ï Fibonacci (based on faith and crowding) ï MA bounces (Random and seen on many data sets) ï Complex multi-timeframe analysis (hard to quantify and use with bar replay backtest honestly without hindsight fogging vision) ï Most well-known indicators for entries These methods are extremely random with weak foundations or are purposefully difficult to test accurately and honestly without overfitting. What is unfalsifiable is of no use to us. Educators push these techniques or similar for plausible deniability when systems do not perform. A bad strategy/model is hard to hold to account if they are 1000 ways to trade it. The use of multi timeframe analysis in trading is fine as long as it is not convoluted, has clear rules, and is tested rigorously. 6 Target and Stop Loss Placement Targets must be placed consistently. Targets are much more important than stops. Entries are more important than targets. Why? Because a strategy is designed to win, in short, it is designed to hit the target, not cushion the stop loss. This is regardless of the win rate that your profitable systems have. The better your entry is on average, the larger the RRR you can exploit it for as long as your trading costs allow for it. The idea that the best systems have either a high win rate and modest RRR or a low win rate and high RRR is not true; what matters is the highest average trade outcome after costs (expectancy). The better your target, the longer you can push average positions (if take prof≠its/targets are used). Stops are solely for risk management to automatically close positions when trades do not work out. Your aim is to make multiples of the stop-loss size per profitable position. If using price structures, e.g., ìsupport and resistanceî (S/R), then define the logic first, then the rules. S/R was a neutral example only; we do not use it. For instance, someone could use swing highs/lows, S/R, clustered wicks (over 3+ bars) or rejection zones. With fixed rules to define and mark them in advance. Price will naturally attract volume at these levels, even if the instrumentís order book volume does not reflect it in real time. Ghost limit orders exist, pending stop orders and order fill algorithms trigger from countless market participants for different reasons. It does not matter what happens when price interacts with these places. It is just more often than not that they are liquid areas. Avoid fixed-distance targets and stops; market volatility is dynamic. For example, a "100 dollar fixed target size" or a "20 dollar fixed stop size" is not going to work universally, even within the same price conditions (trend vs range). The market will not respect it, as volatility is not static. It is better to use dynamic yet consistent targeting methods. A trader must define fixed rules for determining what constitutes a valid target and what does not. A dynamic target, for example, could be 110 points for one trade, 160 points for the second, and 140 points for the next, All placed at predefined rule-based levels (examples are provided in Volume 2). Fixed targets overfit strategies easily. As stated earlier, your execution costs must be factored into your system. For instance, if you use a 1:5 RRR (an aggressive ratio), a minimum target of 100 points, a minimum stop size of 20 points, and a maximum spread on your CFD of around 2 points, that amounts to an instant cost of approximately 10.9% per trade. When adjustments are made to increase limit order fill rates, the cost is often even higher for market order execution. Rule example: ìMy target is always greater than or equal to 100 points on the Dow Jones. My stopís width is one-fifth of the target size." Why? Because it keeps costs at a modest level. 7 Instrument-Specific Rules Again, some markets behave uniquely. You may use existing research (find journals with related articles); a lot of this is defined more in quant-related journals such as JFQA: Journal of Financial and Quantitative Analysis rather than using deep statistics on our own, which are more prone to error and overfitting. ï Nasdaq trends strongly ï Dow Jones exhibits noticeable signs of mean reversion ï S&P 500 can be characterised as a drifting random-walk ï Gold is relatively erratic Note: These generalisations are confirmed by objective mechanisms explaining why they happen fundamentally, as the foundations of a generalisation are more important than the observation itself in markets; without this, It is simply a pattern, descriptive in nature and without predictive value. Mechanical examples are provided later within the material. 7.1 Entry Technique influence Examples: Example 1: If you want mean reversion or early trend entries, Dow is a better choice than Nasdaq. (It is more probable for Dow to reverse for intraday) Example 2: If you want to press trades or let positions run, Nasdaq is a better choice than Dow. This is because trends are more pronounced on Nasdaq compared to Dow for intraday. Either can have a trend or mean reversion model, but different strategies will tend to work better if aligned with the instrumentís nature (key). 7.2 Strategy Risk Management Setup Influence Examples: Example 1: If you have a strategy idea that includes rules to manually trail your stop loss in profit or uses large targets relative to stop size, Nasdaq would likely be a better choice compared to Dow. (Nasdaq trends more during intraday which compliments this idea; Dow tends to mean revert, reducing the potential for home run trades.) Example 2: If you have a mean reversion strategy idea with a hard take profit and stop loss as risk management (most common), the Dow would likely be a better choice, as its intraday trends are less pronounced compared to the Nasdaq. Either market can have trend and/or mean reversion characteristics, but different entry and risk management strategies will tend to work better if aligned with the instrumentís nature. These guidelines are not absolutes. Note: Trending means larger price extensions. Mean reversion means higher likelihood of returning to the average price. 8 Start From Blank Charts Instead of top-down start bottom up. Most traders look at charts for ideas, when we need to consult logic for inspiration and avoid recency biases from recent price action. Testing isnít about hunting for profitable strategies. Professionals use it to validate whether an idea works. Alignment is key. Before building rules based on the chart, define a hypothesis. For example, ìWhat if I traded reversals using a specific trading setup, with a 5 RRR limit-order entry, using Y filters on Z timeframe(s) across these markets (e.g., Equity Indices)?î Then test this on the charts. You are not trying to make it ìfitî, you do it to ask yourself: ï Does this work during valid hours? ï Does the visual match my logic? ï Does the reaction make sense knowing the true nature of price movements? ï Would my setup realistically hit the target often enough to net a profit over time? Only then can you write the rules to test. 9 Write Rules as If You Are Giving Them to a Machine Your rules must be: ï Objective ï Practical ï Not open to interpretation in real time ï Modest costs. For example, keep your costís impact on returns below 20% on a yearly sample, including slippage. The higher the expectancy achieved before costs, the greater the permitted impact of costs. ï Harsh but realistic implied trading costs in the backtest which should result in modest realised cost (real life costs). For example, if you risk $100 and your RRR is 1:5, but after adding spread, average slippages, and other costs, then your new effective RRR after accounting for costs becomes 1:4, which means you only make $400 per winning trade but always lose around $100 before slippage. ò 20% instant costs; in many cases this is too high, so upon discovery we would actively make attempts to reduce costs without overadjustments, such as lowering the ratio on static strategies or increasing the minimum stop size to the intraday bid-ask spread ratio. The following are some examples of bad and good rules. Bad Rule: ìIf the market is ranging, I do not trade.î There is no objective way to consistently identify a range in advance; every range is different. By the time a îrangeî is noticed, it is often too late to gain an advantage. Good Rule: ìIf a 3-wick setup forms between 3ñ9pm GMT time, and the high/low of the setup is beyond/below my filter, I will place sell-limit at the top wick or buy-limit at the low wick.î This rule is not based on intuition and is intuition-free. It is systematic. Define everything clearly: The conditions of execution, the logic, the filter, and why. Our strategies emerge as mature when the prototype trading setup has a set ímechanical sequenceí. To us, the mechanical sequence is a set list of instructions that lead to a limit order or position on a given market, which later develops into a more nuanced system with every possible setup (and each what-if) considered before deployment for predictability and fair testing. We recommend the use of Microsoft Word, Google Docs or Notion for the documenting of each trading strategy once you start building upon basic foundations. Until then basic notes are fine to save time. 10 Stress-Test the System by Breaking It Once rules are written, test them brutally. Ask yourself: Is this rule based on logic or emotional comfort? Be emotionally detached (e.g., break even or partial profits may reduce a strategies net profit -so why use them?). Partials or break-even reduce strategy expectancy more often than not. Does it work over 3+ months of data? (The length of the backtest depends on time frame). Log the data and process it in the backtesting sheet available for free in our Discord. For instance, each day has a number of losses and wins, and you can aggregate them by writing them like so: -1R+4R-1R-1R, in each cell. Essentially, just write all of your data down neatly so you can analyse it later; see Fig. (4). Figure 4: Spreadsheet filled out with each trading days losses and wins to be used for further analysis. What if market conditions flip? Test on conditions against the systemís nature. Test mean reversion and reversal systems on trending weeks. If you are using trend trading systems, then test them on mean reverting/ranging weeks. See your system struggle. An extremely basic test is shown in Fig. (5). For example, August 8th to September 13th, 2024, on mean reversion systems for YM/Dow Jones is a good place to stress test due to the relentless intraday trends exhibited. Instruction on how to identify these periods are provided later in the Mechanism Validation section. What if trading costs rise 20%? Then the size of profits reduces by around 20%. 10.1 A simplified thought process when considering strategy op≠timisation. îAfter the initial rejection candle close, if there is an additional rejection, should I scale in/increase the risk on the trade? The second entry looks like it has a higher win rate as compared to the first when scaling in for my system.î. Testing will confirm whether it is worth doing. Scaling in is only worth doing if the win rate of the second entry is superior to that of the first. For example, a 45% win rate second entry versus a 40% win rate for the first. Most systems do not benefit largely from it, so be careful. The scaling-in ideal The trigger for scaling should ideally be when the entry mechanically shows signs of success, e.g., higher prices followed by a valid entry when longing, and vice versa, or persistence, e.g., price rebalancing at your entry area. Remember that scaling in should be looked at after the base strategy is developed as an optimisation, it is not a requirement. Do not force it. Rules must be consistent and always defined in advance, and drawdowns must remain acceptable in tests. A Standard example of application: an entry is an individual trade execution. Each entry has 1R risk. Two entries would have a risk of 2R, so for 3% risk, that gives 6% total risk if two consecutive trades are running. A Prop firm example of application (more useful): an entry is an individual trade execution. Each entry has 1R risk. Two entries would have a risk of 2R, so for 0.5% risk, that gives 1% total risk if two consecutive trades are running. Distribution of returns and prop firm dynamics The reason why scaling in can be more useful for prop firms is that risk can be spread across multiple positions to benefit from strategy successes, whilst naturally capping trading activity when performance drags. This can help increase the chances of passing, while indirectly slowing drawdown accumulation, especially for trades with aggressive stop sizes. Warning: Some prop firms have rules against scaling in and frame it as îone-sided betsî or use other language against the traderís best interest. If you choose to work with a prop firm, make sure you pick one that does not have such rules when deploying strategies with scaling-in behaviour. I wish I could call out individual firms, but my goal is to guide you safely without creating legal risk. In professional trading environments, scaling is a standard practice; do not let retail prop firms manipulate you, trade elsewhere. In your strategy design process, ask yourself: ï Should I hedge or wait until my position is closed to enter setups on the opposite direction? ï Is it worth holding overnight? ï Do I have enough leverage/margin to trade this strategy on my broker or prop firm of choice (find out the leverage needed maximum per trade with percentage stop distance relative to the percentage risk per trade desired) You are not seeking perfection, you are seeking robustness. If a small change breaks your system, it is most likely due to overfitting. Types Of Trading 1. Discretionary trading Discretionary trading is an approach where an individual trader makes decisions to buy, sell, or hold assets based on their personal judgment, analysis, and experience in real-time, rather than following pre-set rules. Discretionary trading causes decision fatigue and emotional strain for most traders, which often produces poor results. 2. Mechanical trading A rule-based method for making trading decisions, using pre-set criteria to produce buy or sell setups, manage risk, and execute trades. The primary goal is to remove emotion and human bias from the process, leading to more consistent and objective trading. Mechanical trading causes less emotional strain and little to no intuitive decisions. This is what we prefer. Why do we have a preference for mechanical, aside from the psychological benefits? Discretionary trading is viable but cannot be measured accurately, so I have never been seriously interested in including intuition in short-term trading. 1. Modern liquid markets in the short term discover prices fairly consistently: imbal≠ances form, price drifts occur, corrections take place, and undercurrents from news influence action. 2. These are the physics of short-term price discovery. Since this is largely statistical, discretion usually adds noise in decision-making, often eroding the chance of success instead of providing an edge. 3. Prices are fundamentally bound by changes in the state of liquidity, and your strategy, if designed correctly, is simply dealing with changes in market state. Most îdiscretionaryî trading strategies rely on intuition; If two traders use the same strategy, they are likely to achieve different results. One trader could make money, while the other could lose money, making the íeffectivenessí subjective and open to interpretation; every traderís brain is different. Another key problem with discretionary and intuitive strategies is that they try to combine multiple systems together subconsciously. This interferes with profitable outcomes systemically, and here is how: A trader could have a $10000 account trading two different trading setups, discretionary or mechanical. One setup could make 30% while one loses 10%. If we calculate this simply, the ending balance would be $11,700. This is nuanced, but the example is generous, as it does not include the additional costs the losing strategy incurs. Regardless, the returns have dropped by over 10%, which was avoidable. Let us explain. We isolate our risk to one strategy at a time to amplify gains and partition our accounts. We could run two separate accounts for each setup, $10,000 each: one account makes 30% ($3,000), and one loses 10% ($1,000), ending with a net profit of $2,000 instead of $1,700, a 17.64% increase in profits from one step, risk isolation. We use optimised methods to do this properly. The worse a setup performs relative to another, the more money that is lost to the drag another setup may introduce, randomly threatening your best setupís profitability. The more positions that are taken, the worse the drag on returns becomes. Note: When this is applied, this can be on uneven amounts, for example, instead of risking 2% per setup on a $10,000, you can have two $5,000 partitions (separate accounts) risking a value close to 4% per trade. This is just an example. There are more precise ways to achieve the benefits with less capital. If a strategy breaks even, one realises 10% in gains, and the other loses 10% on $10,000, the ending balance is $9,900, not $10,000. If isolated, you would have saved 1%, and these small discrepancies add up over time. The potential benefit can be exponential. That is the power of risk isolation. But you can only gain such benefits realistically with a mechanical trading strategy. Mechanical trading strategies have fixed preset rules that are planned ahead; If two traders followed the exact same rules, they would get the same trade setups, targets, and stop losses, resulting in similar outcomes, making the effectiveness objective if the testing data were of high quality. We explore this in more depth in the psychology section. What I have revealed above adversely affects returns of millions of traders, and they most likely will never notice. This is one example out of many measures we have implemented. We have fine-tuned each part of the trading process, adding to our advantage. For the minority of traders with an edge, trading costs and other hidden expenses undermine their ability to extract meaningful financial gains. After our first profitable year, we sought to increase efficiency in all areas possible. Bonus tip: When in Doubt, Zoom Out Ask yourself: Does this decision happen on every trade? If yes, write a rule. If not, STOP, think, and evaluate the logic. You should: ï Know your risk percentage -make a rule ï Know your stop -make a rule. ï Aim to know target, stop, and entry price(s) before the candle closes. Bracketed limit orders help a lot. Figure 5: Extremely basic test. Old testing data shown from 2022. No edge is possible on this chart, see Fig. (6). It is 100% a random walk and is eerily very similar to a real market. The market is not perfectly efficient, but it is remarkably close. Therefore, you need to be refined in your approach, you need to be accurate, you need to be systematic and calculated. Figure 6: Completely random-walk chart example. No edge exists here. 11 Structure Structure before everything. Logic before data. Consistency before optimisa≠tion. Logic . Rules . Data . Optimisation (Idea-driven adjustments instead of curve-fitting based solely on favourable data) Always ask ìwhyî before ìwhatî. Every rule is based on: 1. What you can realistically do 2. What the market allows (e.g., scalping CFDs is usually not a viable strategy due to higher or exaggerated costs on higher lot sizes) 3. What yields clear, repeatable decisions 23 STSí Interpretation of Market Behaviour (Simplified) Most markets on low timeframes tend to mean revert, but some markets exhibit sharper trends, which naturally spill over into lower-timeframe price action (more trendiness) when you research you will learn the way it behaves and why (just as important). Short-term mean reversion, such as that shown in Figure 7, is a result of efficient trading. The purpose of Figure 7 is to sharpen your perception of price movement(s) before exploring more nuanced trading ideas. Figure 7: A diagram of isolated market behaviours we recognise (simplified) Aggressive liquidity: The aggressive liquidity panel I designed shows market orders that are large enough, as a crowd, to influence price by moving it through successive price levels, and the arrows show the immediate movement until opposing liquidity absorbs the flow, as shown in the passive liquidity panel. Green represents buying that takes offers (the ask side), where a buyer interacts with a sell limit order to get filled. Red represents selling that takes bids (on the bid side), where a seller sells into a buy limit order to get filled. The price moves quickly because these orders exceed what is instantly available at the closest bid or ask (the current prices/best quotes), so market participants consume the available liquidity and the price continues moving up or down until there is enough liquidity on the other side for the price to pause again when passive liquidity (limit order volume) exceed aggressive liquidity (market order volume). Passive liquidity: The green move shows buying pushing price up into a level where sell limit orders are sitting. The thin grey line (ìPassive sellersî) is that sell limit wall/layer. When price reaches it, trades get filled there and further upward movement needs more buying to clear whatís resting at that level. This is the true cause of every single wick we see on charts. Market price efficiency: The price path I generated aims to show normal back-and-forth trading around a reference price. Buys take passive liquidity, pushing price up, and vice versa, and new limit orders (passive liquidity) keep appearing, leaving a tail at the top and bottom. This is materially efficient because price does not run in one direction for long. If market activity remains one-sided, you get drift where the price is skewed toward one direction (a trend) which is inefficient. In Figure 8, we have generated a tighter ranging series to intensify efficiency. Rebalancing: The parabola (curve) shows one direction pressure that lasts for a period, then stops. While it is active, price is pushed away as the liquidity is taken and then when it ends, price drifts back below as the sell limits restore and the buying is not large enough to result in higher prices. Summary: The aim of this diagram was to show the foundations of price movement itself, to dilute the comforting idea of îbuying pressureî and îselling pressureî, as it is oversimplified and almost always undertells the story. The talk sounds good but it provides no edge. One hundred units can be bought, two hundred can be sold, and the price can still move up, which shatters the standard retail narrative of îmore buyers coming inî, îweaknessî, and îstrengthî. Subsequent examples and visual data will illustrate these dynamics in application throughout the material. The Purpose Creating ideas based on sound logic is not just to avoid over-optimising systems that will not perform well in real-time or forward tests. The point is that by using the logic of established market truths in your trading, if your idea or theory is just as sound, it should perform well; testing is used to verify the edge instead of finding it. That is key. This is the ideal mindset before testing. "This makes sense; my idea(s) should work in one of these markets." Remember, this is before you approach a chart for testing; this is in the development and thinking phase. Ideas should be tested on a couple of instruments at most that align well with the idea to see if the approach I am developing is effective post-development. For example testing GC and HG (Gold and Copper) if the idea(s) apply to metals in general or a specific established behaviour present in an individual market e.g., strong trends with a logical underlying cause. Before we explore the consequences of deviation we must go over the general principles of deductive and inductive reasoning and its importance in strategy development. ï Deductive reasoning (in STS strategy engineering) Deductive reasoning starts with general rules or assumptions and works down to specific, testable outcomes. When you are developing your own trading strategy, this looks like taking a well established falsifiable market theories (for example, ìmean reversion tends to occur after extreme deviationsî -Supply and Demand related.) and using that as foundations to turn it into concrete entry, exit, and risk rules. And only after that, we check whether the strategy behaves as the theory predicts when you run it through testing such as backtests and forward testing to confirm its efficiency. This is our primary and preferred reasoning path. ï Inductive reasoning (in STS strategy engineering) Inductive reasoning starts with observations in data and works up to a general rule or hypothesis. In trading strategy development, this is looking at historical price data, noticing a repeatable behaviour (for example, certain strong trending, mean reversion or reversals around specific conditions), then generalising it into a strategy idea that you formalise into rules and test, knowing that the pattern might weaken or disappear out of sample (this is typical, which is why testing is important). This is why we prioritise deductive reasoning paths to find trading edges, as this protects us from confirmation bias and overfitting. Before taking any of these discoveries we pin it to something real, if there is no reason for why it happens e.g., market opens, indirect but consistent market activity underpinned by something that fundamentally happens. This is our secondary more nuanced reasoning path. Common Consequences Of Deviating Overfitted Systems: Overfitted strategies perform well on historical data but do not hold up in forward walks (forward test and live trading). They work well on the backtest but blow up live Overoptimisation: Overoptimised strategies have a long, sometimes convoluted sequence to get the valid trade. For example, data snooping across multiple time frames (using multiple time frames inconsistently) or using multiple different confluences but never the same exact ones and sequence for every setup. This leads to low sample sizes, which can create an illusion of efficiency over long time horizons. This is a data issue, but in real time the system can be extremely sensitive to market changes depending on how it is designed. Optimisations like confluences should only be taken into account as a secondary step. Optimisation of a tested trading idea should only happen to something that already has an edge when isolated, and that should be done with rigour to avoid curve fitting. Your job is to test ideas that tend to work. The second you try to make something work is the second you start ruining your strategyís quality. Strategy engineering should as close to a deductive process as possible in the beginning that is built upon with inductive input selectively later. For example, if X method was effective as a standalone, you would build from there. If it were effective to push it further, you could look for a pre-session higher time frame imbalance as a filter (if you want to use multiple time frames), and the X setup would be a buy or sell trade aligned with said imbalance . But if X method was ineffective on its own, you would be fitting it to the market, which rarely leads people to real market edge. Testing data is used to validate the idea. The Suboptimal Way: Data says it will work; the market may have accepted it, adjusted to look good, has inconsistent execution patterns. Many traders use it, so the probability of persistence is likely close to 50/50 in the short term, with market crowding further reducing it. The Goal: Good logic (first), data accepts it, the market accepts it, no over adjustments, static execution patterns, nobody else trades 1:1 like you, probability for persistence is skewed in the strategyís favour. To us the reasoning behind why a trading idea should work and persist is more important than the testing results itself. If both of these check boxes are not ticked, we do not proceed. You do not optimise solely to improve win rate or net gain. You optimise to enhance the logic behind the system, which often translates to improved performance (net gain). Yes, the first 0ñ20 hours (first few testing sessions) will feel foggy. Then it clicks. You will never know if it works until you test it exactly as written. That is when the market becomes your teacher. If a system implodes/stops working, it does not mean a different variation of it cannot work again in the future. When a system underperforms for a while, it is called "perfor≠mance drag". This phenomenon is normal and recognised in institutional trading. When the market is in a similar regime you can re-test old systems to see if it is working again, if it is, you can re-deploy it. Do not aimlessly discard systems or strategy hop. 1. Keep strategies you choose not to run in an archive. 2. Put your strategies that stop working on ice and defrost them for use when it is plausi≠ble that the market could respect the idea. For example, favourable recent performance followed by market characteristics that the strategy relies on is presenting again to confirm validity (objective signs of effectiveness). Never Dismiss Data Blindly, Always Review It. Disclaimers like ìpast results are not indicative of future performanceî are present mostly to minimise an authorís liability. In every industry, equivalents to backtests and forward testing exist, and they are statistically robust protocols to mitigate issues that degrade the data, such as overfitting and biases. Multiple steps are taken to make sure the outcome is as objective as possible, because money and reputation are on the line. If the data has not been skewed, has not been overfitted, and there is a clear mechanical reason why the effect should persist, it should be taken seriously. Data is everything, regardless of its mechanical or intuitive application. When you have reviewed high-quality data and you can forecast something, that is how you gain an advantage. Although many might forecast the same and be correct, a large crowd are not with the drift: they have inferior data, poorer execution, or they operate in saturated, crowded ways. 1. As individuals, we do not move the price. We can run unique strategies, with our own balance, sizing, and execution patterns. 2. We often have access to flexible leverage and can tolerate higher return volatility, as long as we manage the downside. Our risk management and entry methods often capitalise on that. 3. We keep trading costs low and only step forward with well-tested models. Most retail traders tick none of these boxes, that is our advantage. When we pocket the difference, we gain an edge (positive expected result +R). It does not guarantee success on every trade of course but over many trades the expectancy can assert itself. Reality sets the limits, so we stress test our assumptions against data and stay anchored to what is measurable (falsifiable). Most traders are not serious about engaging in research or reading at all; that is also part of your advantage. Operating in trading without data is like operating as a business with zero market research. Figure 8: A stationary (mean-reverting) series is shown on the top. A persistent (trending) series is on the bottom. The demonstration of Augmented Dickey-Fuller (ADF) is for educational purposes only, we do not try to detect short-term price regimes and use it as a signal, we anticipate them happening indirectly in advance through testing and preparation. Example: If it is normal for price to trend on that specific intraday session, we position ourselves to benefit (our strategy rules do this without requiring us to make any intuitive decisions). Augmented Dickey-Fuller Note: When applied to a chart, these statistical tests only show what is already happening, which everyone else can see. They are descriptive, not predictive so there is no edge in its use for price forecasting. Where it was useful for us: It is a research aid Ali has used in the past to better understand market state through legitimate, well-established processes in mathematics rather than relying on faith. In this case, we moved from observations to rigorous validation and ended with a strong generalisation, which provides us with confidence moving forward (inductive reasoning path example). You do not need to use these tools; we use visuals to wire your understanding as well. When applied to larger datasets (optional and advanced), it can be insightful to know how often a market is typically ranging intraday during certain hours. However, this kind of data mining is only useful if it is falsifiable and backed by something that fundamentally occurs within markets regularly, in other words, a mechanism, for example, consequences of market opens, midday activity, and well-established institutional activity (such as repricing events on metals) as supporting evidence. We do not try to detect short-term price regimes and use them as an entry. Instead, we use testing with an aggressive cost model applied to design strategies intended to perform under specific market regimes, and gain insights on how we approach this throughout the material with useful steps and simplified paths so you know how we went from generalisation to conclusion for each problem (deductive reasoning). We recommend that most people avoid this advanced form of testing, as it is very easy to overfit, and it is not required (we no longer do it). The underlying cause is always the most important. If you believe you have found a specific behaviour, it is important to have a maximum of three real, objective mechanisms for why it happens. This step assesses authenticity rather than existence. Your Path Forward If you already have a trading strategy, we suggest pausing for now. Do not drift; implement the parts of our process that resonate most with your strategy with notes, then come back after you have successfully done so before moving on to Strategy Engineering Volume 2 to retain your momentum. Reading is good, but your role is to shape results. Make this your work environment. Profitability demands application. Your time matters. Our material is demanding because the market is demanding. The public materials are structured to get you competent as quickly as real learning allows. There is no filler. You will use some parts and skip others. The strategy engineering process is about you and your constraints. If you arenít a full-time trader, you must continue reading until you fin≠ish the material and revisit our work when applying the process or ask us directly for assistance. Added Annotations (Notes) Note 1: The specific ratios do not matter. You should not be curve fitting/overfitting your system (trying to find the best ratio). To elaborate, the logic in the example behind using 3-4 RRR in continuation trades is that you should allow for larger movements against your entry because you are entering in the middle of a trend. For example, when trend following, if you are buying, you are executing at premium prices, not at discount prices. More space for error is required. And 4-5 RRR for example, is encouraging tighter stop losses relative to the target for reversals because you are actively going against the short-term trend. The ratios given were example ratios; you can change them based on your ideas. Note 2: We aim for our "Implied Backtesting Costs" to be higher than the "Realised Costs" from real time trading. We assume the costs paid are always at the extreme for strategy robustness and reduced sensitivity to collapses in real time. It is one of many ways we put strategies under stress when testing. Note 3: When I say íconsult logicí, I am referring to market microstructure, which I mention in the document primarily, but it is also about rejecting ideas like îMA bouncesî and îFibonacciî, which are not logical reasons to engage with the markets. Definitions 1. Constraints What limits you -time, capital, lifestyle. These set the boundaries for what you can actually trade. Your system must respect them. 2. Market Type Behaviour of a market: mean reverting, trending, or random/alternating. 3. Valid Trading Window The hours when youíre allowed to trade. Based on where volume and volatility are. 4. Risk (R) The set amount of capital youíre willing to lose per trade. Fixed, consistent. Example: 1R = 3%. 5. RRR Risk-to-reward ratio (e.g., 1:3 = risk $100 to make $300). 6. Price Mechanics Price moves because buyers and sellers are imbalanced. Thatís it. It explains rejections and moves -itís not an edge, itís just reality. 7. 3-Wick Setup Three wicks rejecting a level -signals price has repeated selling activity and wonít break through. Must be rule-based without room for interpretation. 8. Tick The smallest price increment on an instrument. 9. Execution Cost Spreads, commissions, and slippage affecting net performance. Ig≠nore it and your edge vanishes. 10. Backtest Testing your rules on past data. Done honestly, no scrolling, no cherry-picking, no hindsight. Bar Replay below in 13. 11. Overfitting When your strategy works only on the past because youíve shaped it to work on past historical data instead of applying and idea to historical data. Looks good in testing, fails live. 12. Stress Test Deliberately run your system in bad conditions. These are notable periods of intraday chop, low volume on trend trading strategies and periods of relentless trends on mean reversion/reversal strategies. If it collapses, itís weak. Example: Someone could be running a mean reversion day trading system on YM and he could stress test August 8th to September 13th 2024 as an example, where, here Dow Jones exhibited strong trending behaviour which is against the systemís nature. 13. Bar Replay Play charts forward candle by candle to mimic real-time. Helps you test if youíd actually take your setups live. E.g., TradingView Bar Replay 14. Scaling In Adding size after entry. Must be planned and tested -not just done because ìit looks goodî. 15. Hedge Open a position benefiting from movements in the opposite direction. Useful at times, but messy if you donít have clear rules. 16. Breakeven/Partials Closing part/all of the trade early. Often reduces long-term edge unless justified by data. 17. Ghost Liquidity Orders that arenít visible but sit around visible levels. Cause sharp reactions or none at all. Itís just a surge of liquidity that isnít visible on the books. 18. Random Walk Price sometimes moves like noise. Most patterns donít work unless theyíre backed by logic. A Random Walk is a market that is 100% random. In other words, it is effectively a completely efficient market where no edge is possible. Real markets are of course different. 19. Bracketed Limit Orders Pre-set entry, stop, and take-profit. Forces discipline. Removes intuition in real time. 20. Institutional Narrative Fallacy The idea that ìsmart moneyî always leaves clues. Usually marketing fluff. If itís not testable, itís not valid. 21. Data Snooping Repeatedly looking at a data series from different angles to confirm something that you havenít defined ahead of time often leading to insignificant and/or biased discoveries. Essentially looking too hard for patterns and finding things that donít actually repeat. Typically kills forward performance. 22. Drawdown How far your strategy drops from peaks in tests. Crucial for knowing how big your positions should be in advance. For example, a trader could have a max losing streak of 8 but your peak to trough could be 12x your risk (some wins followed by strings of losses repeatedly create this) ñ Super important to track and know. Thatís the maximum drawdown you should be taking into account, especially if working with prop firms. 23. Dynamic Targeting Set targets based on real market structure -swing highs, lows, clusters of wicks. not arbitrary price movements e.g., 100 points, 100 handles, 100 pips, 100 ticks. Market is too dynamic for a one size fits all. 24. Expectancy The average gain or loss per trade. Strategies donít need high win rates -it needs consistency in the data and logical backing: Expectancy = average win ◊ win rate - (1 - win rate) ◊ average loss. 25. Logic-Driven Rule A rule built on how the market behaves -not what a shape on a chart looks like or some untested theory. For example purposes only, using the 3 wicks example. Bar 1 closes with a wick high; this shows that there was selling pressure. If the next candle interacts with bar 1ís high but fails to close above, creating another wick, it shows continued selling pressure. If on bar 3 it happens again, it shows compounded selling pressure. If it reverses, it should do so quickly. If price continues beyond the wicks, price should continue trending. Using a small stop loss relative to the target can create an edge if costs are managed properly. References ï Figure 6 generated by Ron on randomfx.net (no longer exists) ï Other figures were generated using Python, lead by Ron and overseen by Ali. Minor Researching Opportunities (Optional) ï Jim Simons interview https://www.youtube.com/watch?app=desktop&v=QNznD9hMEh0& t=0s&ab_channel=Numberphile2 ï Data-snooping links: https://en.wikipedia.org/wiki/Data_dredging https:// quant.fish/wiki/data-snooping-in-algorithmic-trading/ https://www.nber. org/papers/w3001 ï Order-flow references [Reading 5]: https://www.investopedia.com/terms/o/ order-book.asp https://www.investopedia.com/terms/t/time-and-sales.asp#:~:text=What% 20Are%20Time%20and%20Sales,trade%20orders%20for%20a%20security. https://www.investopedia.com/terms/p/pricediscovery.asp

Strategy Engineering Volume 2

Mechanical vs Discretionary Trading: Clearing Up the Confusion Sentient Trading Society Strategy Engineering Volume 2 Ron & Ali Introduction There is the common idea that floats around endlessly; that discretionary intuitive trading means youíre being flexible and smart, while mechanical trading is some rigid, one-size-fits≠all system that ignores the market context. That is an oversimplification. Mechanical systems can be flexible Think of mechanical systems like flowcharts or decision trees. They can include filters for volatility, time of day, higher timeframe context, session structure basically, anything you want to build in or as many nodes as you want if weíre imagining a flowchart/decision tree. You can even bake ìdiscretionî into a mechanical system if you put in the work. Yes. Really. Traditional discretionary trading, by contrast, often feels smart because youíre calling the shots in real time with intuition. But if you donít have clear rules backing your decisions, youíre prone to what I call Discretion as Reactive Price Making -Ali You are far more susceptible to subconsciously or consciously registering and responding to recent stimuli (the last few trades), recent candles, sharp swings, or your overall performance. All of this is just noise. This is not a structured or a tested logical approach. By acting this way your trading reactions can exhibit recency bias [1]; how many traders reinforce this bias is through post-trade analysis journaling, and emotional trading could be masked as an exact, rigorous process. 1 This is Dangerous. If it is not testable logic; it is reactive bias masquerading as insight. Traders often do this with a post trade analysis journaling process. What you could be doing here is letting your natural pattern recognition (hu≠man biology) override logic which leads to you overriding the process of trading with your instincts. You may think this is not the case, but you must realise that your pattern recognition will come first, and you will try to form some sort of logical reasoning as to why you saw such a pattern emerge on the chart. This forward-looking subjectivity on forward walks [2] leads to a lack of robustness and introduces a severe amount of fragility into your trading. -Ali Every personís brain chemistry is different, so the only person who should participate in a strategy that relies on intuition is the person deploying it. The system may íworkí for that discretionary trader, but intuition-led success can only be replicated coincidentally. It is not objectively reproducible as nobody shares the same Central Nervous System and experience. Analogy: A discretionary trader adjusting to market noise actively or passively is like a Mechanical trader changing their system to produce better results in a back test (curve-fitting), but instead of overfitting a back test, it is your human biology (pattern recognition) pulling the strings on a walk forward. And thatís just as fragile for your systemís frame. Summary: Using intuition does not automatically make you smarter. Without clear, tested rules, it means more often than not for most traders that theyíre trading messier. Traders occasionally find success relying on intuition, but they remain rare out≠liers in a market that rewards systematic logic. Your system doesnít have to be robotic or rigid. But your decision process needs to be accountable and repeatable. Otherwise, youíre apply≠ing guesswork to some of the most efficient markets in the world. Recency Bias [1] -Cognitive Bias when someone favours giving weight to recent behaviours whilst ignoring or downplaying longer term trends influencing trading behaviours. -Basic example. A trader stops trading Wednesdays because the last 6 weeks have had losing Wednesdays but the strategy data over years has been net profitable on Wednesdays. Forward Walk/Walk Forward [2] -Future price action and trading referring to real time trading or forward tests. Curve Fitting [3] -When a strategy is tailored to fit past market data. When a system is tweaked to get better results on historical data. Why Confluence Can Hurt Your Trading Sentient Trading Society Ron Understanding The Outcomes of Using Confluence Confluence is often seen as a way to bring more certainty to your trades, but it can actually introduce unnecessary noise and complexity. Relying on multiple factors to "confirm" a trade can create random elements in your system, leading to more variance and overfitting. It reduces your trading frequency, which might make your system look better in backtesting, giving false confidence in real time, but that is usually not the reality; Confluence should only be used when backed by tangible data, and never faith. For example, imagine your system would normally generate 100 trades in a backtest, but once you add confluence, the number of trades drops to 40. With fewer trades, your results are more likely to look great, with higher win rates or profitability, but this is just an illusion. As you extend the backtest, the performance will likely decay, and you will see that it was not as effective as it first appeared. This is the confluence fallacy. Traders often think adding multiple layers of con≠firmation improves their system, but all they are doing is overfitting it to certain conditions. The data looks promising in the short term, but it falls apart as you test further most if the time. Lots of traders get caught in the trap of confirmation bias or multiple time-frame top-down analysis. They rely on confluence from all kinds of sources, whether it is data-backed or intuitive; this just complicates or obfuscates the process and makes backtesting less reliable. In fact, educators often push this idea of using multi-timeframe analysis. This intentionally makes the strategy harder to backtest and leaves room for plausible deniability when the system fails, and so does using a system with discretionary influence. These added layers of complexity are pushed because they make it harder to expose ineffi≠ciencies in their systems. When these educators sell you a strategy with multiple layers of confluence, it is not because it is the best way to trade. It is because the added complexity makes it difficult to see that the system will ultimately average out and fail over time, or to add confidence to the user. This creates a veil around the strategy, hiding its weaknesses, and making it look like there is something valuable there when there really is not. This is why simple indicators systems and so on, get ridiculed as the veil is easily removed. 1 Ideally when you are on the right side of the market, all the confirmation you need should already be in the price action. That is why I use limit orders to execute. I do not need extra confluence. I just need a solid system and the discipline to stick to it. Especially when shorting, if you are waiting for extra confirmation signals, you are already behind the market. It moves quickly, and the more confirmation you need, the more likely you are to miss opportunities. So do not get fooled. The idea that more confluence equals better trading is a trap. Keep it simple, trust your system, and avoid overcomplicating things just because the ìexpertsî tell you to. 2 Summary Do not use confluence or multiple timeframes just for the sake of it. Be vigilant when analysing the intent behind what you are shown over data. 3 Additional Reading Data Snooping (significant) ï Definition and understanding: https://en.wikipedia.org/wiki/Data_dredging ï https://quant.fish/wiki/data-snooping-in-algorithmic-trading/ ï Lo, A. W., & MacKinlay, A. C. (1990). Data-Snooping Biases in Tests of Financial Asset Pricing Models. The Review of Financial Studies, 3(3), 431ñ467. ï https://www.nber.org/papers/w3001 2 Multi Timeframe Analysis: 3 Trader Examples Sentient Trading Society Ron Multi Timeframe (MTF) Analysis: 3 Trader Examples Traders must stick to a single setup on a single timeframe in the beginning. Multiple time-frames are often an unnecessary complication and easily result in data snooping [5]. See MTFs as a potential optimisation to optimise a strategy, it is not an essential complication. Using multiple timeframes is okay if it is used in a consistent and controlled way [1]. Example: a trader could consistently look at three timeframes, always top down or always bottom up, but consistent [1]. 1 Trader A: Day-Swing Hybrid (Higher to Lower) Trader A uses daily charts, hourly charts, and executes on the 5 minute chart. There are absolutely no deviations. There is nothing open to interpretation and nothing intuitive. There are no thoughts such as ìsometimes I start on this timeframe first, sometimes this one.î If there is variance where he starts, it must be because of a predefined, logical, tested rule. For example, Trader A, who day-swing trades S&P 500, might only short on the hourly timeframe on ES because S&P 500 is unlikely to trend downward for longer periods of time. He does this to reduce his holding times. Factual and logical. Acceptable. That is why he mixes timeframes [1]. 2 Trader B: Day Trader (Lower to Higher) He uses the 5m, 15m, hourly, and daily. Just like Trader A, he uses multiple timeframes on a tested system with no deviations. Nothing open to interpretation and no discretion. Trader B looks for setups on the 5m and 15m charts. When his setup presents itself, he enters. Trader B trades reversals exclusively and has rules for it. Trader B looks on the 5m and 15m charts for entries. Whichever presents itself, if he gets a setup on either timeframe, he will execute the next on the other timeframe too. For example, he gets a buy on 5m first. He enters with 2% risk, and for 15m he enters 2% risk on that trade too (he can frame it however he likes as long as tested), these are for example purposes only. 1 Trader B uses the hourly timeframe for manually trailing his stop loss behind lows (he has a specific ruleset for it). Trader B uses the daily timeframe to set hard targets for his positions. He runs the winner on the hourly and has a hard target using the daily. That is why he mixes timeframes. This sounds fine [1]. 3 Trader C: The Disorganised (Like Most MTF Analysis Traders) He trades most timeframes, from 5m to weekly. Trader C trades Forex USD pairs. He looks at a large set of data aimlessly. Sometimes he will start at the Dollar Index (DXY) to measure things like ìstrengthî, or the pair he is trading, usually high timeframes such as daily or weekly to get ìbiasî, COT reports, or any data needed to confirm his biases. He tries to establish a ìnarrativeî or draw an ìinstitutional storyî through usu≠ally vigorous, although sometimes meticulous, data snooping [5] (repeatedly look≠ing at the same dataset to get a desired outcome), for example, looking at higher timeframe marking points or connecting structure inconsistently. He can justify why he is doing X and Y in real time using words. Let us give him the benefit of the doubt here. Maybe he is even backing it up with logical principles like order flow mechanics. He has a hard time putting it into fixed rules. There are many factors that can influence his trading, some not accounted for in testing. Trader C usually does not backtest properly. If he tried, it would be hard to be unbiased or accurate. It is easy to get a false sense of security in the system since future data visibility can influence backtesting systems of this nature (even if subconsciously). It is very dangerous. Conveniently, this flaw makes his system harder to unveil as ineffective as well (it usually is). Trader C could consistently start on, for example, the weekly, but does not consistently mark the chart. He is looking for something, but it is not linear. For example, Trader C could use fixed, predefined rules to look for something specific on the daily or weekly (X), look for something specific again on the hourly to support what was identified on the higher timeframe (Y), and execute on the 5m using a specific entry method [Z]. The problem is, for Trader C, (X) and (Y) are usually inconsistent, sometimes (Z) too, or he is looking for many setups instead of a single setup, which introduces noise to the trading method (results can depend on what comes first). For example, a retail trader could trade based on Entry technique 1, 2, and 3 lined up on multiple timeframes, but if he has seen technique 2 and 3 first, he could act on technique 4 instead of 3. Or he could enter technique 3, get stopped out, and miss technique 4 if he is rigid either way. This has many logical fallacies. I want you to feel the chaos, the lack of order. When I see traders operate like this, I think to myself, has he forgotten he is analysing a dataset or trying to sell the market an idea or pitch a sale. They look lost in the noise. Naming All The Biases Trader C Suffers From Also remember that not every Trader C is exactly the same. ï Confirmation bias ï Appeal to complexity (convoluted setup) ï Post hoc ergo propter hoc (assuming that just because one technique occurred before another, that it was the cause of that event. For example, technique 2 happening before technique 4 does not mean that technique 2 caused technique 4.) ï False dilemma (possibly if there is favouritism between technique 3 over 4, and not considering co-existence) ï Inconsistency fallacy ï Cherry picking ï Fallacy of composition ï Appeal to complexity (if the system is complex in an overt manner). Trader C could think, ìthe market is complicated so my system must be too.î ï Ad hoc reasoning (if Trader C constantly adjusts rules or explanations to justify results, often seen in traders who journal their trades too) ï Subjectivist fallacy (if intuition is used, a setup could be justified because the trader believes in it, which works for me logically) ï Circular reasoning (if he uses many sources to line up his bias for execution, he may believe something is valid because he acted on it) ï Cognitive dissonance (for example, needing confluence but entering on discretionary setups) ï Narrative fallacy (for example, ìthis is a liquidity sweep before a big moveî) to rationalise a trade that lacks his actual setup alignment [5] Citations (Extra Reading Opportunities) https://en.wikipedia.org/wiki/Data dredging https://quant.fish/wiki/data-snooping-in-algorithmic-trading/ Forward Testing and Edge Decay Sentient Trading Society 1 Introduction In this document I aim to highlight how traders are putting themselves at a disadvantage by depending on forward testing for forming trading systems. Forward tests should not be relied upon in your own trading. Within the industry, yes, they forward test complex high frequency trading algorithms, but this does not mean that you are required to forward test your manual strategy for months. Creating a retail trading strategy where your market impact is negligible is very different from an institutional process. The primary reason for mostly skipping forward testing is simple. The edge lasts for months to a couple of years without any changes depending on the strategyís timeframe(s) etc. If you use time to forward test unnecessarily, you miss out on the edge (alpha will decay) by the time you have executed live. 2 Randomness and Misconception See the simple candlestick generator (Random Walk). Edges are not possible on random walk charts, yet many will convince themselves they can find an edge in a random walk. The market is not 100% a random walk. This is how you find edges in the first place. Forward testing is not a tool for discovery. The way most use it is for confirmation bias, for validation to execute. If you have not stress-tested historical data first, you are just watching your potential edge die in real time. This is called Alpha Decay. Forward testing is not discovery. It is using confirmation bias for validation to execute. If you have not stress-tested historical data first, you are simply observing decay. 3 The Reality of Forward Testing People talk about forward testing as if it is a serious standard. The truth is that if you know edges decay, forward testing is a luxury you cannot afford for long. 1 Figure 1: A Realistic Random Walk Chart The function of a forward test should only be to see if your system can actually be executed live, if it is feasible. When a trader tries to gauge this, it is still a waste of time to demo trade. It should be done on small live capital at the very least to experience adversity such as slippage, especially when executing low-timeframe systems. Most people demo trade half-baked ideas for weeks to months and wonder why they blow up live: ï Edge decay is real. Markets shift daily. ï Demo hides slippage. You will face real costs on small live capital. ï Anecdotes and outliers give false confidence. A simple candlestick generator example (Random Walk) can be seen here: https://codepen.io/yerb/pen/qB There will always be anecdotes and outliers. There will always be discretionary traders who make money for consecutive years, just like lucky people leaving casinos with millions over years. That is all fine, but there is no point aiming for something that cannot be tested and replicated. Luck is a factor, as intuition typically adds noise to results. 4 Analogy Imagine how unreasonable it would sound for someone to say: ìI have something proven that I can take advantage of now. It might not work in the future, but I want to wait a month and think about it first.î That is what the average forward tester sleepwalks into. Educators sell forward testing to give people a false sense of confidence and shift blame when the discretionary systems they teach perform poorly live after a ìgoodî forward test. If the edge you think you have found is real, it will not last long. Market dynamics shift constantly. Volume changes, participants randomly adjust, and volatility rotates. What worked last quarter might already be fading. Market anomalies and edges do not last forever. 5 Randomness and Misguided Validation Markets are very random. Here is another basic example of a random candlestick chart: Random Chart Generator Example. It is not possible to get an edge in a random walk, but people still convince themselves their system works on randomised charts, especially pattern-driven traders. If you spend weeks or months forward testing a setup that has not been properly tested in the past or does not have clear predefined rules, or suffers from discretionary undertones, you are just watching it decay in real time. You are not validating anything. You are observing something that might have worked while it slips away. 6 The Psychology of Forward Testing People perform forward testing because it feels safer. ìLet me paper trade it for a bit. Let us see if it works live.î But that is the trap. If you have not seen it hold up over dozens to hundreds of historical trades first, what are you even forward testing? The live market is not a lab. It is a minefield. Past ìworkingî strategies are imploding daily. The market is already changing. Edges are fleeting. It is only worth a forward test after you stress your system with historical data without overfitting. Even then, the goal is not to see if it works. It is to look for logical fallacies, cracks, breakdowns, and signs of alpha decay. If outcomes start to deviate significantly from historical data, whether positive or negative, it is time to back off and re-evaluate. For most systems, I skip forward testing entirely and start with live exposure as soon as the system has proven itself to be robust over honest, efficient data collection and processing. 7 Key Takeaway Forward testing is not where you figure out if a system works. That is what backtesting and real data are for. Forward testing is a tool. It can confirm that something can be executed, not whether it îworks.î That is the job of backtesting. 8 Summary Do not waste time forward testing half-baked ideas. That is time you will never get back. Build it properly, stress it hard, then bring it live or test it in a prop firm environment and stay sharp. The edge is already dying the second you discover it. Act accordingly. Ron Citation Julien Penasse -Understanding Alpha Decay ìAlpha decay refers to the reduction in abnormal expected returns (relative to an asset pricing model) in response to an anomaly becoming widely known among market participants.î Sentient Trading Societyô Stress-Testing & Out-of-Sample Data Sentient Trading Society Ron Out-of-sample (simplified) Think of it like launching a new product (strategy) in a business (portfolio). You can test it internally all you like, staff, friendly customers, and controlled conditions (in-sample), and it can look amazing. But when you scale, that environment can be flawed because it is ìour worldî and contains what we want to see. Out-of-sample is the equivalent of taking this new product idea and introducing it to an unbiased customer base such as a new region (a different set of data) to verify the ideaís robustness. Out-of-sample in the trading world One of the key mistakes traders make is forcing the idea that a strategy should work just as well in trading conditions that go directly against it. This is flawed reasoning (The Nirvana Fallacy). The reason out-of-sample is performed is to prove your strategy survives conditions outside of the initial testing. It is an anti-hindsight measure to improve the probability of robustness for real-time deployment. We perform out of sample tests in adverse conditions to see if the strategy fully collapses outside the main backtesting data, we call these íStress-testsí. If the strategy blows up, we do not rush to change the rules for survival, we discard the strategy and move on. ` A la Poubelle. (in the trash) Performance drag (permissible): It does not perform as well, but it survives. Example: 100R in-sample + 30R out-of-sample (same time window) is not ideal, but it is acceptable. If there is no return retention of at least 20%, the strategy is not robust. The higher the retention, the better the strategy. 1 Collapse (violation): Violation of risk constraints, for example, more than 100% of the maximum peak-to-trough drawdown statistics. Example: an acceptable drawdown threshold might be -32R out-of-sample when the max≠imum drawdown was -16R peak-to-trough in-sample (main backtesting data). If performance underperforms out-of-sample, it is normal. If it collapses, it is not. Specific, structured out-of-sample guidelines are provided at the end of the STS framework (an exclusive resource for mentorship members). Over-optimisation mitigation The only instance where optimisation after a blow-up is acceptable is if there is an objective hole in the strategyís logic itself (not the data). These tests are done repeatedly during the design process itself without data. Do not use this as a forced reaction after not seeing the ideal. Remember, most of the time, there is nothing there. This must be a tight operation to avoid overfitting. This is one optimal optimisation: you must improve the logic itself, not to appease data, but to improve the strategyís basis. If, after modification, the strategy still does not survive and perform well in-sample, it must be disposed of. Many traders throw out perfectly fine strategies or curve-fit for perfect stats and an equity curve that does not repeat in real time because it was adjusted to patterns that will not repeat in the future. This is why we prioritise deductive thinking over inductive practices to mitigate over-adjustment risk. If you squeeze the data hard enough, it will produce a story, but it will not produce the truth. Adverse market conditions: This helps us avoid deploying strategies that are more likely to blow up. For these, we prioritise small macro shifts with a mechanism behind the movement (because these happen regularly). Black swan events and crises get less priority for day trading strategies (because they are rarer). Context matters. We used to do manual research or look through articles, but LLMs can be used to speed up research, and example prompts and prompt-layering techniques are now provided within the material. We also test favourable out-of-sample conditions. This is to prove that the strategy can still post a positive return outside the data. If it falls to break-even or worse outside of the test after costs, that is a red flag, and we do not accept these models. Time ranges to test your strategies can be researched manually for objective classification, and AI/LLMs can be utilised to speed up your research without compromising your integrity if you actively review their claims. Prompt templates and personal layering tactics are provided later in the material. Regime Shifts and Trading Anomalies: Donít Ignore Them Sentient Trading Societyô (Public) Regime Shifts and Trading Anomalies: Donít Ignore Them Iíve seen many people dismiss major P&L swings to the upside with: ìdonít overthink itî or ìjust stick to the plan.î ìbut the same deviate when their strategies experience shock. If you are forced to retreat, you must do it in a logical, controlled way. We must not give in to fear or mania, or our P&L will feel it. By the end of this write-up, you will understand the process. I issue real time examples and adjustments. If your strategyís real-time performance deviates materially from your testing data, you are usually right to question it and investigate. I last experienced this first-hand twice since the reciprocal tariffs announcement at the time of writing this (Q4 2025 Latest Revision) I had experienced: An intense, extended string of losses on lower timeframes, exceeding anything seen in testing by over 30% (peak to trough). Figure 1: Abnormal peak-to-trough drawdown (never seen in testing) On the other side, I saw a windfall of profitable trades, followed by abnormally large profits due to amplified volatility and other factors. 1 Figure 2: Spike in gains never seen in testing (prior to the losses shown above) 1 lot = 1 unit long or short Dow Jones on this trading platform Long/Buy = . Short/Selling . This forced me to investigate and, eventually, adapt my system. This wasnít normal, it felt extreme in real time. If you trade systematically and suddenly see a major deviation from your historical data, it is not something to ignore. Itís a signal to: Analyse your system to determine whether these anomalies are rare or becoming frequent. If it is a true outlier, you may need additional testing. If it occurs oc≠casionally and within your expected distribution, you might continue as normal. For example, if you experience a peak-to-trough drawdown equivalent to 17 consecutive losing trades, but your worst run in two-plus years of lower-timeframe testing is 11, you may have a problem. Or you hit a 50R trade when the largest winner in testing is 20R and your average winner is 7R. Profit or loss, an anomaly should raise an eyebrow. What looks like luck can be an early warning that your strategy is becoming unstable. Adapt your approach to changing market conditions Make the adjustments required, including fundamental rule changes if necessary. This is not overthinking, it is using your data properly. Another example: Iíve used strategies with no fixed profit target, where I manually trailed stops on reversals. In backtests, I might see more than 10 losses in a row, followed by a single 50R winner (for example, 10 points risked and a 500-point drop from the high). One strategy averaged 7.31R per win, but only because of those monster outliers, so it had to be changed. Hereís the key: Log the outlier trades, but also test the system without them. If your strategy only works because of rare events, that is a red flag. Remove the big win(s) and see whether performance still holds up. The key question to ask Where did the effect come from, fundamentally? A sound checklist: 1. Pinpoint the specific news catalyst. 2. Review economic developments in countries that indirectly influence the assetís pricing. For example, China and Saudi Arabia (via OPEC+) can indirectly influence US crude oil pricing (CL futures), which can alter regimes for months. 3. Review macro shifts, such as central banking or interest rate changes. 4. Review industry-specific news, such as demand or supply strains in commodities (metals, energy, soft commodities). 5. Review geopolitics and escalation risk. Your goal is to judge whether the change in return volatility was genuinely random, or whether it is likely to persist. Response options: 1. Design and run additional uncorrelated systems on the same market or asset class to better fit the new conditions. 2. Reduce exposure and consider withdrawals if warranted. 3. If volatility is abnormally favourable (for example, a never-before-seen +30R in two weeks when the strategy averages +12R per month), withdraw gains and continue only if the strategyís core behaviour remains intact. 4. Ask whether the regime you were exploiting is still present, or whether it has ended. 5. Remember that sometimes there is nothing there. Avoid confirmation bias. If there is no evidence, do not force a narrative. Noise happens. Do not rush to stop trading a valid strategy without a clear reason. A Regime Shift and Reaction (Example) Scenario: A Sentient Trader who has transitioned from metals to energy. For the time being, he trades CL futures exclusively and deploys two strategies at different times: ï An intraday mean reversion strategy (1+ years of data, plus out-of-sample testing) ï A reversal strategy for post-news-release days (2+ years of data, plus out-of-sample testing) He has been trading these accounts for two months and has processed a withdrawal. He notices a modest drawdown in his mean reversion strategy and, within the past week, has seen intense liquidity shocks from news, with large P&L increases from his reversal strategy. He sees CL trending, but still ranging after news-induced shocks. He checks the economic calendar and other sources and finds that Saudi Arabia has made an unexpected announce≠ment that tactically tightens supply by reducing exports to a list of nations. Fundamentally, this does not guarantee an uptrend, but it can increase autocorrelation, which can produce more trending behaviour and sharper reversals. In this situation, the trader can expect mean reversion to experience more friction, but it also opens another opportunity: trend following. Because of this regime, he can be more open to deploying one of his trend-following systems after testing and reviewing the situation (due diligence), to ride the drift with negligible market impact, benefiting in ways that would likely have experienced more friction over the last couple of months of a mean reversion, range-dominated regime (in this example). In this example, the implied regime shift is framed as conditional on confirmation in follow-through. A headline is never enough. The change must show up intraday, be genuinely unexpected, and have an impact that is likely to persist, not something that fades next month. If the market moves on quickly post-shock, adjustments are usually not worth it. Media: Separate Facts from Narrative Mainstream media often misleads with headlines, ignore the titles and focus on the numbers and mechanism(s) at play. Discern narrative from facts when receiving information. Reject the narrative first, look into it and create your own. When you see an article do not take it at face value always look at it with the perception íthey want me to believe that this is happening to X market or economyí. Analyse potential incentives as to why someone may want you to believe something is good or bad for the economy or a specific instrument. Do not accept the headline, pull back the curtain, look at what the numbers are revealing and the possible positive and negative consequences, and why. If you are short on time, you can use AI/LLMs to research for you. Do not assume the AI is objective, you must read the sources and double-check them to form your own opinion surrounding the situation, and if you choose to, you can reposition yourself to benefit. News often reveals the WHAT aspect they want you to believe. You have to look for genuine causation, the WHY, to understand WHAT is really going on and the potential effects. Only then can it be useful information. The front-facing news you receive is often funded or influenced by people who may have direct conflicts of interest. This is not a claim of coordination; it is grounded in observable facts and verifiable mecha≠nisms that were at play. This section was made as a reminder that incentives shape framing when information is delivered to the public. If you must, treat narrative as a hypothesis generator, then validate or reject it using observable mechanics and data. It may align or it may not, and the only way to know is to verify what is actually happening. Defining îPsyopî Correctly (Often misused) A psychological operation (often shortened to psyop) is a planned effort to influence how a target audience thinks, feels, behaves, and reacts by communicating carefully selected infor≠mation, cues, or messages, often delivered in a specific way or at a specific time to achieve a desired effect. In military intelligence contexts (origin), it is typically used to shape attitudes and deci≠sions in ways that support an objective, without relying on direct force. In financial markets, a psyop is used to manipulate peopleís attitudes and decisions in ways that typically support a specific market action (e.g., impulse buying, panic selling), inaction (no participation at all), or another trading behaviour pattern that is beneficial to the news providerís owners, funders, or associates. Example Of A Popular Psychological Operation Against Retail 1. April 2020: The Negative US Crude Oil Psychological Operation Headline WHAT (What they wanted us to believe) ìOil demand collapsed, oil was worth less than nothing. Hey retail do you see it?!î Zero substance, distractionî Mechanism WHY (What was actually going on -> Public information): As storage capacity tightened, longs position who had no interest in taking delivery had to exit (mechanism), they were paying others to assume delivery risk and storage costs. The negative price was a contract-and-delivery constraint (mechanism), not ìall oil is worthlessî situation. There was not enough storage space to accept delivery for the oil barrels and that was reflected into the price. If you cannot accept the delivery you cannot accept the exposure so you must sell. The Market Microstructure Side The price of oil futures went into negative territory because forced sellers had to par≠ticipate in a market with almost no viable buyers, and the few buyers who could take the other side of their trade required payment to assume storage, delivery, and margin risk. Many market makers and others pulled their bids to avoid holding inventory they could not afford to take on. When someone sells using a market order, if there is no bid close by, the price will keep going down until a bid (buy limit) is met to fill the other side of the trade. Proving The Psychological Operation Classification Isnít Just Talk . Figure 3: The United States Oil Fund was underperforming oil, which is the instrument it is supposed to be following. After the negative pricing event, there was major displacement. After reading you will understand why, what and how. (a) The Execution Side: We were unable to trade oil on a lot of platforms, and we were not shown negative prices on many feeds. I also remember my old broker at the time not allowing oil CFD positions (this was common). The same applied on platforms offering the underlying instruments, such as E*TRADE. Many CFD platforms did not quote negative prices, trading platforms were not prepared for negative quotes. The whole situation was bizarre. (b) The Oil ETF Side: Normal people/retail traders treated it as a ìbuy the dipî moment and piled into easy ways to get exposure like the United States Oil Fund (USO), rather than trading the expiring futures contract itself. One widely cited source shown that Robinhood accounts (retail) holding USO jumped up around to 220,905 by 28 April 2020, when it was only 8,000 two months earlier according to Robinhood trackers (Stopped in august 2020) but I found the evidence here. How retail participants got flattened by the psychological operation ï Cash/No Leverage: Many got stopped out or sold at a loss due to the adverse volatility whilst media implied easy money. Others were impatient and sold early as it didnít recover as fast as oil prices did for USOís price to match the price of the high of April 20th 2020 investors had to wait until mid-December 2020, oil prices had normalised and retail traders were still in drawdown as seen in Figure 3. By the time they have realised what is going on it is too late. ï Leveraged traders: People using margin/leverage would have got margin called closing out in steep losses or would have gotten liquidated (forced out of the position) as USO lost nearly almost half of its value from the start of the event until price found the bottom and reversed. ï Options traders: Regular options traders paid for overpriced options because the implied volatil≠ity skyrocketed reducing the return or elevating costs making it much harder to get an edge. (c) Oil Futures Liquidations (key) The subtle narrative retail was sold was: ëif you buy, it is easy moneyí, so a lot bought with leveraged positions. Reality: If you buy at -20, for example 1 CL contract with a $1000 day trading margin and a 5000 USD account, as soon as price goes to -24 your position would get a margin call, as CL provides $1000 in gains or losses per $1 of movement. Some even thought that as soon as you buy in, you get a positive p&l. Some genuinely thought they would get a free lunch. Retail bought up the barrels on margin and got forced to hand them over, they became liquidity to larger participants who could deliver repeatedly. These buyers were turned into bursts of aggressive sell order flow which bids (buy lim≠its) who could take on inventory gladly accepted as they had the logistics to take on delivery. There are other ways institutions did benefit, but they are not rel≠evant here. I want to keep this document tightly focused on what will help you reason more effectively when similar events occur in the future, so it stays prac≠tical and workable for you as a trader, given as we do not have the same resources. Conclusion: Considering that everything I listed is intuitive and obvious to institutional par≠ticipants (especially) that retail would buy. I believe it was fair to label this news cycle a psychological operation. Summary: i. Most of retail chases after the move and later becomes liquidity instead of po≠sitioning themselves on the underlying market periodically to ride the move≠ment back up in a controlled way (with risk limits and targets). ii. Institutions benefit from future prices of oil being a lot more expensive than current value (contango) whilst ETFs structured like USO have to pay that difference when re-rolling on to the next month. iii. Retail acts on the purported story while professionals trade the constraint (liquidity, margin, settlement, and positive positioning) that is the Psyop. (d) How I would have benefited with the same knowledge I have today (One of two): i. 1. I would have developed an aggressive long only day trading trend following strategy (with in and out of sample data) to benefit from the positive drift which skews price to go up post market shock to shave P&L from the market. I would have performed this on CL or Oil CFDs. ii. 2. I would have developed a long only swing trading long only strategy for oil. iii. Note: Points 1 and 2 are not hindsight/post hoc rationalisations, as the liq≠uidity shock was a temporary overextension with an underlying fundamental reason. Once conditions started to stabilise and price quickly regained pos≠itive footing, the larger imbalance still remained and had to be rebalanced. The movement high (positive drift) would be required for the rebalancing, and going long only at efficient prices while taking profits would allow me, within my current style(s), to benefit and compound my trades. iv. Back in 2020 I wasnít consistently profitable. I responded by building a long-only, indicator-driven strategy, but it was over-fitted and did not hold up in real conditions because my process was not mature enough at the time to execute and validate it strategies properly. I have still got some of the original notes saved in my archive. I did take losses during that period (I believe it was on demo), which makes sense in hindsight, I was still learning the craft then. Common Red Flags 1. Selection and omission: The news highlights one or few true fact that are actually true but they leave out the key context that changes the meaning of the event. You see this when you compare one news feed to another or when researching you notice íîthey did not tell me about thisî. 2. Narrative anchoring: They love to give people explanations so they feel like they are in control and stop searching for the real drivers. 3. Authority signalling with appeal to authority (Posturing): ìsourcesî, ìexperts sayî, ìinsidersî, without showing the numbers. (Easiest to spot and the most amusing) 4. CTAs: Call to action: Subtle or direct, ìthis is the opportunityî, ìdo not miss itî or the subtle îexpert says this should happenî or îexpert has an optimistic viewî and vice versa. 5. They often aim to shape the narrative through authoritative messaging, then use sub≠tle guidance to encourage participants to follow along with low questioning (they are experts right) this typically viewers who act into an unfavourable position (free liquid≠ity), or into a lack of pressure from the other side: ìthe stock will be fineî, when the stock is set to collapse next week (not naming names here). Note: Do not assume manipulation by default. But if you cannot define the mechanism that would make it falsifiable, do nothing. If it is a subjective situation there is nothing objective to model. When you see an article paired with a noticeable regime shift that you might be able to benefit from, pause and ask yourself this first. 1. Mechanism: What must be true for this story to be right? 2. Statistics: What numbers would falsify it? (Current price movements, institutional positioning, hedging behaviour, other things specific to the event) 3. Consequence of action or inaction (if suggested): Who benefits if I act on this infor≠mation today or when they imply that I should? 4. What would change my mind?: If price acceptance, volatility, and order flow return to what it was like before the large market levelling news, treat the event as temporary and continue trading as if nothing happened. Return to baseline post shock often shows it was all talk no bite. No need to feel rattled. We seen this a lot with reciprocal tariffs people even called in îTrump always chickens outî TACO. Sure, Market shock it could persist so we bet on it but when the market has rebalanced and the underlying drivers are no longer in the picture if performance drags exposure gets reduced. 5. If this is real, will it persist or continue in real time, and if so, why? If you have an objective reason to believe it will persist (or unwind) over months to years, you can build strategies to benefit from the movement, whether itís skewed higher or skewed lower, indirectly through the strategies you design. Note: Read at least one primary source (filings, regulator notices or releases, exchange rule changes, actual datasets regarding the news) instead of relying solely on commentary. This is super easy to find even with search engines, but AI makes it that much faster to access. Note 2: The reason I needed to reposition was because I was operating on lower time-frames (5m). Personal: adjustments I made during the reciprocal tariffs period: I ran a different model (with a custom, price-structure-based mechanical trailing stop rule set) that I had used before, but applied it to different trading hours to take advantage of the volatility. It produced the largest single-trade profit Iíve ever made (Q4 2025 for context). Applying the reasoning present in this document (real≠time scenario) The reciprocal tariffs period also marked the largest peak-to-trough drawdown. I have ex≠perienced to date, followed by my best recovery (financially, not percentage wise). My tail risk had went through the roof, then settled back down after this I changed my secondary strategy and reduced my capital with withdrawals to mitigate risk of ruin. (Q4 2025 for context). It was my largest account size at the time, which explains the severity of the drawdown when it happened. I took over a dozen losing trades in a row, the fastest accumulation of losses I had ever seen for that strategy, both in-sample and out-of-sample. I recognised the regime change and the resulting edge decay. More importantly, the draw-down acceleration was underpinned by a fundamental driver with objective effects on volatil≠ity that I expected to persist. I replaced the strategy and reduced exposure on the previous one to zero. For usual, less extreme adjustments, I consider opening additional accounts, for example an oil trading account for crude oil that I opened in 2025. What most people saw: the media-led smokescreen. 1. The Narrative: Strength, reciprocal, and ìit trickles down to favour the USA economyî, or ìthis is going to make everything expensiveî, are purely speculative takes with nothing useful or objective. Biased news waffling is common, we know the game. Effect Of Psyop: Investors lost their minds and sold off, buying into the fear. People deviated from their plans, told themselves they would buy the dip, but failed to, resulting in a worse cost basis. Retail traders who might have understood what was going on failed to do so because they were too distracted by the media or by personal cost-of-living worries. 2. What I Saw: The Mechanism A potential excuse to increase prices is to use tariffs as justification, especially in tech. For example, if a phone costs $100 to make and is sold for $500, tariffs come in, and now the phone costs $125 to make, and the seller increases the price to $550. The $25 discrepancy is additional gross profit for the company, especially for essential goods that consumers need. Although this would only hold if the firm has enough pricing power, the incentive is still there, and if multiple companies share the same incentives around pricing, and the media pushes tariffs as the narrative for why everything is overpriced, consumers may buy it and investors may tolerate it. I was thinking weaker demand will force the margin back down, but by then they would have either 1. made their money already or 2. refused to budge, gauge reactions and if there is no bite continue the pricing behaviour. I thought to myself, ìpeople could really take the piss hereî. Even if theyíre skimming a couple of extra dollars from goods at a time, if demand persists, this could boost revenues and potentially gross income, or even margins, if demand holds. If prices increase in foundational sectors where they donít need to, it could spill over into addi≠tional sectors. I put myself in the shoes of the companies and sat and thought through price increases, their purpose, and their potential benefits. I bought more $VUSA (UK S&P 500 ETFs) in response, and made changes to my live trading too, before switching up to reduce downside risk to capital. Objective reasons tariffs could change (and sustain) a volatility regime is that it is a policy shock that could potentially alters future dealings and trade globally it was not about poor politics or strength like the media tried distracting us with. Another thing they hit us with is they hit growth and inflation expectations at the same time, which has added to uncertainty, which was reflected in the VIX quite quickly, making this relatively obvious. Even when a market is not directly tariffed (for example, crude oil), it can still become erratic because of related instruments and hedging against retaliation, etc. What I did and why 1. Wider Stops -> higher target and stop minimum values I was forced to match the newly realised volatility so I did not get ruined by market noise every time I had to make an entry. It is also worth knowing that the spreads were insanely high, often ticking over 10 points, which my strategy was not made for. Even on a CFD broker I had got a quote of 1.3 points when the bid-ask spread on the underlying instrument was over 10 points, which sounds good but the fills were terrible, the book was thin and limit orders were not getting filled as consistently, this made me need to build positions through techniques such as order splitting to get the correct size in which was just as costly. The Key: The costs had exceeded 30% the strategy was no longer tradable. 2. Price-Structure Trailing Stops I employed one of our methods to adapt to expansion and displacement (it trails behind swings and acceptance), rather than assuming mean reversion is still dominant (important because this was on the Dow Jones) 3. No Hard Target I knew that having no fixed target would let me harvest the movements that were likely to happen as the market attempts to rebalance higher and lower post-shock as a result of forced repositioning and liquidity (larger intraday bid-ask spreads are a great tell). I made sure edge became îI will stay in the move until the structure breaks and look for potential re-entry,î not ìI will take an arbitrary RRR at X price level.î Adapting to regime shifts requires deploying a different mechanical model, not applying intuitive discretion to an existing one. Model integrity comes first. The costs would exceed the maximum (30%) by more than double because the minimum stop loss size was 10 ticks. If the strategy was forced to pay 8 to 16 tick bid-ask spreads, the edge would collapse. This was objective reality. I had to stop trading, otherwise I would stand to lose more than twice the amount per stop loss, and over 50% of gains realised (relative to the risks taken) would be absorbed by spreads alone, not including slippage. All discretion is baked into the rules ahead of execution. 1. A specific trading window (hours you trade) is discretion. 2. The maximum costs and drawdown tolerated is discretion. 3. Only trading one instrument is discretionary application. There was no real-time intuitive factor and never should be as a mechanical trader. Remember, intuition is the enemy of consistency, not discretion. Closing Statements Media: Be sceptical of media and review what you see for bias and intent behind mainstream in≠formation sharing. Do not lean on conspiracy, narrative or íexpert predictionsí look at the mechanism and numbers to make your own opinion. Do not be blindly contrarian; although it feels good, it is not correct to label everything a psychological operation without basis. Being contrarian is not an edge. Evidence is the edge. Even people who intentionally go against the narrative often have an agenda or funding which can be a multi-layered psychological operation. We do not trade solely based on fundamentals If an opportunity presents itself we will make strategies to take advantage, it is one of many pathways we use to create profitable strategies that perform in real time. Keep strategies simple for now and when you are ready, you can attempt to branch off to more complex strategies which take macro-economic shifts into account. STS Strategy Engineering: Always optimise. Never get complacent. Profitable strategies do not last forever. When trading anomalies show up, stay sharp and do not be afraid to investigate, even if it comes with unexpected profits. The market does not care if youíve made 10k in a week or 100k in a week. If you get complacent, you are booking a table at the roundtable of financial humility. Never let a hot run lower your standards. Protect the process, keep building. Thanks for reading, Ron Six Rules for Profitable, Sustainable Trading Sentient Trading Societyô Come up with an idea. Logic first avoid charts for ideas. 1 Otherwise, youíll run on confirmation bias & overfit strategies. Come up with a trading idea based on logic. Donít start with charts youíll just end up fitting patterns to what you want to see and building strategies that do not hold up. If you ever want to seek wisdom from the chart you must find supporting evidence to back it up, most patterns observed are random. It is best to learn how the market works to come up with sound ideas, this material will help guide you. Create rules for consistent entries and exits; Underpinned with a plan to behave just like the backtest. If you canít behave 1:1 drop it. Define ahead of time exactly how youíll get in and out of trades. You need to be able to trade the strategy exactly as you backtested it. If you canít stick to it 1:1 in real time, itís not usable. Backtest your system (do not tweak rules) do not curve fit yourself system; if it doesnít work trash it. Test your rules as they are. Donít keep adjusting things just to make the backtest look good. If the system doesnít work out of the gate, move on. Process your backtesting data In a spreadsheet to get important values such as peak to trough drawdown (R) and avg monthly return. Run the numbers. Take the backtest data and analyse your drawdowns, losing streaks, average monthly return, etc. Use a spreadsheet or an alternative backtest platform (it is upto you). You need to know what to expect before you go live. Execute as soon as your system data is processed and ready; Trade it while it works Short term trading edges will fade with time naturally. Once the system checks out, start trading. Edges donít last forever, especially short-term ones. Do not forward test for too long. Do not share your strategies. Keep your edges to yourself. 1 Public strategy sharing contributes to the potential for prop firm expulsions and many other negatives. You have your specific profitable trading strategy; keep quiet. Your edge is yours; protect it. 2 Bonus: How to keep your profits and survive Take Advantage Of Variance To amplify your chances of success make multiple well-tested efficient strategies. It is best to have multiple logically tested systems that are uncorrelated to trade on separate accounts or prop firm evaluations at the start. You must spread but concentrate risk and take advantage of variance; do not take on excess vulnerability at the start. If you limit yourself to one system at a time at the beginning your chances of success can be far lower. Do not fear missing out, spread your risk when the opportunity presents itself. Isolate your trading capital Instead of depositing $10,000 ex. Trade high risk on $2,500. Do not remain overexposed your edge can stop working at any time. Your working capital should always be small relative to total risk. Abuse compounding. Withdraw. You must withdraw at equity highs when your strategy is performing well especially on high-risk models. This cannot be done randomly, we have fixed process to calculate this ahead of the time. Every values needs be precise and known in advance before deployment for controlled outcomes. The exception is withdrawal to supplement goals, again this is planned in advance. You cannot îeyeî or predict when a strategyís performance will drag over low samples; attempting to will often result in random periods of negative drag on your returns.3 Do not get complacent Always test new systems and ideas constantly even if at equity highs; your strategy break≠down is always an unpredictable surprise. Have a replacement in mind regardless of perfor≠mance. When your strategy deviates from its backtesting behaviour ex. Large profits instead of celebrating reduce exposure/withdraw. When your drawdown exceeds maximum peak to trough drawdown on testing drop the strategy and withdraw everything. Buy assets. Skip the cash hoarding. Regardless of what happens trading-wise do not sell what you accumulate. Buy assets. Real ones like ETFs, stocks, property, businesses. Donít sit on excess piles of cash unless you need it. Once youíve built up investments, donít sell them just because trading goes sideways. Those assets are your foundation. Leave them alone. Thatís how you grow and stay in the game. Do not skip the post-reading document; it is just as important as this volume. Context: [1] If you canít come up ideas study basic market microstructure theory or order flow mechan≠ics (why price moves) Consider these reads: https://www.investopedia.com/terms/o/order≠ book.asp https://www.investopedia.com/terms/t/time-and-sales.asp. https://www.investopedia.com/terms/p/pricediscovery.asp https://www.scribd.com/document/349585448/Market-Microstructure-Theory-pdf Learn what wicks and closes represent on a chart and create ideas based on it. [2] All prop firms do not allow people to copy eachotherís trades (copy trading) + If your short-term system becomes widespread, market crowding can interfere with strategy execu≠tion performance or the likelihood of your trade being filled. Itís not worth it. The concern is solely the potential negative impact on system performance. [3] Most traders do not withdraw profit even if theyíre at equity highs. Be the one who Withdraws profit. Key 2018 report in Europe shows î74-89% of retail accounts typically lose money on their investments, with average losses per client ranging from Ä1,600 to Ä29,000.î European Securities and Markets Authority Stats Julien Penasse -Understanding Alpha Decay On the Effect of Alpha Decay and Transaction Costs on the Multi-period Optimal Trading Strategy High frequency market making: The role of speed -Yacine A®it-Sahalia, Mehmet Sa.glam IG Index (Example of a regulated CFD broker) CFD Customer agreement key parts: 12.8b 21.1 and so on https://www.ig.com/uk/customer-agreement Turtle Trading Edge & Alpha Decay Note: Turtle strategyís returns got diluted after media exposure or retail adoption & wors≠ened after structural changes because of electronic trading, etc. Note: Turtle strategyís returns got diluted after media exposure or retail adoption & wors≠ened after structural changes because of electronic trading etc. Does Academic Research Destroy Stock Return Predictability? -Journal of Finance, R. David McLean Published in 2016 Key takeaway: ìPortfolio returns are 26% lower out-of-sample and 58% lower post-publication. The out-of≠sample decline is an upper bound estimate of data mining effects. We estimate a 32% (58% -26%) lower return from publication-informed trading.î

Strategy Engineering Key Insights and Post-reading FAQs

Essential Clarifications & Insights Following Reading Sentient Trading Society Strategy Engineering Volume 2 1 Q1. Does it make sense to look at market structure to filter setups, or is it enoughî if I use a moving average to filter trades? You can and should use additional methods. There are many. There is a lot that you can use, for example, for a trend trading system. For example, a breakout of a recent high in a trending market. Indicators like the moving average can be used, but there are better filters that we use that complement different entry techniques. Remember, you can research to get inspiration and use the knowledge gathered to filter out baseless ideas after you finish the rest of the materials. As my strategy design paper states, dismiss educator narratives and use critical thinking. The logic and insights provided in the free documents can be enough to create a strong, logical information filter when creating unique trading ideas (the goal) or looking for trading ideas for inspiration in general. The common mistake is to rely only on a few types of entries, filters, or exits. Traders can get emotionally attached to their strategies. If it does not produce what you need, it is time to change. 2 Q2. How do I create systems just like Ali and Ron? The way you design your systems can be unique to you. Sentient Trading Society is here to stimulate how you think so you can be self-reliant. STS is all about helping you create a procedure that works for you and around lifeís constraints, without oppressing your expres≠sion. The process is applying real logic to filter out flawed ideas. Complete overrides are not always necessary; however, the filter is required. We are not here to force your hand. We are here to guide it. 1 Traders who apply these principles properly tend to end up with different, often uncorrelated systems, even though the underlying logic is shared. That is a good thing. Keep it simple. As an efficient trader, you need to trade an initial imbalance (so many options) to imbalance (for example, swing high or low, price gaps, price void variations), or imbalance to price extreme (imbalance, for example, swing high target). As a trader using STSí approach, your aim is to create rules that can be tested against the behaviour(s) you are trying to capture instead of having to rely on copying another traderís exact method. I know it sounds clich¥e, but less is more when it comes to trading your system. You should not have many rules or parameters. It should be basic. It is your data collection and data processing that need to be ultra refined. The more things required to line up, the more susceptible your system is to overfitting, and the more likely it was overfitted in the first place. As an STS trader, when your strategy makes gains, you are trained to withdraw. If it has sudden spikes in profits like never seen before, you withdraw even more. When your system loses efficiency, you must create another. If you are running multiple systems or instruments, you must monitor drawdowns and performance. This is done with the intricate systems that we have created to maximise returns by optimising withdrawals, system replacement, capital partitioning, risk, risk isolation, and so on (discussed in more detail in advanced material.) Do not focus on trying to replicate me and Ali one to one in strategy design. Focus on building a strategy design procedure for you that is realistic and fits within your real life constraints. I see a decent number of people attempting to mimic us, which is natural, but this is a bad idea for most. It will be too rigid and difficult, especially if you come from a discretionary trading background. Do not do this, or you will likely be stuck. Take it slow. Traders, remember this is all about you personally. Everyone is different, and we are only human. This is why strategies backed by intuition are a bad choice for people to try to replicate. When we work with mentees, we tailor their strategy design process for them and give two to one support and insights. 3 Q3. What do you think about discretionary trading There is the common idea that floats around endlessly that discretionary trading means you are being flexible and smart, while mechanical trading is some rigid, one-size-fits-all system that ignores the market context. That is an oversimplification. Mechanical systems can be flexible. Think of them like flowcharts or decision trees. They can include filters for volatility, time of day, higher timeframe context, session structure, and basically anything you want to build, or as many nodes as you want if we are imagining a flowchart or decision tree. You can even bake ìdiscretionî into a mechanical system if you put in the work. One hundred percent serious. I have a document discussing this thoroughly. Discretion isnít the enemy, trading with intuition is. 4 Q4. What do you think about multiple instruments on one account? 4.1. Trading multiple symbols: when it is acceptable. If the trader divides their risk based on the drawdown shown in your backtest data. You do not let the order of the trades affect your overall trade execution. Do not create rules that add noise to your real time trading performance. Example: Common trader mistake. A trader limits themselves to three open trades at once while trad≠ing over three instruments. In hindsight, this feels safe, especially in prop firm environments with max drawdown enforcement, but this will cause you to miss trades that would have been placed in the environment you backtested. This creates random trading behaviour that was not accounted for in your backtest (sometimes you will randomly skip trades). Example: A trader is long YM, long ES, short NQ, and the trader has a valid short on CL and E6 but does not take them all because of their max three trades rule. They should have either run lower risk or separate accounts. Random execution patterns equal noise. It is better to trade one or a few instruments you are good at with higher risk. My strategy design model lays this out perfectly. If you have two or more markets that deliver similar results, or you can weigh them, you can diversify. 4.2. How STS-style risk and withdrawal management is superior Example: If a trader originally wanted to risk 3% per trade on a single system, but they have two systems with similar maximum drawdown and return in R, they could risk close to 1.5% per trade. It is still superior to split the two accounts. Three percent risk, for example, with $10,000 as a lump sum is inferior to $5,000 times two. Market 1 for account 1. Market 2 for account 2. $5k USD each. The $10k lump sum example could have strategy 1 working while strategy 2 produces a negative return, neutralising the gains after costs or even resulting in a loss. Account 1 could have a large return, and account 2 is in drawdown, resulting in a net profit. Example: 1:2 RRR system with a 50% win rate (generic profitable model). 10k lump sum ï Market 1: 65.79% gain when isolated 1.5% risk per trade. (35 TPs, 35 SLs). ï Market 2: 26.09% loss (reaches ruin), loss (20 SLs). ï Net result: Approximately 22.53% gain, $2,253 profit at 1.5% risk per trade. ï Assumptions: All trades are compounded in the simulation. In real time (opti≠mistic), they would not be compounded to this extent (overlapping positions), which is another downside. ï Definition: Ruin corresponds to a -20R return per strategy in this example. Ruin is the maximum peak to trough drawdown; a blow-up. 5k + 5k split (Separated) ï Account 1, 5k Partition: Strategy 1: = 164.7% gain with 3% risk, $8,233.84 gain this is an 82.33% gain on 10k (overall) capital) with 35 TPs and 35 SLs. ï Account 2, 5k Partition: Strategy 2: 45.6% loss on 5k (ruin) or 22.81% drawdown on overall capital, $2,281 loss (20 SLs). ï Net result: $5,952 gain, 59.52% (10k). ï Difference versus lump sum: $3699 more profit as compared to the 10k lump sum account... A 164% increase in profits from strategy risk isolation and STS style capital partitioning (22.53% net versus 59.52%). 3% of risk on 5k is the same as 1.5% of risk on 10k, 150; the starting risk is equal, but the whole point is that it stops being equal as compounding kicks in. We benefit from: 1. More extreme compounding steps in our favour with higher percentages 2. Zero multi-strategy interference with each otherís performance mechanically 3. You let your strategies perform solo, concentrating its edge. Drawbacks 1. More leverage is used so low margin requirements are needed. 2. Only applicable to personal accounts. 3. Risk caps/limits exist on prop firms and over environments, making this unfea≠sible in those environments. Here are some simplified examples: ï 1. Let us assume the risk percentage is equal for these trading setups Imagine 10,000 USD starting capital trading two setups; one setup makes 100% after costs, and one setup loses 50% after costs. The trader makes zero because of this voluntary overlap. ï 2. Let us assume the average risk is always 2% on both setups Basic parameters: The starting balance is $10,000 Both have an average gain of 1:2 RRR after costs Setup 1 has a 50% winrate over 50 trades, 25 profitable outcomes +2R each, 25 losing outcomes resulting in -1R each (Very good) Setup 2 degrades in real time resulting in a losing 16.66% winrate over 30 trades, 5 profitable outcomes, 25 losing outcomes (a negative result). If these trading îsetupsî were executed separately here are the values. Setup 1: 10000 ◊ 1.0425 ◊ 0.9825 = 16087.38 Ending Balance This provides a $6087.38 Gain Setup 2: 10000 ◊ 1.045 ◊ 0.9825 = 7342.07 Ending Balance This results in a $2657.93 Loss Setup 1ís gains + Setup 2ís losses result in a $3429.45 net gain If both of the setups were executed on the same account (like most traders do), here is the consequence. 10000 ◊ 1.0430 ◊ 0.9850 = $11811.47 Ending Balance If they had been executed separately, $3429.45 would have been earned by the trader. However, because they were merged, the net gain fell to $1811.47, even though the positions executed were the same. Smaller accounts can also use it, for example, two 5,000 USD accounts with higher risk percentages instead of using lump-sum deposits. $3429.45 -$1811.47 is a loss of $1617.98 from bilateral execution. 50% of profits were wiped out through misapplication, a common invisible cost of execution inefficiency. 4.3. Strategy and risk isolation The main takeaway is to let strategies compound on their own path, the higher percentage a trader uses as risk, the more pronounced the advantage becomes. This takes advantage of path dependence, and other elements present in industry-level statis≠tics. A key problem with discretionary and intuitive strategies is that they try to combine multiple systems together subconsciously, which interferes with profitable outcomes system≠ically. To avoid strategy performances interfering with each other, we suggest you trade your one or few backtested setups on a single instrument per account. Your gains do not get weighed down by another setupís drawdown if it happens. This also makes individual setup per≠formance linear and easier to track. If you insist on trading multiple instruments, it is doable but will require more work and testing. Ignoring this can be the difference between a profitable year and a losing one. The consequences are that serious. Make sure all valid trades are executed and do not interfere with one another. You need enough margin, leverage, and risk available ahead of time. The aim of these protocols we founded was to simplify winning systems to compound at a higher percentage of risk per trade, while the losers shrink independently. This shift, including risk adjustments, prevents effective setups from being dragged down by potential poor performance from others. Sentient Trading Societyô 5 Q5. If edge decay usually shows up in the derivative (not the core behaviour), how do you tell when a derivative has truly decayed rather than just reflecting normal variance or a temporary regime shift? You donít ìknowî decay from feelings; you decide via process. We do this by comparing live results to the strategyís expected distribution from the backtestís drawdown, win rate, and avg RRR (if variable). You only intervene if it is outside what basic variance should allow. If it is suspicious, for example, a 20%+ deviation within a 6-month sample, we re-analyse, and if itís suspi≠ciously good or bad, we explore whether it is reducing our capital at risk. If there is no objective reason we can point to for why the strategy is overperforming or underperforming, we withdraw more (reduce). If there is something with strong support that should persist, we continue as normal (no increase to exposure). We always check the obvious culprits first: costs/execution and venue (CFD or exchange de≠pending on the market). spread, slippage, swaps and prop constraints because the behaviour can still be there, but the derivative stops extracting net positive expectancy. We then move on to checking the regime: if it still works in the conditions it was built for but fails elsewhere, that is usually a mismatch or a missing filter, rather than the be≠haviour supporting the edge fading. A lot of ìdecayî is failure to redesign the derivative, but you cannot, as constant tweaking becomes overfitting if it is not driven by a new, clear mechanism and robust re-testing. It is always best to prioritise prevention rather than trying to fix something that is showing instability. Every trading strategy is different, which is why we push for robust reasoning paired with genuine market understanding and data acceptance. 5.1 Never try to ìfixî a broken system. If the system is done, it is done. If a suitable regime reappears and it starts working again, consider re-running. In documents, I refer to this as putting it on ice. Performance drag is an active shortfall in realised profits relative to the strategyís ex≠pected performance (from backtesting), caused by constraints or mismatches that reduce what you actually gain in live trading. Key mistakes: Discretionary traders make the key mistake of being emotionally attached to a given system. A lot of mechanical traders, especially algos, make the key mistake of sitting through performance drag (return drag) or system hopping with curve-fit ìoptimisationsî. We do not have years to waste. 6 Creating Significant Systems Q&A 6.1 Q6. Should I Be Looking for High Win Rate, High RRR Systems? Answer: It is not that linear. Our best systems ever had low winrates and high risk-to-reward ratios (RRR). Winrate and RRR have an inverse correlation. The relationship for the break-even winrate is. 1 win rate = , (1) 1 + reward where the reward is just the latter number in RRR, such as the 2 in a 1:2 RRR. Most importantly, on their own, they mean absolutely nothing. You do not know absolutely anything about a strategy if you only know one of these. Both must be known. Traders generally like to boast about their high win rate systems or high RRRs, and this is nonsense. Expectancy (expected value) matters a lot more, and so do other metrics which account for risk. You should note that something special happens when you increase the RRR beyond 1:1, the expectancy can exceed 1R per trade (the average return per trade), where R just means one unit of risk. To be clear, R means that a win of size 2R with a 2% risk per trade gives 2 ◊ 2% = 4% profit. } } 2R risk per trade Look at the following examples to see why win rate is not everything: System 1: A 1:1 RRR System ï 80% win rate, 1:1 RRR. ï This strategy has a 0.6R expectancy. 0.6 times the risk is the profit per trade. $60 profit for every $100 risk. ï The downside is that the edge is vulnerable to shifts in win rate. If the win rate is that high, the system is overfitted in most cases too. System 2: A 1:4 RRR System ï 32% win rate, 1:4 RRR. ï Such a strategy will have the market on its knees while still maintaining the same efficiency as having an 80% win rate on 1:1 (very unrealistic and hard to find). ï As a rule of thumb, finding a 32% win rate with an average RRR of 1:4 is a lot easier to do compared to having a high win rate with a lower RRR. ï This strategy has a 0.6R expectancy too. The downside: Your edge is more vulnerable to shifts in costs because the stoploss size is already small. Spreads have a larger effect on smaller stop loss sizes versus larger ones. Costs must be calculated and accounted for properly (use the spreadsheet given in the Discord). Accounting properly for costs is essential if you want backtest and live performance to align. This expectancy/expected value framework can be used to compare and filter strategies. With enough practice, traders can develop their own original entry ideas and test them in a structured way. This is how you get powerful systems, frequently. Your win rate does not hold weight on its own. Your expectancy is a much better measure to use to ensure a positive P&L. Figure 1: This is what happens to a 1:2 RRR 50% winrate system when high costs are applied. Over 60% of the profits were lost to costs, and the longer the system trades, the worse it gets. A lot of scalpers face this problem, and it is not acceptable; this is one of the key reasons we do not condone super short-term trading practices. Note: This figure assumes that the edge is persistent without any real-time edge decay (This is generous to the strategy and over 60% of the P&L is lost, real-time examples are typically worse. To help this mini-section click, I have added some examples below to show what more acceptable costs can look like. Figure 2: This is the same strategy with half the amount of costs. (Still high, but closer to acceptable.) Figure 3: Here is an example with the black line showing the mean (avg) of a single outcome with a 1:2 RRR 50% system and the grey line is with the trading costs applied. To provide you with a more realistic equity curve, I performed 100k simulated outcomes and selected the equity curve with a peak-to-trough drawdown at the 90th percentile. Nine out of 10 equity curves with a 1:2 50% win rate will experience a maximum drawdown better than what I have displayed on this figure. This is not something traders can afford to ignore. The effects trading costs have on trading performance is far more profound than most traders realise, as they can quietly distort outcomes, erode genuine edges, and can easily turn an otherwise viable strategy into a losing one, so take this information seriously and account for costs with care in both testing and live execution. 7 Trading Platform Stimulation And Colours 7.1 Why did we choose to use black candle charts? Neutrality. When we look at candlesticks, we do not want to feel anything; it is a dataset, nothing more. You can tell yourself that red, green and blue donít affect you but subconsciously they are stimulating colours; besides that, it looks unserious too. Ali and I have seen many people trading in public. The lack of colour in black and white charts attracts less attention in public. Red and green or green and black candlesticks signal a lack of seriousness to established traders. MetaTrader 4 and MetaTrader 5 run the clown palette and nobody notices. Red, Blue and You donít need a difference in colour to see your setup. Let us be direct. Why is your chart black and white? Trading doesnít need to be colourful. Black and white is enough; you are there to use OHLC values and other information to place data-backed trades. Nothing else There is endless research showing that bright colours, especially red are stimulating hues. Why would you want to be stimulated even subconsciously when trading? Even if it doesnít develop into action, it can cause additional unease during periods where joy or stress runs high. it is a pointless additive. This is why we hate most retail platforms. To us it seems like something used to gamify the operation. Changing candle colours will not rescue a system with zero edge or fix serious psychological issues. See it as a low-cost way to shave off a bit of emotional noise instead of a magic switch. Do you like them, or did you force them? Pure black candles can appear extreme, but Ali and I have gotten used to them to the point of preference. How can you tell if a bar has gone up or down? You can tell which bars are îbullishî and which bars are îbearishî by seeing where the candlesticks start and end. Okay, I want to do this, how do I start? You can try black (down) and white (up) bars. You can tell which bars are îbullishî and which bars are îbearishî by seeing where the candlesticks start and end. Remove any radiant-looking neon indicators (if any), and keep any lines thin, faint, but visible. You are here to execute, not obsess. The pure black and white layouts we use are to make trading charts look how theyíre supposed to. Dull. The takeaway: If you want true indifference when pressing the button, small adjustments and sacrifices must be made for optimal trading psychology. Bright colours are noise dressed up as signal, we can do better. 8 More Nuanced Questions Dozens of hours of research and experience, condensed into two minutes of reading per inquiry. 8.1 Question Regarding 3 Wicks Counter Trend Applied to a Mean-Reverting Market Statement: ìI have no clue whatsoever if this order filling is actually the reason for why price retraces back into the range.î Neither do we, and this method has a consistently low win rate (below 50%) but does so with high RRR of greater than or equal to 3 whilst trading against the short term trend (mean reversion). The reason why the market exhibits extended market movements (what trends are, even short-term trends) is rarely the imbalance on its own.[2] When smaller traders try to apply meaning to every move or why it happened, especially in real time, it holds them back, as this is ad hoc reasoning, a logical fallacy. You need to create or discover logically sound ideas and build systems around them. What do I mean by [2] 1. Market movements over longer time horizons (many minutes to days) are influenced by something happening under the hood (Market microstructure), for example, the way liquidity is provided in the market you are trading. (researchable) 2. 1-minute charts do not have low autocorrelation (high randomness) for a reason. 3. YM Futures / Dow Jones does not mean-revert intraday for no reason, which is ex≠plained later within STS literature rigorously in the psychology esction. 4. Certain metals are trending or not trending intraday for X, Y, Z reasons that will persist. 5. Every single mechanism mentioned Once you understand the process and reasoning behind each of our generalisations, which I use to create specific strategies, it will clear the path for you to create your own. The three wick counter trend (one out of many execution techniques) is more about getting a consistent point of entry, where it is possible that the trend may start reversing. If it reverses, you collect a large movement relative to the stop loss distance risk. The most important part is getting a consistent point of entry. First, we focus on having it be an efficient point of entry. Then it is about a reasonable, consistent target. And finally, you need an average low-cost exit on failure (stop loss placement method). 9 An Advanced Question Regarding Our Framework Question: In trading system design, the rules used to define events or patterns are often subjective, even if trade execution is mechanical. This creates a risk of bias: patterns may be defined because they appear to stand out in historical data and back testing may simply confirm those choices, which may resulting in a circular and non-robust strategy. How can this bias be avoided when defining events and validating trading systems? Definitions and Implications Non-robust: Weak, the strategy will not perform in real time, e.g., overfitted. Why this would be a serious violation: If there is no objective reason a ëpatterní should continue occurring, it is a data artefact, not an edge. For example, noticing that on rainy days the market was more likely than on non-rainy days (before 4PM) to reach the average price by 5PM last year does not mean there is a legitimate cause for the pattern to continue. The weather does not predict price, nor do random candle patterns or indicators in isolation. That is why falsifiable cause-and-effect relationships are so important. Circular: The belief that something is true because it happens to be so, without a falsifiable basis. e.g., My strategy works because the data says so, without any fundamentals backing it up. (The very thing the STS framework protects you from.) Why this would be a serious violation: Something real can only be taken advantage of if there is a legitimate, objective reason it will occur tomorrow; otherwise, it risks being noise that will not repeat often, resulting in losing strategies after costs in real time due to weak foundations Answer: STS traders are encouraged to separate edge discovery from confirmation entirely, we donít tweak our systems as we go along. It is an important aspect to our process. As soon as a trader deviates, they get overfitted systems; This is reinforced within the material. We operate with a limited set of foundations on every strategy, which are super hard to game. If you try to, the deviation is far too obvious, as it would require total strategy restructuring to suit a narrative. If it doesnít work how we idealised, we move on. Most traders who do this get it wrong by under fitting, but what we do is use specific sessions, anchors (entries, risk management methods e.g. stops, price structures, indicators) and other foundations that are central to that specific strategy. Nothing vague or intuitive is permissible in real time. No maybes, there is just îit isî or îthere is notî The second this is ignored is when strategy design loses its rigour. To avoid narrative waffle, we avoid going from intuition to definition as it reduces tests into a process that exists to validate a belief system instead of a strategy, which is a key trap. Micro Q&A before we continue (a key nuance): îYou say that Dow Jones exhibits mean reversion intraday, but so do other indexes, such as the Nasdaq right?î Answer: Yes, that is true with most markets on super low timeframes, but NQ trends are more abra≠sive, which naturally spills into lower timeframe price action (more trendiness and sharper reversals). If intuition is used in strategy design, we run it through one of these reasoning paths Figure 4: System Class 1: An STS Deductive Reasoning Visual System Class 1: Basic (Default) Intuition (if you must), then we check it shows up in research (objectively), only then entries are designed or picked for it (not dependent on data, but how the entries work). To be clear, entries should not be ífine-tunedí as that is overfitting. We pick or design a suitable (away from the chart/data), and we aim to pick a suitable filter (We have a list of those as well). The end goal is to develop your own process, entries, filters, everything. You will have your own thresholds and rules beyond those universal guidelines to regulate this in a way that aligns with your unique builds. We actively advocate against edge discovery through charts to avoid this problem, as itís the no.1 reason why most retail traders backtest and still get awful results. We see OHLC only as a tool to confirm an idea/edge is real. Discovering edges through charts can happen accidentally, but it will often result in strategies that lack logical foundations to have any genuine edge after costs. Familiarity with the shape/geometry is what looking at OHLC should be used for only as that is required to create ideas. Now we will move on to the nuanced secondary system class. Figure 5: Additional Steps For Class 2 System Class 2: Advanced (Rarer, fewer opportunities per year) Intuition (if you must) then you try to define it by using a few basic measures or data relating to the unique situation (from official reports) if it is not well established in research (niche) if it is based on price discovery test on unseen data only. We want to produce profitable strategies on purpose 1. We will create strategies with OHLC data from candlesticks which to us is a simple preference among many paths. 2. Positive Expectancy after out of sample confirmation is our edge discovery. Question: Do we actively try to define current price regimes? Answer: No Our job isnít to predict regimes from market data alone, as we doubt that can produce an edge when only market data is used for inference. It doesnít matter if itís standard charts, order flow, or something else entirely, a predictable mechanism is required for support. Regime shift plays are about playing when the market deviates based on a fundamental change that will stay and data for existing strategies reflect that through increased effective≠ness. You benefit from a shift in how price should behave by fundamental or foundational changes regarding the instrument itself or what is involved in pricing the instrument. Examples present within the materials are tariffs, oil, gold and a realistic virtual scenario. I navigate them all on how I went from generalisations (acknowledgement) into something specific (your strategies) in the regime shifts document. For the regime talk the perspective is like this You are doing a mean reversion system on Oil [1] for example it is extremely effective, but you also have a trend following system on Oil [2], also one that trades reversals that trail the stop loss [3] You are running the mean reversion one as itís most compatible with typical oil price be≠haviour with researchable microstructural reasons to back up the decision. System 1 is performing well System 2 has mediocre but profitable performance System 3 is profitable overall but not stable or suitable for the typical price behaviour. The mean reversions took out the trailing stop too early in normal conditions. There is a deviation from normal conditions system 3 which was on inactive previously, and it is suddenly performing extremely well. You research and discover that fundamental things are changing in oil that will likely last months to years not weeks it increases the volatility. A news headline is not enough, you have to look more into it and manually review. More volatility = more trending movements and larger ranges. This is the regime shift. Figure 6: Each line represents the average daily return of an individual strategy, calculated as the mean return over 30 days. These allow for strategy structures like [3] to be more effective than usual intraday, same for strategy [3] as shown in Figure 3. Because you can see it performing well in real time the narrative is system one is what got deployed System 1 is intraday mean reversion on oil (example) System 2 and system 3 are on inactive With system 2 and 3 you engage with them when appropriate to do so which requires a trigger such as a fundamental change Only if itís low time frame and thereís a strong correlation to confirm if the event in question doesnít have the longevity for the state to be able to be collected before itís over then it is just noise to ignore. This is why I describe these opportunities to deviate as rare. Without strict conditions and rigour, it is just dice rolling and strategy hopping. Question: Do you ever prematurely (before max DD) stop trading a system, because of regime change reasons. Answer (Common): 1. We keep trading it and run the other one on a separate account to benefit. Full replacements happen when 1. We are forced to because the event in question makes the strategy fundamentally untradable e.g., spreads results in costs exceeding 30%. Premature 2. Over a medium sized sample size e.g., 3-9 months the return distribution deviates by a noticeable amount when compared to a backtest. Premature 3. A full blowup occurs 4. Our constraints change e.g., if I want to do more travelling and theyíll be more potential inconsistency 5. Break from trading (also rare and doesnít guarantee the need for replacement) Question: There could be two entry techniques that would be suitable/logical for the strategy and market, but one can perform extremely positive, and the other one can perform horribly, how is that possible. Answer: One was more efficient in comparison to the other and better suited for how the market is forming prices now (universal) For specific answers you have to take whole system(s) into account. Case-by-case basis. The point of testing is to see which one actually works, not every logical system works, but we at least need to make them all logical to figure out which ones work and which ones donít. In this case, out of sample data would be required to see if the profitable strategy survives, if it collapses completely out of sample its effectiveness is a coincidence. 10 Key Lessons From Years of Chasing Patterns Instead of Understanding Markets The lessons packed into these five minutes took me ages to learn, and many traders never figure them out. After over four years of coping with îThe London breakoutî, ìinitial balanceî, ìopen range breakoutsî and other common strategies, I realised the assumptions were doing the most damage to my chances of success. The turning point came when I stopped asking ìdoes this work?î and started asking ìwhat market behaviour would make this work?î Definitions: Data mining in trading is the process of searching historical market data for patterns that appear effective. Mechanism is why and how something occurs which is important for true understanding. Overfitting is when a trader attempts to íoptimiseí a strategy to perform better on past historical data which never repeats 1:1. This causes poor real time performance. These flawed adjustments are referred to as ícurve fittingí. Coping (In this context) giving into context as itís easier to grasp and accept versus facing reality. Expectancy/Expected value, how much R you average per trade including both gains and losses. For example, (2+2-1-1)˜4 = +0.5 R Expected return per trade / 1:2 RRR 50% winrate (Excellent). In this example each position returns an average profit of +0.5 units of risk [R] taken on average. Out of sample data: strategy performance data collected out of the main testing period to measure robustness in different conditions to check robustness. 10.1 A Conversation: Lessons Learned and How I Grew Data mining The high of the day is likely to be registered within the first hour, this is something Iíve known for years, but at the time I couldnít make an effective strategy from it. The variance is too high. I also tried something for the hourly range where the low is formed, but this was so long ago it is crazy. My struggle I thought to myself, îperhaps I am coping with data miningî and moved onto the next thing. I tried so much madness, including íinitial balanceí during those hours, with zero success, as well as ORB. All this saturated waffle is îoverfitted if adjusted to workî and ìcompeted awayî by algos as the market is highly efficient, or so I thought. So the only way to use these properly is if thereís strong underlying support for its use in real time (at the point of deployment) [1], which might be 6 months a year avg with a specific ruleset assuming If you want a good expectancy (good price regime) but back then I didnít know fk all. Why we stay away from frameworks like IB/ORB For example [1] in a market where mean reversion is tight, fading IB for mean reversion will be profitable indirectly. These setups will happen to work but donít say much when you are looking at price data regardless, each educator seems to have their own narrative with îtrapsî etc. That is hot air to us. The problem with things such as initial balance is there isnít an objective reason it should work, any edge derived isnít from the setup itself; it is indirectly from the price regime. We prefer isolated, consistent setups where price has been dominated by aggression and pauses or there is a dislocation and rebalancing as these show something and provide objectively superior prices and costs compared to standard market orders when itís favourable to long/short[2]. All of this can be predefined, mechanical and filtered. In our materials, we describe this type of entry we design as a îMicro auctionî.[3] The îMicro Auctionísî function is solely to mechanically get consistent low cost fills at more predictable prices vs market orders. The logic may seem restrictive, but we have designed over a dozen price action entries, and we still create new ones occasionally (all microstructure-based). What do I mean? [2] Examples: 1. If price is mechanically íoverextendedí (defined by your rules) it is favourable to short via mean reversion. (Data and the market confirms this at the point of execution). 2. In a mechanically defined uptrend in a trending market it is favourable to buy after a dip has partially rebalanced with a setup to anticipate continuation.[3] 3. When price has partially rebalanced, a filter based on predefined price structures or maths (indicator) will either automatically green-light the position, or the pullback lower will have been too small for the filter to flip. Regardless, we would be looking to go long.[3] No real time intuition would be required, all rules and decisions are baked into the rules before deployment. What these retail frameworks taught me to do The best way to learn is to study the mechanism of why and how these things worked (if at all), and use that understanding to research its fundamentals. When you soak this up, you can use it to help you create sound, unique (important) ideas a lot faster. It is important to avoid falling into the confirmation bias trap as a lot of the time, there is no advantage. The supposed edge purported is often marketing with a nice narrative. The reason we are so vocal in regard to AI/LLM flaws is because it can reinforce faulty ideas that validate traders instead of calling it out. This often reinforces mistakes, which only inhibits growth and hinders profitability. If you ask AI loaded prompts (most people do) e.g., îhow does this system have an edgeî AI will tell you íhowí it has an edge even if it is nonsense. The danger Only someone with the knowledge regarding the mechanism(s) in question can resist nonsense consistently. This information filter can only be applied with genuine understanding, which is why we push it. Guidelines on AI/LLM usage will be provided later, A Basic Example: That is reflected in the well established mean-reversion behaviour during the open. You can fade overextensions on smaller timeframes, or wait for trend continuation after the open on a mid-timeframe (for example, 15 to 30 minutes). Designing robust long entries is far, far easier. Donít be afraid to run long-only models, especially in conditions like these. The real reason it happens is that thereís a consistent liquidity shock (each day) led by market takers at the open. Market makers (MMs) that provide the liquidity and offset it (ideally) within the same window or later MMs wonít want to provide as much size due to the liquidity shock (adverse selection) causing nasty spikes or ranges as MMs are less willing to put up liquidity, making it easier for market orders to move price without intervention from passive participants (placing limits). The reason this is good to know is that you can see the mechanism is real instead of an illusory bias. If it is objective reality, you can design something to take advantage of it, which reduces the likelihood of overfitting or wasting time. Think of this as a generalisation, then work your way down from there. This is an example of deductive reasoning showcased within STSí literature, and every asset has different angles to approach from. Thanks For Reading -Ron

Strategy Engineering Market Microstructure Principles

How Markets Really Work Sentient Trading Society In this document I will break down the difference between real market price formation and common retail perceptions. 1 The Contemporary Retail Perception ï Price builds up liquidity, gives retail fake patterns and trend lines and breakouts and îInstitutionalî traps to enter into. ï A îcentral algorithmî moves the market price to where their stop orders are placed. ï Build up more false entries for retail. ï Return to interact at a demand zone where real orders supposedly rest. ï Price induces to trigger stop orders followed by a reversal. This is a retail trading framework where people believe that price is supposedly controlled by a single central algorithm moving between ìliquidityî and setting ìtrapsî, along the way. This works as a storytelling model for retail traders, but it does not reflect how actual price formation and liquidity provision work at the CME, interbank FX, or equity exchanges. When substance is asked for, it often met with anecdotes or circular reasoning instead of objective mechanism(s) and research. Even market makers do not know where price will go. They only manage risk across prob≠abilities. Modern trading influencers change that into a deterministic narrative where price moves here to grab liquidity, then there to interact, then runs to the next pool. This struc≠ture feels logical and predictive, which is comforting, even if it does not reflect how market pricing works. To keep things simple for now we will define a market maker as an algorithm that provides prices that people can buy and sell at (quotes) which provides îliquidityî, it is more nuanced, but we will explain it in more depth with exam≠ples and figures in more depth later. 1 2 Real Liquidity Provision and Price Discovery A Mini Glossary Liquidity provision: Adding orders, usually limit orders, to a financial market, providing liquidity and making it easier for others to trade. Price discovery: The ongoing process where supply and demand interactions determine the market price of a market or asset. 1. Retail makes up 5% of market participation in Forex, making it extremely efficient. 2. Too little liquidity to manage MM inventory. 3. Too little liquidity from retail. 4. Not how liquidity provision works, and this is not how price movements work on a tick by tick basis. This is fundamentally wrong. 5. Supposed orders. 3 How FX and CFD Liquidity Actually Works Most FX and CFD brokers internalise most of the volume: Most liquidity offered by retail traders is not exposed to the actual FX market. Many brokers take the opposite side of their clientís directional risk. The broker acts as the counterparty, absorbing losses as their gain and vice versa. There are good brokers that do not do this, but this is explored later in the materials without any conflict of interest on our part, as we are not affiliated with any of these brokers. Internalising order flow and offsetting book risk: Even most FX brokers that offset their flow hedge at their discretion with risk limits or hedge their inventory imbalance at market to maintain a delta neutral book. Delta neutral book simplified example: 5,000 lots long, 4,800 lots short, and 200 lots short at market. By offsetting imbalances, brokers, and liquidity providers aim to stay close to a volume net zero directional exposure. Earnings and business model: FX brokers and liquidity providers are close to net zero exposure with marked up bid ask spreads, commissions, and spreads earned for their role in liquidity provision. Market manipulation is real but subtle. Market makers do have predictive models for liq≠uidity to increase market inefficiency and for arbitrage, but they are inconsistent due to distributional decay, making instances of stop hunts and other events similar to anecdotes and coincidences. It is a mixture of confirmation bias and ad hoc reasoning. Unless there is a large payment for order flow scheme where stop loss data specifically is sold, the burden of proof is on the accuser to provide evidence for targeted behaviour during the price discovery process. 3.1 Why No Central Algorithm Controls Market Prices 1. There is not a sole liquidity provider or market maker for Futures (Direct Market Access) or FX/CFDs (Over The Counter) 2. An algorithmic ídelivery mechanismí would imply stable timing patterns, but order arrivals and limit order queue priority at microsecond scales are largely random because how markets discover new value constantly changes. 3. Firms entertaining a deterministic pull to liquidity would suffer a lethal amount of fading because of the predictability. For an institution, funding an operation like this would be equivalent to donating money directly to faster firms. This would be arbitraged, swiftly eroding any edge in the process. 4. If a universal algorithm was responsible for price movements, identical markets across venues would print the same path, yet persistent cross-venue divergences and lead-lag relationships exist, creating price discrepancies which HFT algorithms, funny enough, close. ES-SPY price dislocations are a well-documented example. Figure 1: A visual from The High-Frequency Trading Arms Race: Frequent Batch Auctions as a Market Design Response, The Quarterly Journal of Economics [4] 3.1.1 Reality: When market makers adjust their quotes, it often makes the price tick or causes reactions that influence future price movements in the short term (sequential market inefficiencies). When makers pull or imbalance their liquidity, there doesnít need to be an imbalance between buyers and sellers for the price to move a tick. Algorithms are notorious for creating vacuums that can cause inefficiencies to cascade across multiple timeframes. A Common Claim: îThe big players, such as investment firms, will absolutely create fake 20-point runs to engineer liquidity.î This is a common misconception. Reality ìTradersî at banks are positioned to tactically fill orders and manage risk; their primary job is to handle flow rather than speculate on price movements. Dealers survive by keeping directional risk low while collecting commissions. Here I explain it. If the price is near a liquidity cluster, they might engineer the liquidity by making the price move by, e.g., 1-5 ticks if it is believed to be highly probable. However, 20+ ticks is too risky, and itís not worth the inventory risk they have to put up. Moving the price isnít free, and brazen movements like this result in either fast market participants going against their trades, providing liquidity in excess against them (absorption), or firm position liquidations. The potential gain is too low; it is an asymmetric bet. Additionally, if MMs or other market participants fade that flow, theyíll get absorbed and be forced to offload at a loss. The point is that they believe liquidity is close, and it can be induced, theyíll take it, but they rarely engineer it over large price movements; it is too risky. When I said, îFirms entertaining a deterministic pull to liquidity would suffer a lethal amount of fading because of the predictability.î this was the phenomenon I was talking about. Market makers where possible will position themselves to benefit from stop clustering and to avoid aggressive order flow but MMs do not engineer movement to take that liquidity like purported by retail trading educators. Confirmation bias makes it a satisfying fallacy but we must remember that there is causation and there is correlation and they are not the same. 3.2 What would have to be true for this narrative to be real? Centralised markets, such as equities and futures, would instead have to adopt an extreme form of a Continuous Single Auction (CSA) where you could only interact with a dealerís quotes using market orders. Only the dealer would provide liquidity and traders would have to accept it, the dealer would set all prices, and limit orders would be treated as pending market orders. There would be a single venue and exchange, no off-exchange market makers such as Citadel Securities, no dark pools, nothing. This is the opposite of reality. Dealer (Definition): A dealer is a firm (or desk) that provides two-sided prices (bid and ask) and is willing to take the other side of your trade. They îmake a marketî for people to transact on by acting as a counterparty. After the trade is executed, they often offset directional risk through hedging. Modern liquid markets are a continuous double auction, where traders, algorithms, and other participants are free to provide liquidity through limit orders to other participants who use market orders. This is called a Continuous Double Auction. (DSA) What about decentralised markets? (Over The Counter) For decentralised íOTCí markets such as FX, there would have to be one central spot currency dealer. The reality is that many institutions, e.g., JPMorgan Chase, provide spot FX quotes to be taken by market orders. FX is not remotely close to having a central 1:1 tick-for-tick price movement. Even when comparing one bank to another, without aggregation of prices it is not 100% constant; each dealer has different priorities, inventory risk, and adjusts quotes accordingly. 4 Contrast and Summary Retail Influence Market Perception: A narrative where Institutions îdeliversî price moves to trap retail and collect liquidity. Actual Market Microstructure: The price is a derivative of continuous order matching, bid ask quote adjustments, and risk management by liquidity providers e.g., JPMorgan Chase and multilateral trading facilities for example, LMAX Group in Europe. This material is easy to research if you want. You can explore it further. Use these terms: ï Market Maker ï Market Taker ï Liquidity Provision ï Price Discovery ï Market Microstructure Remember, price discovers quotes. Added Nuances for Clarification For other asset classes, retail volume or participation is 10 to 15 percent depending on the source, alternatives to FX have price movements which are less random, which is better for us. Let us move on to Market Maker nuances. Think Market Makers Are Hunting You? Hereís How They Actually Work. Market Microstructure Principles & Warnings Financial Market Realities Ron Most traders only think about market makers in terms of market manipulation. But market makers are largely your friend. Without them market pricing and costs would be chaotic and inconsistent Everything in this post has been discussed in institutional-grade literature. (listed at end) In the past Iíve read multiple books and papers on HFT behaviour. By the end, you will know: Why we need MMs to facilitate trades in modern markets How îstop huntsî or îsweeps of liquidityî actually work Retail misconceptions on MM behaviour Ways to mitigate vulnerability to market noise indirectly caused by MM activity Only the necessary institutional language and definitions will be provided with zero discrep≠ ancies. Figure 1: Simulation of a market maker algorithm quoting around price 1 1 Why you need market maker algorithms for low trading costs Every time you place a trade in any market, you are relying on someone else to take the other side you need sellers to buy at each price and vice versa without market makers con≠stantly providing liquidity, automatically spreads would be wide, order books would be thin, volatility would be uncontained and costs for execution would be higher and inconsistent, making markets very inefficient. Market maker algorithms are designed to continuously quote both buy and sell prices in huge volumes, smoothing out rough edges making markets more efficient overall. often in fractions of a second. By doing this, MMs provide liquidity where there would otherwise be gaps, they also help correct these inefficiencies. The result for us is smaller bid-ask spreads and more consistent fills for traders of all sizes. They get paid to provide liquidity and we get lower costs so itís a win, win! To add, markets without MMs are less liquid the potential for slippage is obscene. Market maker algorithms quote both sides of the market with limit orders. In a liquid market with market makers, spreads and slippage tend to remain low. When you buy with a market order, you are taking liquidity from someoneís limit order (such as a market makerís), and vice versa. If there is not enough limit order liquidity at the price you want, your order will experience slippage as it trades across multiple price levels to get filled. Market buys and market sells are not matched 1:1, which is a common retail misconception; trades are matched against resting liquidity in the order book, and a single market order can match with many limit orders (and vice versa). 2 How real îliquidity huntsî work (real example) A market maker algo has an imbalanced book at price 20000. (The MMís in≠ventory is net-short.)[1] Simplified Futures Market Example (Linear) The MM needs 400 contracts long to balance his book to zero with minimal market impact The market maker anticipates that at price 19999 there are 1000 contracts that will be executed on the side he needs to get out the trade with zero market impact He knows that he needs 200 contracts to move the price lower to the price of 19999; he does (short 200), and that and the liquidity is taken by market participants, including him; he buys 600 contracts back and pockets the difference, And then price spikes back up = 20000 People would say that the MM algo here îhuntedî liquidity, but in reality they do this to neutralise their risk and are completely neutral. Market makers earn the bid-ask spreads and move on. They arenít invested in long-term price legs like traders are. It is very rare that these adjustments happen over large price ranges. When people say îLow timeframe noiseî, this is the cause! This happens on many price levels and is not exclusively related to stop orders like retail educators purport; itís random and cyclical, happening all the time. usually stop hunting is a coincidence; itís not malicious or intentional; it just happens, just like dealing at any other price level because they front-run flow. Liquidity anticipation is a key thing Market Makers do they make money by providing liquidity. The same thing could be done to anticipate profit taking, but nobody calls it ítake profit huntingí. Confirmation bias makes retail traders want to believe their stops get îhunted.î The point is the event it-self is neutral; they typically donít care if the market participant is realising a profit or loss. All that HFT MMs try to do is quote prices for market participants to deal at whilst keeping inventory risk low, managing adverse selection, etc. Main takeaway: If this happens with your stop loss, remember itís a usually a coincidence in regulated liquid markets especially in Futures and US Equities. Figure 2: Retail perception of Market Makers . 3 Strategies like this do not mimic true MM behaviour This happens several times per day regardless if trades are filled, profits are taken or losses are realised, but trading educators will frame it as îmanipulationî. remember the example [1] shows over a small movement relative to the price only 1 handle / one point / $1 price movement thatís it. 4 Performing these îLiquidity huntsî over larger price movements rarely makes sense for MMs. Hereís why: The marginal expected gain versus the expected inventory risk and potential adverse selection is hardly favourable enough to perform stop hunts regularly on liquid, regulated markets. By committing a lot of volume, the Market makerís liquidity can get used or front runned by faster or more informed market participants. îMarginal expected gainî represents the additional profit expected from a market makerís decision, considering the probability and risk of the outcomes. Retail narrative: Retail educators say that market makers will make large movements to take out the stop losses that are far away from current market quotes, which is absurd because if their volume gets absorbed, theyíre stuck with elevated inventory risk ex. stuck in a 1000-contract long, which would move price further against them if they needed to close their position out in a loss. Even a 10-point move on index futures is large for a market maker. Reality Letís make the current price 20010.00 and the price in focus 20000.00. -10 handles. If a predictive HFT MM Algo anticipates theyíll be 3000 contracts 10 handles / $10 away from the current price and the algo anticipates the market impact per handle to be 200, leaving a +1000 contract discrepancy if the price is met, they wouldnít commit the 2000 contracts to spike the price most of the time even though itís logical because the inventory risk accumulation or chance of adverse selection would be too high even if they spread it out. They could be stuck with -2000 contracts on the wrong side of the market and lose a lot of money; all it takes is for a different algorithm to match their flow to nullify their market impact completely. Hereís the nuance, though: if the price was already trading at that point thatís $10 away from the current price and their predictive model still supports the decision they could provide liquidity at 20000.00 but also influence the price to trigger the orders but only if close and highly probable. For example, if the price is at 20000.50, they could sell a couple of hundred to flush the final buyers to trigger the anticipated order flow. The point is itís extremely unlikely for Market makers to influence larger move-ments/spikes to tap into anticipated liquidity unless the level is extremely close to where price discovery is taking place already. So itís the other market par≠ticipants trading towards that level, thatís the true causation, not the MMs. 5 So what do I mean? Dealing with larger price ranges both on your stop and target size lowers your exposure to the noise introduced by these rebalancing behaviours. The further away your initial stop is, the less likely it is to be taken out by a market maker rebalancing event (e.g., a 5-handle stop vs a 12-handle stop). This is why I donít personally trade timeframes below 5 minutes, and if I was planning to I would make sure the minimum stop size is wider to mitigate costs and to reduce sensitivity to noise. 6 Understanding markets and what to unlearn 6.1 Outdated market opinion and how it weighs modern traders down Whilst market academia is insightful, it typically doesnít include key nuances of real markets; it focuses on theory, so to find the truth, you have to look at multiple peer-reviewed publications and books to answer individual questions. An example of this would be price discovery and the concept of ífair valueí, which are often described in a utopian fashion. Theories often assume that markets are efficient or speak as if they are, when in reality, they are not. In the researcher role, it is challenging to filter out the noise, as markets are structurally different from those in 1995; yet many aspects remain the same, only more nuanced. A single peer-reviewed source just is not enough. Luckily we have done most of the work for you since 2020. (still ongoing) Our framework is the result of years of analysing institutional literature and rigorous quan≠titative testing. 6.2 Sentient Trading Societyís key points 1. Quality reasoning and trading edge before mindset. Once your struggles are studied and your numbers are proven, psychology stops ruining the show. 2. Treat the market as an averaging machine. Judge new systems on large, honest sam≠ples. 3. Backtesting hygiene is non-negotiable. It is either money or time on the line. Large samples, stress tests, no look-ahead bias, then measure expectancy and drawdowns rigorously. 4. Forward testing is not for edge discovery. Use it sparingly on lower timeframe systems to check execution feasibility, because alpha decays while you wait, and day trading edges (especially) are sensitive. 5. Execution is part of the edge. Account for feed differences and spread behaviour when designing your strategy. If you donít, you are finished, especially for over-the-counter markets like Forex, CFDs and crypto. 6. The market is a continuous auction between buyers, sellers (market takers) and the facilitators (Exchanges, Market makers, LPs and so on); they are the immediate cause of every tick directly or indirectly. That is a fact, regardless if the movement was induced from news or dynamics held exclusively within the market. Everything else purported is a distraction. Every movement or quote adjustment is a liquidity-related event. 7. Price discovery is the process of price trading higher or lower to discover ínew valueí, not ífair valueí. The concept of fair value in market literature is a subjective utopian concept that modern liquid markets donít tend to respect. Few think S&P 500 trading over 6500 in 2025 was ífair valueí yet many participated. Price is a negotiation. Not all deals are fair. Different market participants have different models, so the blended result for ífair valueí is never clear; if it was, markets would be 100% efficient. John Maynard Keynes -îMarkets can remain irrational longer than you can remain solvent.î 8. The market requires rigour; it doesnít reward candle geometry or false narratives. The market has no concern for how your candlestick looks or how you feel. For example, trend lines have little to no robust, exploitable predictive value on average, in the tested markets/settings, after costs and other realistic factors are considered. Years ago, I read this paper with over 50 pages proving it. 6.3 How common retail frameworks misrepresent market operations 6.3.1 They claim market movements are purely from buyers, sellers, sweeps and traps Reality: Price movement is also dictated by liquidity offered to participants relative to current buy and sell activity. Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 10002 50 Best Ask 10001 20 Mid 10000 Best Bid 9999 100 Additional Buy Limit Volume 9998 80 Table 1: Depth of Market snapshotfrom Section 6 In this example, if a trader buys 70 units, the dealing price (ask) moves 2 up ticks (last trade 10002 Ask) if there are no additional reactions, but the dealing price (bid) would not move a single tick if they sold 70 units; it would get absorbed on 9999. This imbalance in the liquidity offered can skew where prices go; there can be more units being sold, but the price still goes up. This phenomenon is often behind an îLow Volume Nodeî in volume profiling or îSingle Printî in market profiling (closer), for which the price tends to correct later. We revisit this later within this document. This DOM snapshot/illustration refers to futures with a central limit order book. For spot FX and CFDs, the same exact principle appears as visible or synthetic liquidity gaps rather than through a single exchange. (Liquidity gap = Liquidity inefficiency). If there is a small amount of sell-limit volume offered to buyers relative to buy-limit volume, itís easier for the price to move up aggressively. This is how high-volatility movements occur with low volume or pressure. Takeaway: Market price movements are driven by the liquidity available to participants relative to current buying and selling activity. Casual arguments about buy and sell pressure miss the mark. When market makers adjust their prices, it often makes the price tick or causes reactions that influence future price movements in the short term. When makers pull or imbalance their liquidity, there doesnít need to be an imbalance between buyers and sellers for the price to move a tick. Algorithms are notorious for creat≠ing vacuums that can cause inefficiencies to cascade across multiple timeframes. It is not as simple as a ësweepí and calling it a day. It is either market microstructure or story time, and the latter doesnít make money. A briefly aligned market is not confirmation of your narrative. 6.3.2 Common frameworks dress up the market maker as a villain. Reality: Without market makers your trade execution delays would be horrible, the bid-ask spreads you pay would be obscene, and spreads would gap and vary unpredictably. So-called educators play on tradersí insecurities and appeal to emotions by misrepresenting how market makers actually operate. There isnít a single algorithm; market makers exist as a body. Most market makers are independent liquidity-provision firms, and differences in their incentives, capital, and inventory risk lead to different pricing and behaviour. Analogy: In this example: The supermarket is a trading exchange Ex The wholesaler is the market maker MM The customer is the market taker MT Trading in modern markets without market makers is like shopping for groceries and be≠ing forced to go to separate wholesalers and farms to buy each item, instead of using a supermarket. You would face inconsistent costs and delays. One farmer might be offering a raw product for $3, while another sells it for $1.50. But with a supermarket/exchange that brings the buyer (you) and the sellers (farmers) together, the market becomes more efficient and spreads tighten as a result. The wholesaler/market maker takes on inventory risk from the farmersí supply and quotes prices to the supermarket to pass on to customers. Now farmers are offering $2.20 and $2.25 for the same product, and you donít have to waste hours to get what you need (execution delay) because price quotes and supply are mostly continuous and size is available; you do not need to wait for a matching counterparty each time. When inventory risk spikes, wholesalers widen price quotes or pull stock to avoid losses, just like in real food markets. It turns out electronic markets arenít so different after all. If everyone demands a product and supply runs thin, what happens to the price? It goes up. If there are supply chain issues, what do wholesalers do? They raise their offers. That sounds a lot like a market maker. They are kept in line indirectly by competition and inventory risk. If they operate poorly, customers shop elsewhere for the same exposure via a different venue. Wholesalers facilitate buying and selling, and the supermarket needs them for stable pricing, because thatís what the customer wants. Customers see $3 and $1.50 quotes scattered without the wholesalerMM , versus $2.20ñ$2.25 at the supermarket.Ex The spread is the round-trip fee for convenience. For over-the-counter markets like FX, CFDs and crypto, the dealer often earns via spread markups and financing rather than a commission. Context: bid at 2.20 and offer at 2.25, mid 2.225. Addressing Nuances: Yes, they anticipate order flow; yes, they fade your orders but for consistent, low-slippage order execution, they are necessary. As stated previously, MMs have little incentive to engineer trends, îstop huntsî or îliquidity sweepsî; their goal is neutral risk, so they take constant steps to mitigate exposure instead of increasing it. These algorithms operate within fractions of a second; they constantly reposition themselves for the next small range of ticks up and down, they do not engineer trends. If the movement up or down is too aggressive, they remove liquidity offered to avoid getting hurt. 1. What happens if the customer is willing to buy any offer with no price cap? Price spikes up. 2. What happens if the wholesalers are willing to sell at any price? The market will keep going lower until customers believe it is cheap enough to resume buying (this can cause crashes in value). The Takeaway: The same thing happens in financial markets. Debunking the market maker cartel conspiracy ìWhat about stop runs?î Yes, stops cluster at obvious levels. If price is already trading there because of other factors, a small shove can trigger a cascade. Market makers wonít push the price 20 ticks just to take a stop-loss group (crowd), but if other market participants are already selling into it, market makers will want to be first in the queue to provide that exit liquidity, since they are rewarded for liquidity provision. MMs are not the primary cause of price trading into stop clusters. Okay, I get it I know how Market Makers work but what can I do with this information? 1. Do not chase the market. Avoid joining that aggressive flow. Use limit orders to absorb it when you can. This keeps your entry costs low and predictable, increasing your profitability and order fill efficiency. 2. You must screen current and future strategies for false narratives that deviate from what actually occurs in markets. Market makers are not moving price 20 ticks to take a swing low out; without context, these systems typically return results close to breakeven without look-ahead bias. 3. The core lesson for us was that we should aim to provide liquidity tactically with limits, instead of only taking it with market orders. It teaches us that being part of the team that absorbs liquidity pays well. It is a lesson in letting rushed traders sell into your level. 7 So how do I use this knowledge to influence my trad≠ing strategy design? Genuine îstop huntingî is a rare market event, severely misrepresented by trading gurus to sell illusions and push false narratives. An MM moving price by a point to ìsweepî liquidity is not the same as an MM moving price by 10+ points to induce/sweep liquidity; itís far too risky for them to do that, with rare exceptions. Larger engineered moves as shown in trading guru videos are super rare because they would expose market maker algos to too much directional risk, except in very thin markets or during macroeconomic news releases. Provide and remove your liquidity tactically Try your best to make your entries at efficient prices, getting filled preferably with limit orders. The more often your winners get low drawdown before going to target the better. Anticipate the flow instead of being apart of it. I only use limits. If you are larger, you can use order-slicing techniques like we do or use pending market orders or other methods to get filled. Only let your orders get filled when your context still respects your hypothesis. Example: only get filled on limit orders during liquid hours during london and new york hours. Reframe your mindset Donít design strategies based on the idea that market makers are targeting retail stop loss flow because when it happens it is a coincidence and MM behaviour is largely inconsistent. Key Market Microstructure Lessons: 1. Market makers quote the prices you see, and market participants influence them. MMs govern the impact. 2. You placing a market buy order and getting filled is not someone else selling it to you with a market sell order; you are buying into someoneís sell limit order. 3. If all market participants, including market makers, pulled all of their bids/buy limit orders, the next market sell order would cause a flash crash, as the price would continue moving lower until a willing buyer is willing, and vice versa. 8 Final Note: Expect and accept the short-term noise from inventory balancing, and other events. Understand that HFT MM algos are involved in general price discovery rather than trend creation. Understand that algo-driven liquidity anticipation is largely cyclical and random to slower market participants because of their complex predictive models, so focus on adapting risk management rather than attempting to predict îmanipulationsî. Now we will escalate theory in to action. Now we will discuss how this affects our bottom line and how to mitigate risks when interacting with financial markets, considering the nature of these algos. We will start with centralised markets like futures and equities, and then we will explore decen≠tralised situations with spot FX, CFDs and other the counter products. This will help you avoid making key mistakes in execution, judgement, and positioning. Why The Best Traders Keep Their Strategies Private, Part 1 Sentient Trading Society Exchange-Traded Markets 1 Introduction Financial markets are not completely random. Traders who follow a disciplined, rules based approach, especially one grounded in price action, logic, and data, can carve out a real edge. But that edge is delicate. One of the fastest ways to lose it is by broadcasting the strategy or allowing it to become overcrowded. A real trading edge comes from staying ahead of predictable behaviour instead of partici≠pating in it. Sharing or selling a working strategy may inherently degrade it. This is why serious traders rarely share profitable systems widely. Strategies that truly work rely on consistent execution and a degree of uniqueness. NDAs in firms exist for a reason. 2 The Black Box Principle Once strategies become common knowledge, their effectiveness fades. It also explains why most people selling signals or trading systems are not offering anything genuine. They are often capitalising on hope instead of results. As soon as volume is predictable on the books, you are finished. This is about understanding the nature of modern markets rather than îbeating market makersî on an exchange. 3 The Nature of a Trading Edge or Profitable System A trading edge comes from consistently spotting opportunities where the odds tilt in your favour, where the potential reward is greater than the risk over many trades. These oppor≠tunities are not random. They show up in patterns or setups you can recognise and repeat over time. Whether it is through reading price action, tracking flow dynamics, or spotting order book inefficiencies, the key is finding those moments where the risk reward balance works for you. 1 Your edge exists only under the condition that: ï You execute it with negligible market impact. ï Your specific strategy and/or its triggers are not widely known or acted upon by a large number of market participants or a few with high trading sizes. ï Once a strategy becomes common knowledge, your edge dissipates due to crowding. 4 Why Profitable Traders Do Not Share Their Strategies If a specific trading system becomes widely adopted, the following can happen: ï A large number of market participants starts entering and exiting at the same levels, making liquidity concentrated and easily predictable. ï Market participants, especially market making algorithms, front run the strategy, which can erode a strategyís profitability. ï Prop firm expulsions. Most prop firms do not allow people to copy trade, increasing potential consequences for strategy sharing. ï People with conflicts of interest start taking advantage, with large volume benefiting. ìWhat if I get others to copy my trades directly. Would not that push the price in my favour, making the strategy more profitable?î Only in fantasy land. The more widely a strategy is used, the more likely it stops taking liquidity and starts providing it, often without the trader realising it. When that happens, you are no longer one step ahead. You become the target. Once you are the one supplying liquidity, you are more likely to get picked off by faster or smarter participants. Even in a high value market, for example Dow Jones/YM futures, if there is a day trading crowd and the guru enters before everyone else does, the liquidity is still predictable. If it is consistent enough the algorithms may anticipate and go against into the order flow to get filled. This could soften the initial expected spike or remove it entirely. 5 False Incentives in Selling Trading Strategies People often ask, ìIf your specific strategy works, why would you not share it or sell it?î Answer: because there would be no legitimate economic incentive if it were truly profitable. Any real trader understands that the mass adoption of a trading strategy, especially in instruments with limited liquidity, kills its edge. In contrast, those selling systems or signals usually fall into three categories: ï Frauds: Selling dreams and back tested fantasies like premium indicators, automated systems such as MT4 EAs, and individual trading strategies. ï Pump and dump operators (small market cap): The so called guru manipulates crowd behaviour to push the price temporarily, giving them a chance to exit with a profit after getting in ahead of everyone else. ï Online creators or influencers: Constantly posting strategies to collect advertising revenue from engagement and to direct traders to affiliated brokers and prop firms. 6 Why ìMore Buyers = Profitî Is Not So Simple While heavy buying can push prices up, it is the imbalance between buyers and sellers that moves the market rather than the number of participants. Remember, market movement is determined entirely by size and volume relative to available liquidity. Key Misconceptions: ï Support and resistance levels are often arbitrary. Breakouts occur not because of those levels but because buying continues after the level is crossed. ï If too many traders try to buy at the same level, they compete for fills. Many traders will experience slippage or remain unfilled. ï If market participants know that buying happens at a specific price, others, especially high frequency traders and market makers, can anticipate and trade against that flow instantly and faster than any human could. This is why predictable systems become targets for front running when crowded. Sharing is the easiest way to become the sucker. 7 Market Makers and Flow Anticipation Modern markets are shaped by the interplay between market makers, the liquidity providers, and market takers, the liquidity consumers. High frequency trading firms use algorithms to: ï Detect patterns in order flow. ï Quote prices that anticipate incoming orders. ï Modify spreads to price discriminate against predictable participants. Relevant citation: ìHFT may engage in predatory quoting strategies, or price discrimi≠nation, against impatient liquidity consumers by exploiting his order anticipation skills.î If you are following the crowd and acting predictably, you will become a target for faster and better equipped traders. It is not malicious or directly targeted. It is just how it is. Market makers will provide liquidity in excess at stop clusters and benefit indirectly but they do not engineer large movements to take your stop loss as that would be too risky (directional risk and fines). 8 The Myth of Orchestrated Buying Power It may seem appealing to have a crowd you can direct, telling them to buy when you do, but this fantasy fails in real market structure: ï You will likely not get filled at your desired price if many others try at the same time. This is even less likely for day trading systems since it depends on concentrated volume. ï Your actions become trackable and exploitable. ï Algorithms may anticipate the behaviour and either fade it or use it to exit their positions with minimal slippage. ï FX and CFD liquidity providers that are not direct market access often hedge client risk in real underlying markets to compensate for imbalances in exposure if risk is not internalised. It is important to understand this: 1. Market Makers do not have your individual order info. Order history is anonymous. The algorithms do not need your private order, they often react to public, observable behaviour (price action, order cancellations, market depth changes, correlated flow, predictable time and predictable tick ranges. 2. Market makers practice what I demonstrate in Section 8.1 to mitigate directional risk, rather than from any direct intent to cause harm. A Market Makerís ideal position is neutral and they actively aim to be risk neutral when aggressive buyers come in it interrupts that. 3. Market makers are not forced to take risk Market makers are not required to continuously quote both sides of the market all of the time. A quote is a bid and/or ask price shown to the market with a specific size available, which others can trade against when price reaches it, unless it is cancelled (a resting limit order). When it is executed an exchange of market exposure takes place, this is how buyers and sellers meet. Market Makers do not have your individual order info and past order history is anonymous. The algos do not need your private order, they often react to public, observable behaviour (price action, cancellations, quote changes, correlated flow and predictable timing) and on post-processed public information from other sell side firms, with minimum delay. 8.1 Market Microstructure: Collapsing The Theory Here is an example of a marketís order book reacting to a crowd of 200 traders using the exact same strategy at the same time trading small size (2 contracts per trader . 400 contracts). Example figures only. Remember, only the ask has to move for this to happen, the bid price does not need to move. A simple ask spike without the bid moving is possible too. Section 8.1 is one of many ways the same adverse event can take place. Here is the market before they step in to buy. To make this memorable let us remember this as a íCrowd-induced slippage eventí. Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 20003 95 Best Ask 20002 85 Best Bid 20000 75 Additional Buy Limit Volume 19999 85 Table 1: This is the order book and pricing just before the crowd hits the market buys, consider this a snapshot just before the market order wave actually lands. The bid-ask spread is now 2 ticks wide. The order book appears normal and liquidity is provided relatively evenly between the current bid and ask. Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 20005 50 Best Ask 20004 30 Best Bid 20001 70 Additional Buy Limit Volume 20000 55 Table 2: The best ask on the Table 1 was absorbed immediately, causing a liquidity shock. The 20002 and 20003 offer levels are instantly taken, the bid ask spread has widened by a tick, and the Market Makers reduced their ask size to avoid getting hurt. The crowd of traders have partial fills and are now experiencing slippage. 180/400 filled assuming they are the only buyers consuming the ask, the reaction has potential to be more extreme if other buyers step in without passive sellers stepping in with sell limit orders. The bid-ask spread is now 3 ticks wide. In the institutional world this is referred to as íAdverse selectioní. The hurt the Market Makers are trying to avoid is getting short exposure at potentially unfavourable prices. By moving the ask up and providing less liquidity to buyers, they provide sell volume at higher prices. In other words, Market Makers start quoting higher to reduce the chance of losing money in response to the aggressive 400 contract buy that they are providing sell exposure to, Market Makers can also choose to not quote at all (quote pulling). Market makers will act automat≠ically depending on perceived risk based on their exposure being net-long or net-short with the goal of being neutral (rebalancing). Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 20005 140 Best Ask 20004 20 Best Bid 20002 90 Additional Buy Limit Volume 20001 65 Table 3: Market Makers have reduced their offers significantly, but the spread narrowed by a tick because random market participants placed 20 briefly (a group) It is taken quickly and the offers move back up instantly. 120 filled. 300/400 Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 20006 60 Best Ask 20005 10 Best Bid 20003 105 Additional Buy Limit Volume 20002 75 Table 4: 70 filled. 370/400 Depth Of Market (DOM) Price Available Volume Additional Sell Limit Volume 20007 40 Best Ask 20006 30 Best Bid 20004 120 Additional Buy Limit Volume 20003 85 Table 5: 30 filled. 400/400 filled. Key: NumberOfContracts ∑ PriceFilled(Ask) Fills: 85 ∑ 20002, 95 ∑ 20003, 30 ∑ 20004, 100 ∑ 20005, 90 ∑ 20006 85 ∑ 20002 + 95 ∑ 20003 + 30 ∑ 20004 + 100 ∑ 20005 + 90 ∑ 20006 8,001,615 VWAP = = = 20004.0375 400 400 AvgSlippage = AvgF illP rice - P riceRequested AvgSlippage = 20004.0375 - 20002 = 2.0375 ticks Cost Impact and Crowding Risk (10 point stop, 20,000 price level) To put into perspective how bad this is, if these traders had a 10 point stop loss, they have increased their trading costs significantly on this trade (assuming the spread was 2 ticks without slippage), which will add up in drawdowns unnecessarily and erode the strategyís edge. If the person sharing this, such as a YouTuber, actually trades this system live with a larger order size, the slippage would be even worse. To top it off, this entire scenario assumes that nobody else was buying. Other buys make the situation even worse for the people trading it (bid-ask spread could increase further making costs worse for buyers if no passive sellers step in with sell limits). If the trading is predictable, people could attempt to buy just before they do, causing this effect before they get a chance to join the queue for entry. If there were 20 traders or a single larger trader buying 40 contracts, chances are they would get filled maybe without any slippage keeping costs lower. This scenario is based on 50k video views with a follow-through rate. 0.5% is a low value considering the view count but that is on purpose. If the strategy is shown in a serious manner, it is much closer to 1%. If the trader is established and shows data the number of live traders who deploy could exceed 1,000 (2% based on 50k views, modest). Treat these numbers like a conversion rate, it reflects how many people out of 50,000 were sold on the idea. Remember that 2 contracts is an average some traders trade with more capital than others, some overleverage and some risk low with micro-sizes. If they use limit orders, it is a different problem. The crowd of traders will be competing for liquidity in the same market, forced to rely on enough aggressive orders to fill the entire position. If price runs away after partially filled limit orders, people will chase the move with market orders, increasing costs or risk not being filled entirely. Other market participants may aggressively sell into their orders, absorbing their liquidity, using them to unload positions, or actively trading against them. Is this just a hypothetical arenít modern USA markets more liquid? Absolutely, but even 0.5-1 ticks slippage due to publishing your strategy is too much. Mar≠kets being more liquid softens the impact but it does not make the core issue disappear as a strategy sharer. Main Takeaway: If a serious trader is actively using their system the reality is it is not worth sharing en masse, it can only degrade their profitability, this is one of many potential consequences. 9 Summary Volume alone does not make you profitable. Order placement, timing, and order flow me≠chanics matter far more. If a strategy is widely known, it becomes noise or prey for better equipped participants. Trading ideas or rules where the logic behind the hypothesis depends on market crowding, for example traditional support and resistance or Fibonacci, naturally are not viable long term. The part people avoid talking about is what makes a price efficient. Every trade leading up to that point is what creates the price, and profitable strategies push the price closer to where it theoretically should be. This is the beauty of price discovery, but it is also a problem if you are a serious trader operating a saturated system. If someone is selling signals or their strategy most of the time they are not making real money trading. They are making money off you. If their system was decent and robust, they would be using it for themselves exclusively, and they would not want anyone else touching it. 10 Guidance for Traders What do you do as the trader. ï Create an original trading strategy. You can take inspiration from ones that exist, but the system must be your own. ï Do not curve fit your systems. ï Use logic and data. Back test your system without hindsight bias or curve fitting. Bar replay is best. Once data is collected, execute. Do not share. ï When you develop working strategies, do not share it or allow third party services to track your trading activity. Your trading activity is one of many data groups that can be shared with data brokers and other participants, even if set to íPrivateí. 11 Assumptions and Clarifications This assumes that the day traders using the strategy aim to enter at a similar price and have the same or similar stop losses and targets. That is, they are following the trading strategy as taught. This is about potential disadvantages surrounding fills on a tick by tick basis because of sharing. Taking inspiration from working trading ideas to create your own is not the same as copying the activity one to one. Creating a strategy based on a trading technique is different, and nothing is wrong with it. This document talks about copying a day trading strategy one to one. Our individual final strategies are private, but people can reach out for our individual entry techniques, filters and other processes we use. We share privately at our discretion, but it will never be public information. Thanks for reading -Ron 12 Context and Additional Reading Market Maker Versus Market Taker Key Information ï Market makers: offer prices to buy and sell, providing liquidity. Engage in arbitrage. Short term orientated. Earn a spread. ï Market takers: buyers or sellers taking liquidity. Traders, investors, producers, and consumers. Earn or hedge from price movements. High Frequency Market Making High frequency market making: the role of speed. Yacine A®it-Sahalia, Mehmet Sa.glam. Sources: Journal article View the full paper: SSRN copy Key Part: ìThird, we show how the HFT may engage in predatory quoting strategies, or price discrimination, against impatient liquidity consumers by exploiting his order anticipa≠tion skills, modifying the spread between his quotes in the process.î Alpha or Market Edge Decay Alpha or market edge decay and why no profitable trader would sell or give away their strategy for free. Julien Penasse, Understanding Alpha Decay Highlights that alpha, the edge over the market, tends to diminish. Alpha decay is generally a non stationary phenomenon. Julien leverages studied anomalies for credibility. Key Part: ìAlpha decay refers to the reduction in abnormal expected returns, relative to an asset pricing model, in response to an anomaly becoming widely known among market participants.î Why The Best Traders Keep Their Strategies Private, Part 2 Sentient Trading Society Over-the-counter (OTC) Markets 1 Introduction This document breaks down an educator sharing their FX or CFD trading strategy would actually hurt how well it ìworksî for them assuming they are actually trading it live and the strategy works (called edge decay), the realities of trading CFDs and retail Forex, and why liquidity and order execution often are not what most traders expect. 2 Clarifying Assumptions A specific trading strategy is not the same as trading methodology or idea. This assumes their system is profitable and the educator trades it live. The educator directs traders to his broker via affiliation or casual mention (CFD talk). Alpha decay means your trading systemís edge, your ability to make money, fades over time, especially if lots of people use the same strategy. 3 The Unique World of Forex and CFDs Forex and CFD trading is very different from futures or direct market access instruments: ï Liquidity: The FX market is huge ($6 trillion a day), but retail traders do not get to tap that full liquidity. Brokers might only have small inventories (like 5ñ20 lots on buy or sell), so bigger orders face slippage or rejection. ï Synthetic Order Books: Many CFD and retail FX brokers use decentralised or synthetic books, not a centralised exchange book. That is why prices vary between brokers, even for the same instrument. ï Price and Tick Differences: CFDs imitate the underlying assets but often have different tick sizes and pricing models (e.g., US30 versus Dow Jones futures), which affects fills and slippage. 1 4 Market Makers and Broker Mechanics Most CFD brokers are market makers. They create their own market and manage their exposure by hedging in underlying markets or with liquidity providers. Market making is not bad or ìtrading against youî. It is how they manage risk and keep their books balanced (delta neutral). They make money via spreads, commissions, and profit from overnight charges (not always). Different brokers offer different quotes and liquidity, which explains why prices do not always match across platforms. The difference between A Book (orders sent to the market) and B Book (internalised risk) brokers is complicated. Many brokers use a mix of both. 5 Practical Trading Issues with Order Size When trading larger sizes on CFDs or retail FX, liquidity issues become obvious. Orders might not fill at your target price or require market orders that cause slippage. Spreading trades across brokers or instruments can help, but has limits. Even I have run into fill issues on retail platforms when trading buy limit for approximately 125 units (25 YM contract equivalent). Although manageable, it is rare (because I do not have a large crowd consistently trading behind me). 6 Why Symbols Differ in CFDs CFDs use alternate symbols like US30 instead of DJI because they are synthetic contracts for difference. They mimic but are not identical to real futures or indices due to legal and pricing model differences. The nature of FX and CFDs (for those interested) is explained towards the end of the material. 7 Important Warnings About Public Trading Strate≠gies Many ìgurusî show ìprofitableî strategies without factoring in market impact or real world fill problems. Assuming a strategy works without testing how execution holds up at scale is risky. If a strategy gets popular and a lot of people use it, it will lose edge because of alpha decay. Educators often skip over liquidity, slippage, and broker mechanics, giving a false sense of simplicity. They also conveniently skip over that no prop firm, regardless of whether retail, scouting, or professional with a base salary, allows copy trading activity. It is not allowed, another net negative. To share oneís edge reduces personal P&L potential. 8 Summary and Final Thoughts Edge decay from crowding can be a reason a retail systems fail. (Turtle trading strategy returns are less and less impressive the more time elapses.) Forex and CFD retail trading have unique liquidity and execution challenges due to synthetic books and limited broker inventory. Market makers play a key role but can cause price differences and fills that do not match expectations. Big trade sizes expose these problems clearly. Smaller traders often do not notice. Be sceptical of public ìprofitableî systems without understanding market microstructure and real fill conditions. Managing risk, liquidity, and execution takes knowing how brokers and markets work, some≠times even using multiple venues (if you are trading FX). If you want to trade seriously, grasping these realities is crucial to protect your edge and avoid common mistakes. 9 Considerations for Liquid Markets Such as Futures If you trade big size solo, like 100 contracts in futures, you might cause some noise but will not really move the price during active hours. But if, for example, 200 traders each trade 5 contracts at the same time using the same system, that adds up to 1000 potential contracts which could influence price or HFTs, especially outside NYSE hours like London session and after-hours. This concentrated pressure can cause slippage or bad fills when scalping or day trading, which would eat at the educatorís edge (market crowding). Even a couple of ticks of adverse price movement can wreck scalping or day trading strategies. Bottom line: Although uncommon, sharing an actual profitable system risks losing or de≠stroying your edge even on liquid markets because the combined trading has potential to very briefly influence the market at consistent levels. Our sole focus is on the potential net negative for strategy sharing. 10 Algorithms and Market Response Market Maker Algorithms are not out to get you. They are automatic programs working to make the market more efficient. When lots of orders cluster at predictable prices, these algos notice and adjust, often by pulling liquidity or moving prices to avoid risk. So, when you feel ìhuntedî by íMMsí, it is random and just the market reacting to too much concentration or imbalances. 11 Main Point: Over-The-Counter Markets FX and CFD traders can start getting worse fills on live trades (relative to the prices they want) if they share and their strategy becomes popular on their broker. 12 Main Points: Centralised Markets The consequences for market impact are often sequential; something brief can influence a lot of future dealings, for example, traders could cancel orders in response which influences other participants and so on. That is why serious traders do not share trades in real time. Not because they are hiding some trading cabal secrets, but because the moment you turn a trade into a crowded event, order fills become less favourable and trading becomes more tactical. A final point: So if licensed professionals suffer from it and actively avoid it, ask yourself: why would a sane retail trader willingly create the same problem by broadcasting entries to thousands of people? Do not rely on signals or public strategies. You must create your own strategies. The edge is in trading in a way that gives you a unique execution pattern that aligns with genuine market behaviour for enhanced probability of success. 13 Additional Reading and Context ï Julien Penasse ñ Understanding Alpha Decay ï On the Effect of Alpha Decay and Transaction Costs on the Multi period Optimal Trading Strategy ï High Frequency Market Making: The Role of Speed ñ Yacine A®it-Sahalia, Mehmet Sa.glam ï The Role of Financial Instruments in Reducing Exchange Rate Risk ñ Vlora Berisha, Rrustem Asllanaj ï Also, note that: Total Return Swaps (TRS) and Contracts for Difference (CFDs) are similar in that both allow you to gain exposure to an assetís price changes or perfor≠mance without owning it outright. You benefit from price changes and, depending on the contract and type, even receive or pay income like dividends or interest. Both involve paying financing costs if you hold positions overnight (swap fees). ï IG Index (example of a regulated CFD broker) CFD customer agreement key parts: 12.8b, 21.1, and so on. https://www.ig.com/uk/customer-agreement ï Turtle Trading Edge and Alpha Decay Turtle trading story and rules Note: Turtle strategyís returns got diluted after media exposure or retail adoption and worsened after structural changes because of electronic trading.

Strategy Engineering Volume 3: Logic Optimisation

Logic Optimisation & Intraday Trade Timing Strategy Engineering Volume 3 Ron -Sentient Trading Society Introduction Key concepts covered in this volume include: ï How to choose intraday time ranges and filters using research and microstructure logic instead of overfitting to price data. ï Useful market microstructure concepts and vocabulary. ï Why and how to use variance to your advantage through multiple uncorrelated strategies. ï The two main STS strategy logical foundations. ï Why a real edge needs an underlying cause rather than price action in isolation, and how to think in terms of drift. ï Why you should define the job of a system before testing. ï Why îinstitutional order flowî is a marketing distraction, what it is used for in banking and what retail cannot access. ï How to think critically about educators and public market narratives, and how to spot and resist poor information, conflicts of interest, and unfalsifiable claims. Contents 1 The Statistical Reasons Why Retail Traders Lose 3 1.1 Behaviour matters, ícandlesí are for favourable entries to ride the drift up or down. Nothing else. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2 STS Strategy Types 1 3 Using Variance To Your Advantage To Succeed 9 3.1 How doyou know your systemsareuncorrelated? . . . . . . . . . . . . . . . 13 3.2 If you are choosing to trade intraday, attempt to identify broad time ranges where your specific strategyís logic actually has edge and why through research insteadofrelying solelyonprice data. . . . . . . . . . . . . . . . . . . . . . 14 4 Retail Educators and Public Market Opinions 15 4.1 WhatitTakes To Win .............................. 15 4.2 Artificial Intelligence and Large Language Model Usage . . . . . . . . . . . . 16 4.3 Critical Thinking and Third-Party Information Review . . . . . . . . . . . . 17 4.3.1 Information Rejection .......................... 17 4.3.2 One-Size-Fits-All and Conflict of Interests: . . . . . . . . . . . . . . . 18 4.3.3 Information Acceptance ......................... 19 5 Sentient Auction Framework and STS Market Microstructure Principles 19 5.1 BasicDefinitions ................................. 19 5.2 Addressing the îinstitutionalî order flow distraction . . . . . . . . . . . . . . 22 5.3 Three Wicksí Market Microstructure Context. . . . . . . . . . . . . . . . . . 24 5.3.1 TheTruth About OrderFlow Tools ................... 25 6 Dealing with Market Microstructure The Sentient Way 27 7 Understanding Order Flow Mechanics and Dynamics 28 7.1 OrderFlow Mechanics (OFM).......................... 29 7.2 OrderFlow Dynamics(OFD) .......................... 29 7.3 Liquidity andMarket Efficiency ......................... 29 7.4 TheSTSInterpretation ofMarket Efficiency . . . . . . . . . . . . . . . . . . 31 7.5 GeneralMarket InefficienciesandOHLCdata . . . . . . . . . . . . . . . . . 34 7.5.1 Why usecandlesticks? .......................... 34 7.5.2 DefiningOHLCMarket Inefficiencies . . . . . . . . . . . . . . . . . . 35 7.5.3 DefiningMicroscopic Inefficiencies.................... 35 7.6 Avoiding Ultra Low Timeframes (Technical) -Ali . . . . . . . . . . . . . . . 37 8 STS Entry Technique Fundamentals and Order Flow 8.1 Whatmarket efficiencylooks like ........................ 38 8.2 OurProcess: .................................... 41 8.3 Aheuristicto simplify:.............................. 43 8.4 The STS Time Series Inefficiencies and Market Profiling . . . . . . . . . . . 44 9 Backtesting Quality 48 9.1 Good Backtesting Hygiene Must Be Prioritised . . . . . . . . . . . . . . . . 48 9.2 You Cannot Rely on a Single Strategy for Long-term Success . . . . . . . . . 49 9.3 Intraday Trading SessionOptimisation . . . . . . . . . . . . . . . . . . . . . 49 10 Why Some Losing Strategies Still Look Like Winners and Variance in the Distribution of Returns 53 10.1 Simulating aLosingStrategyísEquity Curve . . . . . . . . . . . . . . . . . . 53 10.2 Simulating a Breakeven Strategyís Equity Curve . . . . . . . . . . . . . . . . 54 10.3Whatatrueedgelooks like: +0.4EVSimulation ............... 55 10.4WhatOverfitting Looks Like ........................... 56 10.4.1 Example: Bull market long-only indicators bait . . . . . . . . . . . . 56 10.5 Market Cycle Theories and Traditional Auction Market Theoryaredated ................................. 58 11 The STS Glossary 59 12 STSí Market Principles 61 1 The Statistical Reasons Why Retail Traders Lose Before we get into the details, weíll lay out a simple lens for reading what follows, what tends to work, what tends to fail, and why. 1. Baseless Execution Patterns: No accounting for microstructure or genuine market mechanics, e.g., the continuous auction aspect of financial markets, trading sessions, microstructure events, e.g., session events. Decision noise from intuition-led decision-making. 2. Using Variance Incorrectly: Inconsistent execution patterns randomising P&L outcomes. Relying on one strategy instead of multiple when not established. Using multiple uncorrelated systems on different accounts increases your chance of success statistically. This is explained on Section 3. 3. Statistical Laziness. For example, thinking 3 months of data is enough. 4. Poor Trading Cost Assumptions: Wishful assumptions, minimal first-party testing, review or research. 5. Short Term Price Discovery Consequences: Poor autocorrelation between bars in the short term. (near 0 for 1 minute charts) In plain terms: if the last candle was up, that does not reliably make the next candle more likely to be up (and vice versa). The ìsignî of returns is close to random a lot of the time. This makes îprice actionî useless as an edge when isolated in day trading. If the bars are largely unrelated to each other, they cannot be used to forecast future results accurately on their own. Edges are possible on lower timeframes, but you need an underlying reason for why the price would go up that is unrelated to the price itself, underpinned by established reasons that are unlikely to disappear. Each component provides a small edge that creates a large edge (profitable strategies). Key word: Drift In plain terms. ídriftí when price is skewed to go up or down. It is the average direction the market is being pushed over time. The core of our style is identifying specific behaviours (common, generic tendencies) or îregime-shiftsî induced by fundamental changes (niche and rarer) to ride in advance (prefer≠ably) before testing, with the expectation for either to persist justified with objective evi≠dence. The test is whether you can profit from the market adhering to that behaviour or regime over time with your entry technique and risk management tactics. If the data validates it, we are clear to run. At the point of entry, we position ourselves at the point of absorption and with the position we want to ride the drift the condition, session or behaviour provides. If it doesnít go all the way to the top of the auction (auction tails or mechanically defined swing highs and lows), we have mechanical references such as clusters to aim for first. There are endless possibilities, and microstructural inefficiencies that are indirectly ex≠ploitable by retail traders are much easier to find with refined search engines, public research journals and LLMs (all research must be triple-checked for validity). 1.1 Behaviour matters, ícandlesí are for favourable entries to ride the drift up or down. Nothing else. Edges live in temporary market states rather than calendar years. Remember, Logic first, Why before what. 1. An Established Session Behaviour: E.g., îDuring this trading session or time range, this market tends to trend so I want to position myself to ride that drift up or down, as price is likely to persistî E.g., îDuring this trading session or time range e.g., (11:30 AM and 2:00 PM ET) this market tends to range/mean revert so I want to buy it if itís stretched to the downside and I want to short if the price is íoverextendedí with a pre-defined consis≠tent entry technique. The trend is likely to turn around. E.g., On this trending market, this market tends to mean revert during these hours, so I want to create a reversal system to position myself to get an entry and ride the potential post-range trend. (A hybrid example) 2. An Established Market Regime/Behaviour. This market mean reverts intra-day, this market trends intra-day or the S&P 500 has positive drift over long timescales (simple examples). 3. An Emerging Market Regime/Behaviour: This market has changed character or lost stability because of an economic policy (e.g., Quantitative easing efforts, Quantitative tightening, Interest rate decisions), company changes, or supply shock(s), and it should persist for a while. For example, a relatively stable market is experiencing whipsaws or trends. The point is the market in question has changed how it discovers new value for a reason that will stick, and we position ourselves to benefit by anticipating its continuation. A medium to long-term example: Gold accumulation cycles during geopolitical tension, e.g., Russia vs Ukraine 2022 esca≠lations and asset restructuring efforts, e.g., People moving to gold after asset seizures. Potential Actions: (a) Long only strategy development. (b) Aggressive longs + predefined scaling in tactics. (c) Strict shorting criteria on long/short systems. A shorter-term example: US Market Volatility Shocks e.g., 5th February 2018 In this example, the VIX jumped over 100%, which meant US equities had shifted from a îlow-volatility regimeî into a îmedium-volatilityî regime (VIX reading >15) followed by additional volatility. This naturally makes trending more likely even as the volatility normalises. As a result, markets that typically mean-revert or range intra-day, and are heavily influenced by US equities, will exhibit more trending due to the volatility expansion. Potential Actions: (a) Be more open to intraday trend-following strategy development for mean reverting markets influenced directly by US volatility such as Apple stock, YM Futures (Dow Jones) or Eurex Futures (Euro Stoxx 50). (b) Develop strategies that take advantage of these persistent trends where they al≠ready exist with scaling in or manual trailing stops in markets that trend intraday, such as NQ Futures (Nasdaq) or GC Futures (GC). (c) Develop reversal strategies in mean reverting markets which have manual me≠chanical trailing stops at price structures to capture overextension whilst exiting where the market is liquid. 4. A fundamental reason (Advanced): A favourite example for swing and position trading. Citrus Greening Disease in Orange Juice (2020+); an obvious non-hindsight threat to supply which was likely to provide positive drift even if it normalised. Potential Actions: (a) Long only positioning (b) Aggressive long only strategy development (Including scaling in) Why we can do this? The only reason we can casually engage with the market like this and win is because we do not move the price. We can trade like this position to position with thousands or millions per strategy, but not billions. Adapting with higher trading sizes: When we trade order-book-influencing size, it is possible to conceal our intent by spreading our orders across multiple venues, using over-the-counter or non-exchange access to market exposure. This can be a tactful way to increase scalability per strategy, but only up to a certain point. Personally, my largest position was equivalent to 106 YM contracts short via a CFD (written in 2025 for context), which is high enough to briefly influence short-term price discovery on the central exchange (DMA). When we choose to trade in the short term, as soon as we start to influence price, algorithms notice, and our edge becomes vulnerable and is often faded. Institutions have this problem; we do not. 2 STS Strategy Types We design two strategy types: 1. Class 1 is behaviour anchored in structure. 2. Class 2 is behaviour driven by regime change. In the documents, there are multiple examples of the type of things that could be modelled. For instance, the most basic ones are anchors, like why a market tends to range intraday versus trending. Most of the time, what it tends to do more is range intraday. We will refer to this as system class 1 (common) Research is used to verify if it is just a pattern or based on something real. In the documents, I have outlined why, for example, the primary instrument we choose to trade, the Dow Jones, actually ranges. In markets, generalisations without a mechanism supporting them are descriptive, not predictive. Align your setup with the assumption that the market is going to continue this behaviour when discovering new prices. You create a setup, for example, an O H L C chart-based entry, with the aim to get filled at a price that is cheaper after a signal has completed. This gives you: 1. A price advantage at the point of entry, and a cost advantage. 2. Using limit orders will get you filled at the price requested, or better, in most instances, 98%+ depending on size. You will rarely not get filled because your size is too small as retail. There are also tactics I use for CFDs to get good fills. We use a filter to support the setup, we refine it, and then we refine it, and then we earn our edge after rigorous design and testing. Figure 1: Simplified Visuals for STS Deductive Reasoning STS System Class 1 System class 2 (niche, rarer but very profitable when done properly) The other way is to use the fundamental changes within the market, e.g., supply chain changes or product restructuring, that will persist for months. We have provided multiple examples (more than three) within the materials (Strategy Engineering Volume 2) and how to approach this objectively. For example, gold and other commodities especially, will trend for months for a reason. If it is rooted in something fundamental and real, then even an instrument that tends to mean revert, for example oil, can behave differently for a period. You have grounds to deviate with a trend-following system. View system class 2 as a side-arm and system class 1 as our primary weapon. Figure 2: Simplified Visuals for STS Deductive Reasoning STS System Class 2 3 Using Variance To Your Advantage To Succeed As mentioned in Strategy Design Volume 1: Do not enter the arena with just one strategy; When seeking establishment we must have multiple logically tested systems that are uncor≠related to trade on separate accounts or prop firm evaluations to increase our probabilities of capital accumulation. Have at least three well-tested trading strategies in place before you start. Run each one separately so that one strategy cannot adversely affect others. Not all trading systems need to be short-term strategies; all of them could be swing trading strategies (multiple days average holding time) it is up-to the strategy designerís constraints and goals. An average retail traderís opinion: îI only need one trading strategy.î Reality: This is common industry practice; variance is leveraged, and it is what we did with prop firms to get in the door. Yes, it is not required, but you only reduce your chances of success if we do not take advantage of what variance provides (split paths). Success relies on a large sample of positions that perform well on average (Positive R). Institutional incentives for diversification that we do not have to follow: 1. Market impact reduction (market impact costs money) 2. Lower return volatility (less extreme returns in favour of stability to satisfy investors). 3. Adhering to risk limits to satisfy investors (we set the risk limits and standard trading accounts and prop firms allow high position sizing) What this means for us 1. We have more flexibility in the strategies we run, and we have to worry less about order fill probability and slippage. 2. As smaller participants, we can take disciplined, asymmetrical positions, e.g., 1:3 RRR, that optimise our risk-reward profile in ways larger institutions cannot, without com≠promising our overall capital because of our capital partitioning and risk isolation techniques. Institutions generate extremely efficient returns, but they lose flexibility due to their strict risk boundaries and liquidity constraints related to their trading size. 3. The Main takeaway: The flexibility I describe enables us to structure strategies with higher-upside potential paired with pre-defined, controlled risk. This is a balance that traditional investors are generally unwilling to allow, but that our accounts or retail prop firms will permit. How we see it: We do not aim to compete directly against institutions; we are simply travelling across the waters, taking different constraints into account. Institutions operate more like a ship built to cover long distance safely and steadily: they are constrained, slower to change course but more stable. Lower cost percentages and a lower chance of failure, but with more expensive foundations and lower but efficient returns. We operate more like a speedboat: we are agile, able to change direction quickly and exploit shorter windows of opportunity, but we are more exposed to waves (variance) which can throttle us if we depend on only one route. Each strategy has its own path. Figure 3: Three boats in harsh waters, two out of three optimised routes succeeded This is why we split paths. Multiple uncorrelated systems on separate accounts is our way of increasing the probability of reaching the destination (profitability) without natural streaks of rough conditions forcing the entire journey off course. We do not try to directly compete. It is a matter of engineering differences. Remember, it is our boat; we have the flexibility to steer our own course. Once a trader is financially established, they can scale back the number of systems they choose to run or shift their priorities in response to changing constraints as needed; larger vessels are more stable. Our style can evolve over time as well; but it is up to us to satisfy ourselves instead of risk-averse investors. Institutional structures should not be copied blindly. Diversification should not be done for the sake of it. It is important to use variance for growth, if at all. To gain any potential benefit (a) Each system has proven positive expectancy (b) Correlations are genuinely low in testing (c) Total risk is still capped Without these conditions, variance accelerates failure just as efficiently as growth. Figure 4 shows that a strategy as good as 1:2 with a 50% win rate after costs (a very good system) can still lose money over time (unlikely but possible). Each strategy had equally profitable stats, but regardless, multiple paths ended in a minor profit or even a loss. It is best to think of each strategy as an individual path with the potential for success. Trading success is path-dependent. Figure 4: Here is a reproducible equity curve simulation of 100 uncorrelated trading strategies 1:2 RRR 50% win rate strategies with 5% risk, one strategy per account. Key points to take away from this section: 1. Developing and running multiple unrelated systems increase your chances of trading suc≠cess. Trading success is path dependent; it is natural. It only takes one to play out as expected, or better, to establish yourself as a trader. 2. Having 3+ good, uncorrelated systems across 3 separate accounts is like having 3+ un≠related profitable businesses in different areas that you can scale. Split, but concentrate your risk(s) safely. 3. Running one strategy that has a genuine edge at a time, for the sake of it, only reduces your statistical probability of measurable success in any field, trading included. You only really have the flexibility to run a single strategy at a time if you can comfortably absorb a long period of underperformance (even years) without it putting you under financial pressure. With the importance of variance and uncorrelated strategies established, we can now move on to how they are classified and separated. 3.1 How do you know your systems are uncorrelated? 1. Different Asset Classes Index Futures Stocks (Individual Equities) FX (Currencies) Metals Energies Commodities 2. Distinct Behaviours Exploited By Different Strategies Fading short-term trends as the core approach in markets that tend to mean-revert intraday. Following or anticipating trends on markets that trend intra-day. Market Opens or Post-News Effects, e.g., fading over-extensions. Trading Around Predictable Liquidity Events Uniquely. (a) Different Session Times (b) Different Timeframes (c) E.g., Session overlap volatility expansions, e.g., Asia, Frankfurt and London over≠lap. (d) E.g., London and New York Session îOverlapî (e) E.g., Predictable times of price spikes and mean reversions that follow (surround≠ing market opens). 3. Different Risk Management Style(s) Single position at a time Scaling in Manual Trailing Stops Hybrid 4. Different Strategy Dependencies Examples: (a) Different derivatives can be used for execution (for example, spot FX vs back-month FX futures or FX forwards). (b) Nuances between completely different instruments e.g., Stocks vs Options (c) Depends on a different market condition or regime. Each derivative stated in 4a provides exposure to the same currencies, but pricing models and cost structures differ. For example, with back-month futures or an FX forward, a swing trader can take advantage of forward curves. With regular spot FX, this is not possible because there is no quarterly rollover. 5. Decide the job of the system in advance Remember that îTrendingî or îMean reversionî are common characteristics, they are not permanent market features. Your system will not produce amazing returns every quarter or every year. Prioritise logic and context before data, do not let test results own you. Run multiple edges that take advantage of different behaviours. If your trading frequency is too high, add a filter that complements the underlying logic of your idea. The time when we can deviate and create systems that go against the marketís status quo is when fundamental changes justify it as a mechanism; examples are provided within the material. These are the more advanced strategies, which I classify as sec≠ondary (System class 2). End Note: Remember that a strategy only needs one or two key differences to follow a completely dif≠ferent path. Creating entirely new models is secondary to ensuring adequate non-correlation between systems; avoid deploying minor variations of the same strategy, as this will introduce noise into the returns instead of the non-correlation we seek. 3.2 If you are choosing to trade intraday, attempt to identify broad time ranges where your specific strategyís logic actually has edge and why through research instead of relying solely on price data. A common trap is when traders try to treat their strategy as a permanent personality trait of the market. The edge(s) you see only exist within specific market conditions. This is why out-of-sample testing is important for gauging true strategy robustness, both in favourable conditions that are unfamiliar and in adverse conditions. When conditions change, traders typically blame their setup and start tweaking parameters until the backtest looks smooth again (overfitting). At that point, we are not improving the system, we are just fitting it to a chunk of history that never repeats 1:1, which ruins real-time results. If there is a single, simple filter (for example, a price structure, indicator, or foundational adjustment) backed by a logical reason that can improve performance, use it. For instance, if trading during a particular hour is statistically unfavourable because price is likely to move sharply due to a specific microstructure event, such as a market open rather than data, cut that hour. Changing the times you trade a system often makes a noticeable difference without overfitting. Do not try to perfect your hours or fit them to the data. Instead, test the hours that fit your constraints and your strategyís idea. The earlier they are defined, the better, preferably chosen before testing To avoid key issues, never ask ìwhy did my entry stop working?î in isolation. Ask ìWas a fundamental adjustment responsible for its collapse, is it realistic for me to benefit from this in typical intraday conditions?î. If the edge only shows up in certain types of days e.g., news days, make the system selective, collect more data, and let the strategy sit out the rest. Discard any strategy that requires adjustments exceeding 20%. Modifying parameters to force universal performance across varied market conditions directly degrades the strategyís integrity. Specific instrument(s), specific times, rules, timeframe(s). What to take away from this: A traderís technical analysis can be íperfectí on paper, but it can only produce positive returns when the market conditions support the behaviour(s) your model relies on. The ìedgeî is not solely in the candlestick formation itself, it is in the conditions that make it statistically meaningful; that is a part of our logic. If you are running a trend-based trading strategy in a ranging market, the edge usually breaks down, and even a small positive expectancy like 0.1R per trade can get wiped entirely out by costs, slippage, and variance. We outline the scientific basis for using candlesticks throughout the document, starting in Section 7.5 and continuing beyond. Guidelines on picking other tools are provided at the end of Execution and Venue Mechanics. We know that no single profitable strategy lasts forever, and that is normal. Every industry goes through change, so there is no need to feel discouraged or intimidated by it. Let us take this step by step and focus on what you can control: how to accept quality information and reject poor information. 4 Retail Educators and Public Market Opinions In this section, we will be helping you sharpen your critical thinking skills, which are essential for adaptability in your trading career. It is not enough to suggest what to do, so we will also be discussing in thorough detail what to avoid doing to prevent slip-ups. 4.1 What it Takes To Win Rigour. As a trader, you do not need to be sharp between the ears, but we need to be determined and sharp with our decision-making. Success in trading is path dependent. You need a sequence of mostly high-quality decisions to succeed, for example, running a 0.3R expected value strategy verified by data would be a higher-quality decision compared to running a strategy not verified by data or one that relies on intuition (not that this is the optimal metric, it is just an example). It is all about rigour Many traders say that ìit is all risk managementî, ìit is all trading psychologyî, or they say ìyou only need an edgeî to succeed. This is nonsense. Sometimes it is wise to step back and wonder why most traders say the same thing and why most traders are unsuccessful. The material condenses several years of work and observation into a shorter format so the same costly mistakes can be avoided. Psychology and mindset (to perform) . Edge (to win) . Risk and capital management (to survive): all of these elements must be in check. You need to be serious. It is not a casual game if you want sustained profitability. Having a lucky profitable run is not the same as having multiple years of profitability. 4.2 Artificial Intelligence and Large Language Model Usage To start off, we want to be extremely clear on this to avoid being deceived: do not use AI to outsource your thinking or decisions, use it as a research tool. AI is trained to be agreeable and often outputs rehashed retail opinions and distractions. If you are going to use AI, only use it to help with specific definitions. For example, if you do not understand what îLimit ordersî are, you can have an in-depth back-to-back discussion to fast-track your learning. Besides that and speeding up research efforts, AI is not 100% reliable for raw information. -Stated in Q3 2025 for context. AI is also useful for vibe coding tools; if you cannot program efficiently, they are great for creating indicators or other assists. Canít code an indicator? Just submit a loaded prompt describing the formation and desires with a lightly annotated visual. In a couple of minutes of pressing and debugging, it is done. The Reality Surrounding AI AI is not sentient; it does not think. It processes. AI cannot perform basic nuanced tasks such as citing well-known papers without error, paraphrasing, or mislabelling. If AI gets to the point where it has real intelligence, it would still only be as good as the data it is trained on. Modern AI has been exposed to a lot of retail education and techniques. AI can push flawed ideas and frame them as decent, indirectly encouraging poor strategy logic. AI is easily manipulated; Users do not need to manipulate AI. LLMs are skewed to agree with the user, as they are trained to. In trading, this is dangerous, as AI can reinforce your mistakes. A lot of todayís winners will get wiped out tomorrow. Publicly available AI is faster than you but not smarter (yet). Use it to speed up tasks towards the solution. If you rely on AI for a solution, you will end up with a saturated or overfitted system most of the time. 4.3 Critical Thinking and Third-Party Information Review 4.3.1 Information Rejection To know what is good information, it is best to know what bad information looks and sounds like. Do not fall for these narratives. They can hold your progress back for up to years. Red flags to look out for: Dismissive reactions instead of logical rebuttals when you question the narrative. Posturing with experience, gains, or materials to make a point (appeal to authority fallacy, as no evi≠dence is given). If an educator feels the need to posture in order to appear as an authority, it is almost always deceptive, in bad faith, or driven by feelings of inadequacy. Nine times out of ten, when you ask for trading statements or an unedited trading platform video from a regulated brokerís PC platform (not MT4 mobile screenshots, but actual proof), you will see what they are really about. The classic responses like ìI do not need to show you anythingî, ìWhy would I show you that?î and ìI do not need to prove myself to anyoneî, or my favourite, ìI am not selling anythingî are nonsense excuses. It takes a maximum of two minutes to provide evidence if you are successful. If anything needs to be censored, such as names or account numbers, they can easily be cropped out. Instead, gurus would rather waste five minutes distracting and manipulating rather than simply presenting the evidence on a regulated platform from a computer. If they use intuition or discretion in trading, their success cannot be replicated unless their strategy can be sequenced mechanically, making their trading subjective and non-reproducible. In this case, it is best to focus on the basics they showcase and filter out the nonsense. Rigour or A Lucky Art Salesman? How did this educator succeed? Was it sharpness or lucky drawings? Most educators are teaching people how to depend on a single trading strategy instead of how to survive and grow as a trader. This is about showing the process. How did they arrive at the profits they showcased? 1. How do they manage risk? 2. How do they collect data? 3. Risk management techniques? 4. Do they have decision-making models? 5. Is their strategy intuitive and therefore non-reproducible? A ìprofessionalî trader who cannot show any rigour, who does not know their costs, etc., should raise immediate red flags. Years ago, we used to direct message or call gurus and ask them, îWhat are your trad≠ing costs percentages?î and they did not know, or issued vague answers, because they do not trade live. 4.3.2 One-Size-Fits-All and Conflict of Interests: A good educator should absolutely encourage you to adapt your OWN strategies to your own strengths, risk tolerance, and goals. If they push you toward a single trading strategy (limited long-term usefulness) or a specific tool, then it is a conflict of interest; they are more interested in selling you a product than in teaching you how to adapt to the markets, creating dependence instead of adaptable, sustained success. No single approach works in every market condition, and flexibility is a core component of successful trading. There is also more than one way to beat the market as long as the setups are grounded in real logic. We are not here to force your hand, we are here to guide it. Be cautious of anyone who issues absolutes. For example: Anyone who says trend following does not work is wrong (many traders say this), Anyone who says mean reversion or reversal trading does not work is wrong (many traders also say this), Why? Both can work on any market with different time horizons/trading hours, that is the truth. People who spew dogma regarding market navigation are posturing. Non-absolute version: Each market and timeframe is different, when you start to hear e.g., trend following works better on X,Y,Z markets you can lean in and listen but even then, you must verify with research done by yourself, including the part most skip. Checking if their observation is a coincidence or something that will persist because of existing market structures and dynamics that will stay. This is the type of thinking that matters; objective binary answers rarely exist in financial markets. If in 5 years you still need us to tell you how to think about financial markets etc, I have failed. The goal, is to build enough skill that you no longer need the Sentient Trading Society. 4.3.3 Information Acceptance Just as it is extremely important to recognise and reject the bad information, it is just as essential to understand how to identify and accept good information. Trading success requires a mindset capable of ruthlessly distinguishing signal (value) from noise (nonsense). Good information in trading, or any other field, has the characteristics cleared up. It can be verified, replicated, and explained logically, e.g., with peer-reviewed papers, without relying on any personality, stories, anecdotes, or emotion. When someone presents data, statements, or results, they should be transparent about their process and open to reasonable questioning. A genuine educator will not feel threatened by your curiosity; they will welcome it be≠cause it shows that you are thinking critically. Gurus do not want critical thinking because people will question and resist the narrative. Talented traders use market microstructure principles and data instead of anecdotes, to explain their reasoning. 5 Sentient Auction Framework and STS Market Mi≠crostructure Principles We have defined this term to describe the act of combining market microstructure knowledge and unfinished auctions from AMT to candlesticks rigorously. It is named this to help with your learning. This is what logic is to us. Unfiltered. 5.1 Basic Definitions In-depth explanations covering key nuances have been provided. Price Delta: The net change in price over a specified time, e.g., a bar closed 10 dollars higher = +10 delta. 10 dollars lower -10 delta. Volume Delta: The net difference between the volume of buyers and sell volume. Time Series: Data plotted over time. For us, that would be candlestick charts. Liquidity: How easily an asset can be bought or sold without moving the market price. In markets, it also means buy and sell orders. Poor liquidity means difficulty in executing trades. Liquidity implies active buying, selling, or both. Liquidity Provision: Adding orders (usually limit orders) to a financial market, providing liquidity, making it easier for others to trade. Price Discovery: The ongoing process where supply and demand interactions determine the market price of a market/asset. Order Fill: A position being executed in the market. Market Order: A request made by a market participant to buy or sell at the current or next available price. Limit Order: A request made by a market participant to buy at a specified price or better. Market Maker (MM): A participant who continuously quotes buy and sell prices, earning the spread and providing liquidity. When you go long or short, you are buying from or selling into another market partici≠pantís order, taking their liquidity. To be clear, although we tactically provide and remove liquidity, we are not market making. We engage in passive order placement and removal. Market Taker (MT): A participant who uses market orders, consuming liquidity by trad≠ing at available prices. We take liquidity from the market when positions get stopped out. Inventory Risk: Inventory risk refers to the potential risk market participants have, e.g., traders or market makers, due to holding an îinventoryî of assets, e.g., units/contracts long or short on an instrument. The risk is from the price fluctuations of the assets held, which could reduce the value of their îinventoryî For example, a market maker can hold a large amount of a single asset; the price decreases, and they could realise losses on their position. This is called an îimbalanced bookî Buy-side Participants: Traders, investors, practitioners and money managers, producers/consumers, and people hedging against market changes. Sell-side Participants: market makers, dealers, HFTs, and investment banks operate on the îsell-sideî. Adverse Selection: Adverse selection is where one side of the trade has superior informa≠tion to the other regarding the market traded, leading to an imbalance in the transaction. During adverse selection, these traders enter the market, exploiting that imbalance in in≠formation, leading to unfavourable outcomes for other market participants, such as market makers, causing them to reduce or remove their liquidity entirely and increasing the bid-ask spread that buyside participants have to pay. Market Crowding: Market crowds emerge from a concentrated amount of liquidity clus≠ters at price levels that are predictable to HFT algorithms, such as popular retail trading strategies (e.g., opening range breakouts, Darvas box breakouts, turtle trading strategies, traditional S/R and pivots). Market crowding erodes market edges, even small ones, because larger market participants will anticipate the crowdís order flow. When advantageous, they will use or induce these price points to get filled in or out of large positions if the levels are close by, e.g., just a couple of ticks from the current price. For instance, an institution might cause a brief spike of a few ticks to capture thousands of anticipated contracts, absorbing the market crowdís liquidity. Institutional Manipulation: Retail traders often believe market makers deliberately move prices several ticks to hit stop losses, but in reality, it is too financially risky for MMs or institutions to do this because of inventory risk and adverse selection. For most liquidity hunts in most high-liquidity mar≠kets, the inducements are a couple of ticksí distance at most. Hunts are institutions taking advantage of the price movements towards the stop loss driven by other market participants to get filled on a large order size with low market impact. Market makers take advantage of it very rarely to rebalance inventory risk to zero. Another factor is constant regulatory oversight through systems such as the consolidated audit trail in the USA. Firms face significant fines from regulators if they are suspected of preying on other market participants. Regulated exchanges do not allow this behaviour, as it undermines market integrity. If regulators fail to correct malicious practices, it can reduce buy-side participant confidence, reducing volume, liquidity, and exchange income. This forces institutions to rely on subtle inducements rather than overt manipulation if they participate in it at all. Whenever institutions overreach, six to nine figure fines are issued. For example, Brett Falloon and Flatiron Futures Traders LLC were fined hundreds of thou≠sands of dollars by the Commodity Futures Trading Commission (CFTC) for spoofing in the S&P 500 and Nasdaq futures markets in 2022, and other firms have been fined for inaccura≠cies in reporting. In the 2010s, Barclays faced a £284 million fine (approx. 375 million USD) for forex ma≠nipulation and a group of investment banks were fined over 1 billion USD as a group for a similar operation by the European Commission. Market manipulation simply is not worth the risk to institutions. Level 3 Data (Market-by-Order): Every single order and every change are presented in sequence, providing a high depth of information down to the minute details Post-processed L3 MBO data is the most detailed and premium form of order flow information available. L3 data shows which specific orders were matched or cancelled at each price level. Of course, the individual participants are not identifiable, but they can still be tracked, and their flow can be anticipated, especially in crowds. Anonymous IDs are used. L3 data allows you to see exactly where participants matched and when, providing a complete sequence of events that includes all amendments, partial trade fills, and limit order cancellations. L3 MBO data reveals all active market participants, their orders, and order sizes at each price level, allowing high visibility of market behaviour. Level 3 is a lot more direct compared to simpler solutions like Level 2, which are limited to generic order flow and market depth. Level 2, footprint charts, volume profile (POC), and other traditional public order flow tools do not show the contextual depth institutions require to maintain their edge. This information, with zero-millisecond delays combined with the freshest tick data, is a powerful tool for institutions (not retail) to map, predict, and anticipate order flow while also supporting quote-pulling strategies to mitigate adverse selection before everyone else. These operations contribute a lot to alpha decay and edge decay if your flow is predictable, as you can get picked off by algorithms that operate by the microsecond. This is why we say to create your own trading strategies. If you are trading like everyone else, you will either get unfavourable fills due to slippage (this is from algos buying just before you do) or increasing bid-ask volume, absorbing retail flow in a way that is temporarily disadvantageous for retail traders on the other side. How this looks on a chart: Price gaps up on a bar close, or price moves quickly as soon as you and everyone else are buying, causing slippage against their orders. Or your volume will be absorbed in ways that are unfavourable, nullifying the crowdís market impact. How this looks on a chart: If, during price discovery, the market maker predicts that an uninformed crowd of traders is likely to buy at the next 5-minute candle close, they could increase the sell limit order quotes to provide excessive amounts of liquidity. Other buy-side participants looking to go short, e.g., institutions, could also utilise this liquidity, turning what would be a noticeable upward movement into a wick high rejection or continuation down against the retail crowd buying. To us, depending on a TPO/market profile in real time is like going downstairs to check your phone to see if itís raining instead of looking out your bedroom window (the charts) to see whatís going on. Both give the correct answer, but the weather app provides other information that may be useful, like how it rained earlier today. It is a decent tool, but it just is not required to know if it is actually raining now; candles give us that instantaneous reading. 5.2 Addressing the îinstitutionalî order flow distraction When you see older footage of ítradersí at investment banks looking at îinstitutional order flowî using time and sales and market footprint charts, they are using it to fill large orders from large firms at good prices when asked, they do not actively speculate or predict price movements, institutions (especially on the sell-side) manage assets in ways that mitigate directional risk if they do not, they blow up. When people call it îinstitutional order flowî, it is for marketing purposes; anyone can access it. It just reorganises the order book and time and sales for you visually to see the cause behind movement. A traderís job at an investment bank desk is to: 1. Fill large client orders whilst ensuring good execution quality over time. 2. Monitor conditions to judge when the market can or cannot absorb the requested size. 3. Avoid signalling their clientís intentions to mitigate adverse selection risk. This is not îinstitutionalî. îinstitutionalî is what you as retail cannot get. These traders have additional insight into flow from the firm they are working for, potentially over sell-side firms. That is what makes it institutional, not the information and tools that anyone can get access to. So when an educator says îI use what institutions useî, they are posturing because, without this context, it appears far more serious than it actually is. It is valid but not required for smaller trading operations under 50m USD. As a practical rule of thumb, data available for under $10,000 per year is unlikely to qualify as genuinely institutional. A lot of institutional sell-side dealer activity is automated these days anyway. You do not need order flow tools as retail; you need to understand the fundamentals of price to understand the effect. Remember your goal: You want to isolate moments of temporary market inefficiency within a formation and then anticipate a return to the reference point. The other edge is from understanding how the market(s) you are trading behaves at the specific time(s). For example, on YM futures / Dow Jones / US30 it exhibits mean-reverting behaviour setups intraday, so I will look for setups that either benefit from the reversal to return to the average or overshoot repeatedly. You can also benefit from the short-term trends in the middle too. It is all about having fixed rules to take advantage. These behaviours are well established and documented. NQ exhibits strong trending behaviour, so if you can identify overextensions, this can also be exploited (harder to take advantage of with limits, so I prefer other markets, but it can still be done). Figure 5: STS Three Wicks 5.3 Three Wicksí Market Microstructure Context. In a stretched market, three wicks against the overextended trend show that aggressive orders are potentially getting faded multiple times, or that excess sell-limit liquidity is being provided by market makers over time, as seen in Figure 5ís case. The inefficiency lies between wick oneís high and wick twoís high. For a mechanical constant, we typically operate on wick oneís price. Three repeated upper wicks into the same price area in a stretched market show failed pushes and potential absorption, which can offer a favourable entry when combined with a proper filter. If the effect from the fade persists, the trend reverses. For trend trading, when longing, three-wick lows can also indicate that aggressive sell interest was repeatedly absorbed, either by traders or by market makers over time. As a trader using this approach, regardless of the formation used, we want to position ourselves to benefit from being a part of and ideally ahead of the queue when this setup presents itself in alignment with market behaviour. The protocol is either to fade the third interaction or wait for the completion and use it as confirmation to enter a different setup (confluence). If we are fading for superior queue priority, I place limits as soon as the setup is confirmed. I know my target area, stop, and size it up many minutes before placement, as the position is visibly marked, and the longer you wait to place the order, the less priority you have for execution, reducing the chance of your limit being filled, especially if your entry price is close to the current price. What are we doing here? You are positioning yourself as a passive liquidity provider at a price where the market has repeatedly shown absorption on the timeframe you are viewing, instead of chasing price. In microstructural terms we are making a market for aggressive, less informed participants (when we are right) between our limit in (entry) and limit out (target price) with a maximum risk threshold at our stop. In auction terms you are anticipating a failed auction with a sell limit (after candle 2) due to an overextended market combined with other rule(s). 5.3.1 The Truth About Order Flow Tools Do not get me wrong, order flow tools can be useful, especially if you are trading large institutional sizes or doing high-frequency trading. But for regular trading sizes, the effects you see on the chart, which come from solid underlying causes, are more important in the short term than the cause itself. For example, price moving 10 ticks after interacting with that low and breaking out of this marked high is more insightful than îMarket makers adjusted their limit orders in response to a large buy order, and people cancelled their sell limits at the swingî for a typical profitable trading strategy. It sounds and looks fancy, but you are still buying the pullback. Order flow could be used to fine-tune the entry inside the unfinished auction to get a slightly better price, which optimises the edge, but it wasnít needed to create an edge in the first place. So we do not use raw order flow tools; we have developed our own way to profile markets to design ways to benefit from the effect. Scalpers need order flow for accuracy because it is super competitive. That does not mean you should attempt it as the chances of failure are amplified. We do not condone scalping. Work smart, not hard. Time is the one asset you can never get back or file a return for. Figure 6: Source: BookmapÆ You must ask yourself whether it is worth putting all those extra hours in, using all those extra resources to increase your strategyís edge potentially by 0ñ5%. Is it worth all that resource commitment to turn a 1:2 RRR 50% win rate system into a 1:2 RRR 53% win rate system? If it is, and you have the time, then you can proceed, but I have been trading for a while and it is not a requirement for me yet. The same price effects can be seen on candlesticks, the only additional insight is the exact cause. Order flow would just be another thing that you have to look at and focus on in a fast moving market, and an additional angle that you have to test. It costs extra focus and adds fatigue in exchange for accuracy. To add, if you are new, order flow data and tools are not free, and financing the tools is not fun either. If you arenít running modest risk with 20k+ capital, the rolling costs for data and platforms erode the benefits. What would change things for me: If I were trading 500 contracts per trade (over 100m USD position value on margin (not account size) on average, perhaps I would be forced to use it. But I have traded over a fifth of that in a thinner market (CFDs) via order splitting, so I think I will be okay for a while. Order flow is not magic. Knowledge provides the edge from knowing the mechanics of the effects in the first place. Just like everyone can see the same highs and lows, everyone can see the same order flow, and as retail, one of the few guarantees I can issue is you will never get to see or act on it first. It is not easy or free money like it is implied in marketing. See order flow tools as something that makes your 1:2 RRR 50% win rate system into a 1:2 RRR 53% win rate system, with additional labour or inputs required to earn the return. It can take away from your edge too. See it as an indicator. Order flow isnít going to turn a terrible idea into a great idea, it can fine-tune an already excellent idea, it just isnít going to be the thing that makes the difference for 9/10 traders, and it typically leads to more confusion. 6 Dealing with Market Microstructure The Sentient Way As a trader putting this into practice, your goal is to make a market at favourable levels by tactically providing liquidity to enter and exit and by taking liquidity when conditions are unfavourable to get out. We aim to absorb/fade aggressive orders whether the market is DMA (e.g. futures or stocks) or OTC (e.g., CFDs or Swaps) 1. Superior entry prices compared to market orders 2. Superior order queuing versus when your entry is equal to the best bid/ask. The orders placed first get priority execution from the exchange. The trade-off is other market participants are aware of your potential intent to buy or sell earlier. For CFD markets. We get rewards either way. We position ourselves to benefit by 1. Designing strategies that get accurate, superior entry prices compared to market orders 2. Mitigating vulnerabilities to delays and liquidity provider discrepancies by using limit orders exclusively. 3. Scaling to size with order splitting techniques (Highest trade size ever: a 106 index futures contract size equivalent) 4. Get positive slippage from providing liquidity instead of absorbing negative slippage from taking liquidity from a synthetic book. 5. Operating with CFD firms that are regulated and show transparent market depth. We desire entries only where recent liquidity anomalies or inefficiencies are present, and want our profits to be taken where past inefficiencies are present. Limit in, limit out, and limit in, stop out for losers. We use research to know what a marketís behaviour should be, and we ride that drift by exploiting market logic for cheap entries. The market is supposed to be highly efficient, so we want to provide liquidity where it should be absorbed, and we want to get out at deciding points where liquidity will be concentrated, with both sides driving a breakout or anticipat≠ing a reversal; attractive places for price discovery. Inefficiency to inefficiency, liquidity wall to liquidity wall. This process has been presented carefully over several months to enhance accessibility and ease of use. 7 Understanding Order Flow Mechanics and Dynamics Positive delta (positive net buy volume), e.g., +1000 delta (buy volume and sell volume combined), means that there is a 1000-unit imbalance on the buy side (buyer dominance). Negative delta (positive net sell volume), e.g., -1000 delta (buy volume and sell volume combined), means that there is a 1000-unit imbalance on the sell side (seller dominance). Price Delta (What we care about the most): The net change in price over a specified time e.g, a bar closed 10 dollars higher = +10 delta. 10 dollars lower -10 delta. Important Nuance: Just because there is positive volume does not mean that the price will go up; it is the relation between the liquidity being offered to buyers (sell limit volume) compared to the buy market orders that moves the price up and vice versa for movements lower. Depth Of Market (DOM) Additional Sell Limit Volume Best Ask Mid Best Bid Additional Buy Limit Volume Price 10002 10001 10000 9999 9998 Available Volume 50 20 100 80 Table 1: Depth of Market snapshot In this example, if a trader buys 70 units, the dealing price (ask) moves 2 up ticks (last trade 10002 Ask) if there are no additional reactions, but the dealing price (bid) would not move a single tick if they sold 70 units; it would get absorbed on 9999. This imbalance in the liquidity offered can skew where prices go; there can be more units being sold, but the price still goes up. This phenomenon is often behind an îLow Volume Nodeî in volume profiling or îSingle Printî in market profiling (closer), for which the price tends to correct later. We revisit this later within this document. Hit pause for 10 minutes if you need it. Then dive back in. Every page counts. Assuming you are reading this, we are halfway through. If you feel your focus slipping, it is perfectly fine to set a 10-minute timer for a break before continuing, as we are about to probe into a new level of depth below. What market efficiency looks like Figure 7: A Market that is 100% efficient does not move. The needs of buyers and sellers are constantly met with zero deviations in supply and demand. Important note about the îStraight Line Figureî Since all information is already reflected in the price and no new information emerges, this example serves only as a heuristic, because such a market state is highly unlikely to persist. Remember, price movement itself can be efficient as long as it truly reflects information justifying the price. It may seem confusing at first, but it will begin to click over the next couple of figures. The takeaway is in a globally î100% efficientî market, there is no inefficiency to reprice. We will discuss local and global inefficiencies later on. Figure 8: In a highly efficient market, prices fluctuate randomly but correct themselves as new information is introduced. Over time, excess returns tend to average out to zero. Figure 9: Simulated Trending Conditions What is being simulated here? A low-volume grind-up on a lower timeframe. 8 STS Entry Technique Fundamentals and Order Flow 8.1 What market efficiency looks like Important note about the îStraight Line Figureî Since all information is already reflected in the price and no new information emerges, this example serves only as a heuristic, because such a market state is highly unlikely to persist. Remember, price movement itself can be efficient as long as it truly reflects information justifying the price. It may seem confusing at first, but it will begin to click over the next couple of figures. The takeaway is in a globally î100% efficientî market, there is no inefficiency to reprice. We will discuss local and global inefficiencies later on. Explaining Inefficiencies Global inefficiency: A structurally driven imbalance or mispricing that exists over a larger time horizon. For example, think of constraints like regulation, latency, capital limits or volatility shocks caused by news which persists over days. Feature: It is often persistent and often caused by natural limits or events that interfere with efficient price discovery. Efficiencies are sequential, unfolding over time and influencing other decisions, which gives them that îglobalî characteristic. Local inefficiency: A short-lived imbalance in price behaviour over a small window of time or space e.g., 3-5 bars. For example, think time series inefficiencies such as uneven price movements and one-sided trading (most common). Feature: It is temporary and caused by microstructural dynamics (the current buying and selling environment). Order flow: We do not use complex order flow, for example, Market By Order, as I believe deriving an edge from data this high resolution would realistically not be feasible for a standard human participant in real time (not in a backtest), so we apply what I know about it to time series charts instead (candlesticks). Focus should be isolated on changes regarding a marketís state and brief deviations from market efficiency instead of individual îcandlestickî shapes. When the market discovers new prices, not ífair valueí (subjective), but ínew valueí. What gives you power in understanding mechanics is that you know the causation and the effect it creates; for example, every wick is a tail in the auction. Order flow is about showing the cause (important for extreme short-term algorithms). Tools like candlesticks and TPO are about representing the effect which we use to structure our trading activity. Modern liquid markets are continuous auctions; an imbalance in supply/demand is created, leaving an inefficiency/unfinished auction behind which is typically followed by a correc≠tion/rebalancing of supply and demand later, after this correction. We look for an expansion towards a place where new value could be discovered. We instead opted to study the cause and effect for clarity when working with markets; order flow tools did not seem useful to me for larger price movements, especially because most limit order liquidity gets cancelled over time. Understanding the mechanism of price was essential; research provided this clarity. Source: YM Futures (Dow Jones) Examining the Auction the Sentient Way AMT stands for Auction Market Theory Classic AMT Narrative: The price was pushed up to ítestí how far up buyers are willing to pay. Missing Context: or where liquidity was provided in excess to buyers, e.g., by MMs or Traders Classic AMT Narrative: The price went down to ítestí how low sellers are willing to accept. Missing Context: or where liquidity was provided in excess to sellers, e.g., by MMs or Traders Classic AMT Narrative: The price had settled back to an area where both sides were roughly in agreement on ífair valueí. Missing Reality: It is not ífairí value; new value is discovered briefly (a change in price), and the market continues on to the new auction. Any íagreementí collapses on the next tick. What to take away from AMT for efficient trading (Valid): A complete auction is where both buy and sell demands are met, settling the auction. Typically resulting in a range or an îefficientî market. (The marketís aim) An unfinished auction is a deviation from market efficiency where one side of the marketís demands are not met, typically showing price runaways or îoverextensionsî from an area, which highly efficient price discovery tends to revisit later. A failed auction is a deviation from market efficiency where attempts are made to go beyond a reference (high, low) That breakout cannot attract follow-through. (Not enough volume or excess liquidity provision/fading) causing price to return to prior value. This is shown in Figure 15 for targeting and Figure 5 for entries. We do not use retail interpretations of AMT. Most versions you find online are misleading and not worth your time, but these listed concepts are worth retaining as foundations for your own ideas. 8.2 Our Process: Price Imbalance -> Market Inefficiency -> Stabilisation 1. Price Imbalance This represents a mechanically defined movement away from the current market value within the timeframe under observation. Price moves enough to surpass or deviate from its current path or structure. The way to see it is a one-way expansion over one or many bars How it typically presents: (a) A mechanically defined large movement (b) A sustained movement beyond a defined baseline (c) A movement quickly undone by an aggressive movement The market microstructure origins are thin liquidity, liquidation, reactions to news, or aggressive order flow (more market order volume relative to what is available over multiple prices). What it is not: îBuyers are taking controlî, îsellers are taking controlî. What it is: A change in the state of how liquidity is being provided to market order participants (market takers). This is not supposed to be the sole point of entry, as price movement in isolation is not the opportunity. How should this market behave? Should the market be ranging, should it be trending? This is one of the most important foundations for our entries, A financial marketís data feed alone, regardless of tools, software, or depth, is not enough to optimise the odds of success in speculation over many trades. We must research, even lightly, to understand it at the surface level, then make generalisations based on how and why it behaves as it does in order to develop specific, effective trading strategies. One can have a profitable period by accident, but we choose to be purposeful to increase our probabilities. This is what provides us with the privilege to achieve a positive P&L with simpler real-time data-sets such as OHLC bars. These questions will be answered by your research before strategy testing and deploy≠ment. This is embedded into your rules in advance. A lot of traders make the mistake that ígapí type entries in isolation have material predictive value. To make the differentiation click (analogy) Most traders see inefficiencies as a ìmagnetî with an unconditional pull, but the correct way to see it is like an electromagnet that only springs to life when the current (rea≠son) is strong enough to justify the pull (correction); otherwise, it is noise. Someone can push deactivated electromagnets together without a current, but there is no mag≠netic field. Similarly, the market can push price into the inefficient location without a consistent mechanism; one large random sell order shatters the fantasy. 2. Market Inefficiency Retail traders tend to translate îinefficiencyî into: Gaps must be filled Imbalances must be traded Or in plain terms: îPrice has to come back hereî Most people often treat îinefficiencyî as a guarantee instead of a market tendency. Inevitability is assumed and intuition fuels biases lowering decision hygiene. What inefficiency actually is: An îinefficiencyî in price refers to the structural changes that remain after an anoma≠lous price movement has occurred. Efficiency is where order is restored. This stage is where your strategy should have rules in place for interacting in advance to identify where price failed to adapt efficiently. 3. Stabilisation This is where we want to get filled. Stabilisation means price slows down or undoes the movement; it becomes briefly stable or price delta neutral. A local time series inefficiency, such as price gaps, do not have a îmagneticî characteristic that draws predictive value. Gaps do not need to îcome backî before doing X movement. As price continues, the gap becomes less relevant; we call this phenomenon Market Inefficiency Decay, which is managed in every strategy we develop with guardrails to prevent overfitting. The idea: Be intentionally late so you are not paying for randomness. Instead of chasing price, follow a logical, mechanically defined approach to get filled at favourable prices. An additional incentive: With certain derivative products, such as regulated CFDs or FX offered by a serious, reg≠ulated firm, you can even be paid via favourable slippage on limit order fills due to how synthetic ìliquidityî works on their platforms. This is discussed later in the STSí Execution and Venue Mechanics material, which includes guidance on how to filter out poor brokers. There are few conflicts of interest, as we are not affiliated with any brokers and do not promote any. 8.3 A heuristic to simplify: 1. Imbalance (Anomaly) 2. Inefficiency left behind (Something to be corrected) 3. When price meets the movement, it is balanced briefly (Price delta advantage) We get filled only at efficient prices (a good price compared to a market order at a candle close); that happens on the correction when price meets entry. An edge has multiple components, each providing a small, sensitive advantage; one weak link in the chain can be enough to destroy it entirely. -Ali The entry component is about getting superior costs and pricing relative to a blind market order. Just like in business, be the person who provides value to the market instead of the one who takes it. Do not be a customer unless forced to. A market order charges the immediate bid-ask spread, and slippage often comes second, especially in FX/CFDs, while well-chosen limit orders help you avoid both. You are not an institution trying to fill 1,000+ contracts in one block; you do not influ≠ence the price or benefit from that. With your own unique execution patterns, you are less likely to be part of a crowd. Being faded is a coincidence instead of a setup; less predictabil≠ity is part of your edge and its scalability if it persists. Let the impatient customers sell into your level; if they persist, concede at the stop loss. The mechanics of filtering or trend bias are to apply a technique where it is favourable on average to carry out your trading decision. The target is to benefit from the expansion towards ínew valueí which price discovers whilst trading from A to B over and over again. You earn an edge through the understanding of how data works, understanding the auction, and picking up what deviations from efficiency are. With each idea you aim to exploit: 1. The Marketís Process (Market Microstructure) 2. The Marketís Desires (The Auction) 3. Data Science Principles (Statistics & Mathematics) These are three robust characteristics that will not fade unless things are restructured aggressively. The only thing HFTs have done is amp up the speed. 8.4 The STS Time Series Inefficiencies and Market Profiling Because we use candles indexed in time, there is a temporal aspect to it, so I only get filled with limit orders. If the imbalance isnít corrected quickly (e.g., by the next candle), I pull my order (expiry set in advance). If there are multiple candles, I wait for it to be anchored to a logical dynamic value related to the strategy. Time Series Inefficiency Context: 1. A formation that should not persist without natural rebound in a highly efficient time series data set. 2. An opportunity revealed by price to absorb aggressive order flow to provide an edge You do not need TPO/Market Profile as a retail trader to see what has happened; OHLCV reveals enough. For example, you donít need a market profile to see a single print, and I have provided proof below. All a market profile is, is another chart. We find it to be a complex way to get the same answer. Of course, it can provide slightly more entry precision in some cases, but the extra commitment is not worth it for most traders. We do not use these tools. All time highs explained: All-time highs just mean that there is no overhead inventory (longs) to offset. There is less to ìrebalanceî because it is a guarantee that anyone holding longs is in an unrealised profit before costs. Anyone net short is guaranteed to be at a loss, which later turns into more buy volume. Swings are essentially market profile tails... Nothing special; they can be defined with standard OHLC bars mechanically using our methods. This shows the starting point but also the reset in the counts; this is an example of making consistent rules that are so mechanical to the point that if you had to code it, the classification would be consistent without interference. To be clear, aiming at swings should always be the secondary choice; aiming for swings alone can lead to suboptimal aiming at complete/finished auctions. To mitigate this, we aim for clusters or other inefficiencies if available. Price Management If a swing high starts forming within a swing high and then completes, the latest is prioritised mechanically, which removes the need for any subjective discretion. Order Flow Mechanics: Context We will call candle 1 the imbalance candle. On the time series (chart), the price must close on or above a wick high, we will call this 2nd candle the breakout candle, and the third candle (reversion) must close on or below candle 2ís starting price to show completion. Multiple candles may form before the reversion candle appears, indicating that the reversal is not always immediate. This is illustrated in Figure 15. Order Flow Dynamics: The Development of Price Candle 1ís high is to identify the area that needs to be broken (to show active imbalance). Candle 2 is used as evidence to suggest buy interest was significant in relation to the liquidity available (sell limits/asks) to the point where it breached candle 1ís price extreme on the next step in the time series chart. The close is to show its registration, which takes place and shows that the move was sustained. If there is only a wick, it indicates that the price was not held beyond the price level for the time series close. Candle 2ís opening price is a reference for price delta neutrality (efficiency/zero movement). If the price returns to candle 2ís open, there is a net zero change in price (efficient). If it closes below, it confirms a negative price delta (inefficient). With reference to Figure 15, multiple candles can form before the reversion candle hap≠pens, but it can be the same for the breakout candle, which is classed as candle 2. The extreme of the swing (the highest point on a swing high and the lowest point on a swing low) we interpret as the peak of short-term inefficiency, which is often great for targets. Candle 2ís open is used as a mechanical reference for neutrality, while Candle 3 serves as the marker of correction or over-extension, showing that price has undone the movement. The exact opposite is the case for our mechanical swing lows. Order Flow Mechanics: The Function Figure 15 isolates an attempt by the market to discover higher prices that fully pulls back and overextends, flipping into a failed auction (as the price fails to hold the breach). This formation shows that either sellers (market sell orders) regained control of that small auction by outpacing the available size offered on the books (buy limits), or that market makers were not as willing to provide liquidity due to adverse selection (which can also leave an unfinished auction). For this to be a selectable target based on minimum target sizes, the price would have had to continue for it to be a potential targeting area, increasing the chance of it being a failed auction with unfinished auction(s) left behind or at the target, which price tends to correct or exceed later. Takeaway: The entry component of your entry strategy should be designed to get an efficient entry to catch the marketís drift (from a trend, reversal or mean reversion), benefiting from the rebalancing (return) of price to finish the auction. You anticipate the marketís efficiency. Click here to learn the STS Wick Clusterís targeting logic. Note: The 3:00 to 6:10 segment includes STSís wick cluster logic. Ignore the rest of the footage. These are generic ísingle printsí in standard market profile but adjusted to the immediate highs and lows offered by the market. Our finding in Figure 17 is that the inefficiencies revealed through standard îmarket profile or volumeî analysis may provide fewer data points to work with. While it can still serve as a useful filter, it is not strictly necessary; OHLCV data are often sufficient. The same time-series inefficiencies could have been identified without relying on market profile or market≠by-order data. To be clear, this is not instruction on how to use TPO/Market Profile; it is showing how OHLC charts look. We believe that these tools are largely hindsight but useful for understanding markets and their basis; do not get distracted. 9 Backtesting Quality When you curve fit, you rig your backtest to look amazing on paper, but it will never work in real conditions. When you backtest properly and respect data science rules for quality data, you will see how tough it is to backtest something that shows strong profitability. If you have something over-fitted, you will suffer. Unless you have a strong understanding of market logic, your chances of finding something exceptional on each backtest are low, even then, it is a hit or miss. Remember, I said exceptional. What gives your understanding value is that you know the data is high quality. 9.1 Good Backtesting Hygiene Must Be Prioritised Such as preventing data leakage (look-ahead bias). Look-ahead bias is when a trader subconsciously lets future data they have already seen influence their trading decisions during a backtest. For example, a trader with poor discipline may choose to go long where they would normally short because they saw that the completed daily candle has a strong bullish close, even though this is not in their rules. This happens a lot when retail traders backtest discretionary multiple-timeframe systems, resulting in subjective outcomes that appear effective. When people automate, this happens when their system can see future data, which influences future decision-making in ways that are not reproducible in real time. Look-ahead bias is super dangerous, as you never have the luxury of seeing exactly how the future looks when trading live. Hindsight Bias This is where you believe that an outcome was obvious after the fact influencing your back-testing data. îActually, I would have placed this trade...î, this phrase is said by a hindsight trader after the movement has completed. Recency Bias Recency bias is where a trader prioritises recent patterns to influence future behaviour instead of larger data sets. Data Snooping This is when a trader analyses the same data repeatedly without a predefined hypothesis, e.g., multiple timeframe analysis applied inconsistently. Trader example: one setup uses 4H and 5m, and the next setup uses 1H and 5m. Curve Fitting Curve fitting is when a trader tweaks a strategy to improve the result of the data instead of doing logical adjustments which complement the hypothesis. e.g., simple changes, such as using a market order fill instead of limit orders for positions with no pullback and a fast reversal. There are so many things that traders do that are wrong when backtesting, and it ruins their chances of success. The worst thing is people often do it for years without noticing. 9.2 You Cannot Rely on a Single Strategy for Long-term Success You must know how to design profitable trading systems because when market regimes change, strategies will lose effectiveness. I refer to this as edge decay. Do not become emotionally attached to your system. If you cannot adapt, and you rely on a single strategy, you will always be wiped out. Eventually. Markets change as they are dynamic. People expose large amounts of their capital instead of isolating their risk in smaller amounts. We call this capital partitioning. If you do not take the steps, you can have a good run, but eventually you will be wiped out. Keeping risk well managed, especially with withdrawals, is a part of the game; greed tricks us into thinking it is a betrayal. It is important to never give in. Industry/Market Practitioner Capital Management It is not uncommon for similar procedures to happen in the industry, for example, with algorithms, with prop desks, risk is spread out. There is no lump sum deposit where one strategy is deployed, and that is the single thing that is running indefinitely. Our Capital Management Principles It is better to deposit lower amounts with higher risk percentages to amplify compounding effects to maximise growth per strategy (must be done properly to not worsen drawdown). We also isolate our risk per trading setup, so multiple trading setups do not pull each other down, as we want strategies to compound on their own. 9.3 Intraday Trading Session Optimisation Times are listed in ET/EDT (New York Hours), adjust to your timezone post-reading. There is a lot of why and what here. The values presented in parentheses are Standard Deviation values, which show variability in price movement in this case. A sample of >22 years was used to calculate these Sdev values. Remember, times below are in New Yorkís time. STSí Volatility Expansion Windows (VEWs) [1]: Generic (Main VEWs) 3:00 to 4:00 London Open (Also the DAX, Germany) The mechanism at play (The Why): The FTSE in the UK and the DAX in Germany open simultaneously at 3am (21.46 Sdev [03-04] spike from 15.78 Sdev [02-03]. The What) 9:00 to 11:00 New York Open (36.11-42.28 Sdev) Mechanism: The îopen auctionî starts, and fresh liquidity floods the market; the response causes repositioning across multiple asset classes, which is why the volatility is not exclusive to US equities and index futures. (This is key) 8:00 to 9:00 Europe and US overlap Mechanism: The reason why this hour is significant is that the NYSE starts publishing additional information about the market state before the open. Furthermore, high-impact US macroeconomic data is released within the same window, such as CPI and NFP. This is the mechanism, the increased variability is reflected within the Sdev value (28.76 Sdev [08-09] spike from 19.78 Sdev [07-08]) [1] 11:00 to 12:00 London and DAX close (Nuanced) Mechanism: There is less information to process as the market has been open for over an hour, and the European marketsí closing only reinforces the decline in incoming order flow, not the other way around. As a result, the standard deviation drops for the S&P 500, since the US equity market is dominant. This happens with or without these European market closures because there is less information for the market to factor into the price. 42.28 Sdev [10ñ11] to 32.96 Sdev [11ñ12]. 15:00 to 16:00 New York Close Mechanism: Repositioning before the close rattles the market once again (Over a sample of >22 years), and Sdev jumps from 36.80 [14-15] to 50.79 [15-16]. These highlight where volatility spikes. There is a U-shaped distribution: volatility spikes on the open, stabilises and comes lower as the session progresses, and variability spikes again at the close. I like to refer to these as volatility shocks. They happen at the open as pending instructions to buy or sell flood into the market. As orders are filled and the market calms, it becomes more efficient, resulting in less volatility. Towards the close, repositioning creates another shock. [2] An example of instructions: An investor decides to buy or sell VOO (S&P 500 ETFs) at 19:00 ET on Vanguard. Since the market is closed, Vanguard queues the order as a pending market order, which is executed at the next trading dayís open (around 9:30 ET). An example of this playing out: At the open, a lot of buy and sell orders hit the market, causing sharp price movements. As orders are filled over time, the market calms down and volatility lowers. Towards the close, traders reposition, and prices move sharply again. It is important to remember that the price itself does not show a U shape because liquidity is not provided equally to buyers and sellers. STSí Optimised Trading Windows (OTWs) [1]: OTWs are private, pre-selected, testable trading windows we have built from session be≠haviour and volatility structures backed by mechanism and verified over a 22 year sample with Standard Deviation measures from an institutional source (The Federal Reserve Bank). We provide these windows in our proprietary materials. These windows are considered ìoptimisedî due to their underlying mechanism and the values they generate, which can effectively support various strategies. Opportunity during intraday timeframes is not distributed evenly throughout trading ses≠ sions. Recurring intraday session intervals can give rise to identifiable shifts in liquidity, activity, and short-term market volatility. Such factors are important for system construc≠tion, but they need to be considered more broadly than precise trade entries. OTWs are designed to make each strategy runnable in a specific context without overfitting timing. They are practical trade deployment time ranges chosen to align a strategy type with the hours when the mechanism the system relies on is most likely to hold. We will share one window below as an example (mean reversion). The NY U-shape trough (A Nuanced window) ï 11-14 (32.96, 29.17, 30.52, avg Sdev 30.88), this is the relative trough within the sessionís U-shape volatility distribution; it is not low in absolute terms. Figure 18: U-shape profile An example: 42.28 Sdev [10-11]| B: 29.17 Sdev [12-13] |C: 50.79 Sdev [15-16] Note: I included 14-15 ETís values in the figure for accuracy purposes. The mechanism is that there is less information for the market to process at the trough, as price discovery has taken place for hours, fulfilling past demands. This increases intraday efficiency, as there is less new information to factor into the price. This typically results in more efficient price behaviour, with lower autocorrelation in returns. Mean reversion strategies can benefit from a temporary increase in efficiency. We also combine time ranges to take advantage of predictable volatility expansions through≠out the day. We refer to these as îdouble rangesî, and this was the same tactic we used in the early 2020s to trade efficiently on low timeframes despite high trading costs. The takeaway is to avoid trading all day, it is inefficient for most strategies, it is not required. I added a VEW for Asian STS traders too. The Asian Market Overlap Context: Japan is open at 19:00 ET, Hong Kong & Shanghai Opens at 20:30 ET. This provides a surge in global market activity which spills over into other asset classes which influence other markets indirectly. References [1] Staff Reports: The Overnight Drift by Nina Boyarchenko, Lars C. Larsen, and Paul Whelan -The Federal Reserve Bank of New York I pulled Sdev values from Tables VII, Table 1 on pg. 37. January 1998-December 2020 data. [2] A Theory of Intraday Patterns: Volume and Price Variability by Anat R. Admati and Paul Pfleiderer -The Review of Financial Studies. A widely cited study examining the mechanism-driven U-shaped volatility curves we have presented. Now we will revisit variance below. Consider a 10-minute timer before continuing. 10 Why Some Losing Strategies Still Look Like Win≠ners and Variance in the Distribution of Returns 10.1 Simulating a Losing Strategyís Equity Curve Before you jump on that ìamazingî strategy you saw online, stop and think about what I am showing you in Figures 20, 21, and 22, and what it means. People can still make money through luck with losing trading or investment strategies. The figures highlight simulated returns across various strategies. 10.2 Simulating a Breakeven Strategyís Equity Curve 10.3 What a true edge looks like: +0.4 EV Simulation Takeaway: It is up to you to do high-quality tests to get a true edge. You can roll the dice and maybe get lucky profiting by accident with structure-less trading, or you can do it on purpose with robust, tested strategies; it is your choice. We both know what is required. Rigorous backtesting changes lives. Most strategies will not survive a high-quality backtest without look-ahead bias. Exposing strategies through testing saves wasted money on deployment and, most importantly, time. Click here to access a simple interactive equity curve simulator. 10.4 What Overfitting Looks Like 10.4.1 Example: Bull market long-only indicators bait Mark Cuban -îEveryone is a genius in a bull market.î Many educators, especially in the stocks and options space, promise easy gains with mediocre indicators such as RSI, Bollinger Bands, The Stochastic Oscillators and many other ípre≠miumí remixes of the same tools. These indicators appear to work really well in bull markets but the next sustained price movement down exposes their lack of robustness, giving back up to years of gains and some. Indicators are fine as components but not as systems. The concept of indicators lagging must be understood. Indicators are mostly reactive but not predictive. They are not useless, but they must be used selectively and only where they add measurable value to your strategyís underlying idea. Indicators should be used to filter trades instead of predicting movement itself. In backtests where indicators appear to be solely predictive, they have likely been overfitted and will rarely survive live trading once costs are included. Itís illusory and mainly used for marketing. Many indicators appeared to work better in the 1980s-2000s because short-term trends used to last longer because markets were less efficient compared to todayís more competitive and liquid markets, so trends lasted longer, but trading costs were much higher. Indicators are fine as components. Retail trope: îYou can take something like a 9/20 MA crossover and be profitable with proper risk management.î This is true even for coin flips, but profitability is not the same as market edge or alpha; it is the amplification provided by leverage that makes weak systems look strong. Important Distinctions: Sustained profitability is different from a lucky profitable run, and a system realising a profit is different from a system outperforming the market. Indicators should not be used to define a regime; they should be used as an aid to anticipate well-established market behaviour -Ali Donít use indicators to define the market state. If you can see the market is ranging, it is already too late. If the market is trending, indicators often lag behind. In markets where trends persist longer, they can be used as an aid. Itís all about context and fit. Takeaway: Treat indicators like thermometers instead of thermostats. A well-built indicator can tell you how hot or cold conditions are for your system (trend strength, volatility, compression). It should not set the heat or fire the entry on its own. The market does not care much for arbitrary math or indicator readings; price only responds to flow. If you choose to use indicators, use them for price filtering instead of signalling direction; indicators should not be used as a foundation for strategy. They are an add-on. Your entry logic is the meal. Indicators are seasoning. Iím not a Michelin-star chef, but seasoning alone has never fed anyone. If you are going to use indicators, only use ones that complement your initial idea for entries. 10.5 Market Cycle Theories and Traditional Auction Market Theory are dated Traditional discretionary theories from the 1870s-1980s are not realistically going to pay the bills in todayís ultra-competitive electronic markets Ornaments in trading feel nice, but it is the effect that matters. Price is not bound to any ícycleí. Modern price discovery aims to be efficient; modern liquid markets do not care about ífair valueí; price only moves because of changes to the state of liquidity. Market makers provide the price by quoting, and high size from traders shifts it; MMs react by moving the price. Large aggressive order flow dominates, but quotes move the needle. Traditional auction market theory: A complete auction on the low shows that sell interest has been exhausted (dated talk). Nuance (real life): Market makers were super interested in increasing their buy limit order liquidity provision, absorbing the market takers via quote skewing (realistic talk). Any market cycle analysis that was effective in the past would be competed away by now. If it were as simple as accumulation and distribution cycles inferred from candlesticks, I would be a billionaire. Someone refuted the îedgeî of dow theory over 90 years ago for us Reference: Can Stock Market Forecasters Forecast? by Alfred Cowles 1. The study shows that Dow-based signals produced roughly 12 percent in annual re≠turns. 2. A passive buy-and-hold approach produced about 15.5 percent per year over the same period. 3. Cowles concluded that the forecasting ability was statistically no better than chance. Why I and Ali respect this source 1. He performed a rigorous backtest in 1934. 2. Context: It directly targeted Dow Theory by examining the official interpretations that were considered its active application at the time, making it free from misrepresentation or incorrect application. This is the modern equivalent of comparing a trading strategyís performance to the S&P 500 benchmark. Wyckoff is fundamentally discretionary and relies heavily on hindsight. While many have attempted to test it, the lack of objectivity remains a common criticism. This results in two key issues: first, the framework becomes unfalsifiable; second, those who apply it randomly may succeed purely by chance. Over time, many have created spin-offs, adding more îaccu≠mulationî and îdistributionî phases to capitalise on this principle. The biggest flaw in any trading strategy is its unfalsifiability; without a way to prove or disprove its effectiveness, its true value remains uncertain, and lack of proof becomes convenient. What holds these theories back the most in modern markets is that short-term trends persist for far less time compared to less efficient non-electronic markets in the past. It is the same reason why effectiveness in derivatives of the Darvas box, like the Turtle Traders strategy, has fallen sharply. Make sure you use this document in conjunction with Strategy Design Volume 1 for Strategy Design. Use these resources as an assist. Now that the impurities are gone, it is time to move on to Mental Frameworks & Discipline References [1] Yacine A®it-Sahalia and Jialin Yu. High frequency market microstructure noise estimates and liquidity measures. Annals of Applied Statistics, 3(1):422ñ457, 2009. 11 The STS Glossary 1. Rigging the Game: Deliberately designing your rules, instruments, and sessions so typical market frictions and behaviours tilt the odds in your favour, rather than relying on luck. 2. Price Structures: Predefined areas or formations in price, for example swing highs or lows and clusters (in between), that you mark in advance to anchor entries, targets, and stops. 3. Target Setting Methods: Rule-based ways to place take-profits, preferably dynamic and structure-aware rather than fixed distances. 4. Time Series Inefficiency Context: (a) A formation that should not persist without natural rebound in a highly efficient time series data set. (b) An opportunity revealed by price to absorb aggressive order flow to provide an edge 5. Personal Liquidity Provision and Removal: How your own orders add liquidity, usu≠ally with limits, or remove it, and how to use that choice to reduce market inefficiency decay (When an entry becomes less relevant to price over time) and adverse fills (aggressive movement against your position). 6. Consistent Execution Priority: A fixed order of decision-making that removes choice noise, for example, always acting on signal A before B and using the same entry protocol every time. 7. Optimally and Properly Scaling Into Trades: Pre-tested rules for adding size after entry that respect risk, structure, and costs instead of discretionary averaging in. 8. Sentient Scaling-in Techniques: STS-specific, rule-driven scale-in patterns that aim to add size where liquidity is favourable and risk is contained. 9. Market Maker Intricacies: The practical nuances of quoting, hedging, inventory control, and quote adjustment that shape spreads, trade fills, and market price. 10. Sentient Trading Operations: STSís operating practices for data handling, routing our orders, picking our instruments, time windows, and cost control that underpin our execution style. 11. Market Classification: Categorising a market by dominant behaviour, for example trend≠ing, mean-reverting, or mixed, to decide which strategy archetype to run to maximise the chance of edge discovery. 12. Efficient Markets Facilitation: The role of liquidity providers and venues in keeping spreads tight and matching continuous flow so price discovery is smooth. 13. Main Liquidity Providers: The primary institutions that make prices and supply depth, for example, banks and non-bank market makers serving FX and CFDs. 14. Informed Traders: Participants whose orders are guided by superior information or models, increasing the risk that passive quotes placed by market makers are taken advantage of. 15. Front-running: Entering ahead of anticipated flow to benefit from expected orders, typically discussed as a microstructure risk rather than a retail tactic. For example, if someone is sure a group of people will buy thousands of units within five minutes, moving the price up, they could buy just before they do, benefiting from the price movement before it happens. 16. Adverse Selection: Losing when you quote because counterparties trade only when they have a better view, meaning your fills are systematically bad. 17. Liquidity Anticipation: Forecasting where and when demand or supply will cluster, then adjusting quotes or execution to reduce inventory risk, which is what high-frequency trading algorithms are famous for doing. I like to refer to this as ífadingí 18. Algorithmic Behaviour and Distributional Decay: Automated strategies that exploit patterns, and the tendency for their returns to fade as the underlying return distribution shifts over time. In documents, I call it îdistributional decayî to keep things short. 19. Market Maker Function and Regulation: What market makers are expected to do, provide two-sided quotes and orderly markets, and the oversight frameworks they operate within. 20. Proprietary STS Capital Partitioning: Splitting capital into separate, purpose-built accounts so each strategyís risk and behaviour stays isolated. 21. Capital Optimisation: Allocating and resizing capital to maximise long-run growth within drawdown limits, for example, sizing rules that stabilise equity. 22. Risk Isolation: Keeping strategies, instruments, and risk limits segregated so a problem in oneís unfortunate poor performance cannot contaminate others, improving the compounding quality of successful trading systems. 23. Optimal Withdrawal Time to be Safe from Edge Decay: Withdrawing profits at defined equity milestones or intervals to bank trading gains before time-limited edges fade. 24. Capital Reinvestment Models: Prewritten rules for recycling withdrawn profits into new or existing strategy buckets without raising day-to-day stress or discretion. 25. Tails and Tail Risk: íTailsí in finance refers to extremes in measures or outcomes; ítail riskí refers to volatility in returns. High tail risk indicates that the strategy has a high potential for both positive and negative extremes, while low tail risk suggests a stable return rate, which can still be high but remains stable. 12 STSí Market Principles Written by the Sentient Trading Society 1. Intraday market movements are highly random when isolated but the market itself is not 100% a random walk making trading edges possible. 2. The market is an averaging machine. 3. A frameworkís logic should always be tested first. 4. Once emotional decision-making enters the process, it becomes gambling rather than trading. 5. In markets, following the crowd usually means buying high and selling low (loss of edge). 6. The only way to make a profit from buying is if people buy after you do, and the only way to make money shorting is if there is sell volume after you. 7. Markets are neutral and emotionless. They reflect information and behaviour rather than fairness or morality. 8. We Cannot Rely on a Single Strategy Forever for Long-term Success. 9. The edge is already dying the second you discover it. Act accordingly. 10. A real trading edge comes from staying ahead of predictable behaviour instead of participating in it. 11. Forward testing is not discovery. We think of it as using confirmation bias for validation to execute. 12. The only reason price moves is that there is an imbalance between the buy and sell volume offered. Nothing else. 13. Liquid market prices behave this way: imbalance, inefficiency, rebalance, over and over again. Nothing grandiose or special. 14. The STS workflow: Logic -> Rules -> Data -> Optimisation 15. Good Backtesting Hygiene Must Be Prioritised 16. Decision Fatigue Mitigation: The Hidden Edge in Trading Is Removing Decisions 17. Structure before everything. Logic before data. Consistency before optimisation. 18. Market makers will provide excess liquidity at stop clusters and benefit indirectly from the absorption, but they do not engineer large adverse movements to take your stop loss, as that would involve too much directional risk and potential fines. Everyone would see the manipulation(s), and institutions have already been fined hundreds of millions (USD) for misconduct, even in over-the-counter markets without a central exchange. 19. Most people who overcomplicate with ësmart moneyí or ëinstitutionalí talk are waffling. 20. Logic before data, why before what. Sure, your strategy did well on a backtest, but why would it continue to? Thanks for reading -Ron

Logical Fallacy Handbook: Cognitive Blind Spots

The Logical Fallacy Handbook The Sentient Trading Society Ron Introduction If you want to fix your psychology from the roots instead of medicating the decaying leaves, this is worth the read. This write-up offers guidance on how to conquer your trading psy≠chology, addressing multiple nuances. Whether you decide to take the red pill today or in 3 years is your choice. Solid trading extends beyond the basics such as charts, data, and execution. Every click, decision, and trade you take is filtered through the human mind, and the human mind is far from rational. Besides my personal experience, studies such as Born to Choose: The Origins and Value of the Need for Control by Lauren A Leotti, Sheena S. Iyengar, 2 Kevin N Ochsner prove that humans are irrational and will often give things up unnecessarily just to feel in control. As humans, it feels good to believe we have a choice or a say in the matter. This is why intuition in trading is so appealing. However, it comes with too many drawbacks and lacks reproducibility. Cognitive biases push people in the wrong direction, and it is innate in us to be nescient. The only way to overcome these flaws almost permanently is to study and understand our own flawed reasoning. This awareness creates the cognitive dissonance needed to pause, reflect, and improve. Most people will make psychological mistakes in the beginning, and that is normal. But if you want a successful trading career, you cannot allow these mistakes to persist. Markets are neutral and emotionless. They reflect information and behaviour rather than fairness or morality. What often determines a traderís success is their clarity of thinking while executing their edge. Mastering trading psychology without the acknowledgement of your cognitive biases and your weaknesses does not exist. îFugayzi, fugazi. Itís a whazy. Itís a woozie. Itís fairy dust. It doesnít exist. Itís never landed. It is no matter. Itís not on the elemental chart. Itís not fucking realî -Mark Hanna, Wolf of Wall Street. 1 Highlighting the difference between correlation and causation In trading people often make the costly error of confusing correlation and causation. Correlation describes when two variables move together. Causation means one actually drives the other; it is the underlying reason why a move≠ment happened. Why before what. Always demand the mechanics of why something happened; never settle for a pattern alone. Coincidences are often justified with inductive reasoning, this is a common mistake. Inductive reasoning is drawing conclusions from observations or examples instead of using correct principles to come to a conclusion. Mistaking correlation for causation example: îIn the last three months whenever the Nasdaq (NQ) bounces exactly goes up exactly 1.5% in a trading day, gold (GC) starts trending up so itís probably the algorithms switching their flow.î this remains a correlation, not proof of causation. Speculation at best. Two completely unrelated events that appear to happen together the example given above is a narrative correlation which does not prove causation. The reason this happens is that the human brain is naturally pattern-seeking. Itís an innate desire for us to connect the dots and make things seem predictable. Weíre hardwired to prefer certainty, or at least comfort. The Common Culprits: Statistical laziness (Common): Many traders rely on coincidence and short data samples instead of rigorous testing. Emotional validation: Correlations that support a bias feel like proof or íenoughí. Traders often feel better accepting a small sample size to validate an idea to BELIEVE in it rather than putting in the work to potentially invalidate what they are invested in believ≠ing. Many traders procrastinate with backtesting large samples to protect themselves from disillusionment. How it happens: A lot of the time, the more a trader commits to an idea, the longer they procrastinate, delaying the actions required to avoid the potential pain and additional effort needed if the data invalidates the idea, such as developing something else to take its place. How to resist it: Having an appropriate sample size of 100+ per instrument, per setup is like taking the red pill; settling for low samples is the blue pill. Strategy decision noise is holding many traders back so I will cover it quickly before we begin I refer to this as îimbalanced execution priorityî, and it is related to chaos theory/the butterfly effect, where the initial conditions have a large impact on a data setís path/results. Many modern traders trade multiple setups or instruments on the same account without ac≠counting for how they rotate the positions. For example, they could trade 4 markets looking for 2 setups on each but only allow 1-2 running positions at once, randomly missing trades whilst holding others. Imagine there are two identical traders using the exact same strategy method. One begins on January 6th, the other on the 7th of January. Because of that minute single-day difference, they end up in different initial trades. Those first trades then affect everything that follows for that trading strategy: which setups get triggered, which stops or targets are hit, and how the strategy evolves from that point onwards. So even though the rules are identical, the two paths diverge immediately, caus≠ing noise in the results both for backtests and real-time deployment, such as in forward tests. This path dependence guarantees subjectivity in data. This makes it likely that the or≠der in which events occur determines the outcome rather than the events themselves, which is not ideal for trading strategies. You cannot allow the day your strategy begins trading to influence all future decisions. This is inappropriate, as it makes trading dependent on luck rather than on the relationship between work input and output. For example, a trader could run 2+ trading techniques and multiple instruments. But the trader only has 2 positions maximum running at once This introduces noise in your trading results because you miss trade executions every time the strategy overlaps. For example, a trader could get filled on 2 setups, and whilst those trades are active, 3 more setups form, which are ignored as youíre filled on trades already. Even if you take account of this in a backtest, the results still have noise because the execu≠tion priority is random. This means the day you start backtesting or the day you start trading influences up to all future trading decisions, making the path on walk forwards random. This is not acceptable; the trade or day you start on should not affect future decision-making, as it is bound to add randomness that can have a persistent effect. The day you start will produce a different equity curve, which is normal, but remember that if it influences your future positioning, subjectivity is guaranteed because if X tradeís outcome was different (out of your control), the next trades will change as seen in Figure 1 Figure 1: Imbalanced Execution Priority This could be the difference between having a profitable or negative result. Designing strategies while being aware of this prevents it from happening. Design strategy rules to ensure that the starting day does not affect future trades. To mitigate the chance of imbalanced execution priority eroding your strategyís data quality, we have provided guidelines to prevent it. 1. Run strategies within a consistent time window that fits your constraints as stated in Strategy Engineering Vol 1 2. For swing trading, we suggest a cut-off point at a fixed time. For example, end of month (monthly close) or end of quarter, to close positions and reset used in combination with maximum holding times for trading costs reduction. Note: There are other ways, but these have minimal interference with overall execution pattern. It is important to avoid trying to optimise the time you close out positions, as that is curve fitting and only reduces the quality of the strategyís data. 3. If your strategy takes multiple entry triggers into account, separate them. Instead of cancelling one setup in favour of another inconsistently, we isolate each setup on a different trading account to manage risk and compounding. If your trading environment, e.g., a prop firm, has maximum exposure limits, separate and split trades to execute between different accounts. Do not just írun the best oneí and believe that price will discover quotes in the same way it did 1:1 this is idealistic thinking. Light application of these principles (Universal) 1. STS style risk isolation How: One strategy per account applied on one market or one asset class. (e.g., This market only, Metals only, Indices only) 2. Consistent session starting times for day trading How: Not îsometimes starting at this hourî. Always that hourî 3. Maximum holding times for swing trading How: e.g., max 3 days holding time before exiting based on the setupís logic and costs. Do not curve fit this value Do not attempt to try to find the îbestî value historically, use a value that fits within cost and strategy constraints instead. Contents 1 Learning from trading communities, debates and modern society 6 2 How to Detect Emotional Traps in Real Time 10 3 Vanilla but costly. Common Logical Fallacies to Avoid 11 4 What to do with what youíve learnt 13 Before we immerse ourselves in psychology, it is important to accept this. To succeed in most ultra-competitive fields, such as trading, you have to go against what is natural. Our trading psychology material is about escaping what is natural. Discipline and fortitude do not come naturally in this environment, they have to be developed through conscious deviation from instinct. Anyone can feel smart when following examples that have already been curated for them. That is not the same as thinking for yourself. That is not the same as building something of your own. That is not the same as being able to survive in the markets. A tutorial tells you what to do. Guidance teaches you how to think. In trading, this difference matters a lot, as fixed step-by-step instructions may have a place in narrow technical tasks such as arithmetic, but serious trading development depends on reasoning, judgement, testing, and refinement inside a clear framework. This is why we put so much time into learning. Trading is not a joke. It is not a game where you collect concepts and somehow become capable by being around them. Regardless of the path you choose, at some point you have to accept the burden of doing your own thinking, along with all the confusion and friction that comes with it. This is not a choice. Now we will move on to the mental side . 1 Learning from trading communities, debates and modern society For a psychological baseline we must understand first how inefficient thinking works in soci≠ety, which is witnessed in both social environments and media. This wonít only help your trading but also social perception and interactions between traders. Many people will pause and ask why. In markets, suboptimal reasoning doesnít just lead to insentient arguments; it leads to poor trading performance, which results in financial harm and disillusionment in the end. Every false assumption, emotional appeal, or manipulative narrative erodes objectivity. If youíre serious about trading, you cannot afford to let this subjectivity take over; if you do, it will be reflected in your P&L. Key fallacies: 1. Posturing, False Assertions and Appeal to Authority Psychological Posturing: Posturing is when a trader attempts to position themselves as superior or the authority confidently or aggressively, even when they are uncertain of an outcome or topic being discussed. This usually involves making bold claims, or asserting themselves with high levels of confidence, even if they donít possess rigorous evidence. (False assertions) It is all ego, all in an effort to curate or maintain a certain image or reputation, not to help you. Supposed experience is not an excuse for posturing. Being domineering is fine if one can back up their talk with unedited, desktop-regulated platform footage, trading statements and peer-reviewed or reproducible evidence, E.g., backtests that arenít overfitted. The problem is many traders have a dogmatic attitude, but when evidence is requested, the person asking is often dismissed. It is super important not to fall for this. Advice on dealing with posturing: It is your job not to buy into their BS. Remain well composed; do not engage in character-based attacks or mockery (ad hominem), instead disengage or ask for evidence if you care about their claim. 2. Sunk cost fallacy The sunk cost fallacy occurs when a trader continues using a strategy even after it has lost its effectiveness, clinging to it instead of adapting. This happens because it feels easier to stay in the same place than to change. As a result, some traders remain stuck in the same patterns for years, unable to move on. 3. Straw man fallacy The straw man is a classic tactic used in bad faith to manipulate someone or misrep≠resent their argument in an attempt to make them easier to attack. People in modern societies, including bad-faith traders, try to manipulate the situation by giving subtle answers that donít address the question or by making bold misrepresentations such as: Trader A says, ìI only long ES Futures (S&P 500 Futures) and itís a big contributor to my edge.î Trader B replies, ìSo you think we should only long ESî By distorting the original statement, they shift the debate into something completely different This can only take place if you allow this to happen. Do not let people derail, interpret what people are saying back to you when learning. 4. Appeal to Tradition / My strategy will work forever E.g., a trader says, îIíve been trading this successfully for 12 months-years so it must continue working for much longer.î Reality: Markets evolve. What worked in one cycle may fail in the next. While markets arenít close to being 100% efficient, the characteristics of liquid markets often resemble effi≠ciency, which erodes market edges over time. We refer to this as edge decay. The causation of edge decay is algorithmic patterns in liquidity provision change over time + other underlying properties such as macroeconomic-induced drift influencing how the market discovers new prices. Traders who are not adaptable die out. 5. The loaded question / Common educator manipulation A question phrased in a way that assumes someone is incorrect or guilty. For example, a trader could say, îWhen did you stop overtrading?î when you havenít been overtrad≠ing, to reduce your mental breathing room when you are replying, making you easier to manipulate if unaware. Some bad faith educators use this to posture. Another example is when a trading guru says, îDonít you think itís time you stopped experimenting and started trusting my strategyî It is disguised to look encouraging, but the educator has embedded two assumptions: the studentís independent testing or analysis is a waste of time, and the educatorís trading strategy is the only reliable approach, shutting down critical thinking. If the trader says yes, they give up their cognitive freedom; if they say no, theyíre dismissed as unteachable or hard to work with. Classic Emotional Manipulation. Donít fall for any of that. Good educators want to push you to do your own independent testing and analysis. Our focus is isolated on rules, edge and execution. Many educators lean heavily on chart artwork. We try to avoid that. 6. Begging the Question and Circular Reasoning Begging the question: This is when a trader assumes what they say is true without evidence. Example: ìThis strategy is reliable because it always provides very accurate entries. Check it out!î The assumption of manipulation is taken as fact without proof, e.g., backtesting data. Circular reasoning (Very Common) Using the conclusion as proof of itself. Example: ìPrice will rise because this setup has formed on this timeframe.î Thereís no logic or evidence, just repetition of a narrative. Circular reasoning feels certain, especially from an authority figure but explains noth≠ing. In plain terms, circular talk is when something (A) is framed as being the way it is because it is the way it is. There is no (B) that adds objective substance to (A), so the reasoning simply loops back around A. Examples of circular reasoning are (a) îPrice moved X way because I believe price behaves like Yî (b) Question: Why does your strategy work? Answer: (Circular reasoning): My strategy works because my testing shows that it works, or because the X educator said Y. Answer (Good reasoning): My strategy works because X,Y,Z happens in the market I am trading, which is well established and occurs due to how this asset is structured. Here is proof [Proof] to add my testing data, and real-time data shows success. 7. Techno-solutionism Techno-solutionism in a trading context is the belief that most, if not all, of your trading-related problems can be solved by technology alone, for example AI/LLMs. People with this fallacy assume that AI, if tweaked correctly, will automatically pro≠duce their desired outcome (an edge) without the hard parts, such as developing market understanding, testing skills, order handling skills, and risk controls. AI can be used to outsource testing and research (if you already understand mar≠kets), but you cannot realistically outsource thinking to achieve trading success with AI (written in 2025 for context). An example is when a trader is not aware of what makes seemingly good data poor quality, there is no way for them to trust its accuracy without wishful thinking. This is one of many nuances that can only be solved by learning what is required. 8. Honourable Mention: False Equivalence Comparing two things as if they are equal when they are not the same thing. Example: ìTrading is just gambling.î While both involve risk, rigorous trading is structured around probabilities instead of chance. Take a step back and think about whatís being compared to each other, the com≠mon mechanics, and what differentiates them. E.g., Trading and gambling are both underpinned by statistics (mechanics) and how they differ. The causation of their outcomes are completely different. In gambling the odds are fixed ahead of time (with nuance). In trading it is a path-dependent process with varying odds and potential gains relative to the risk committed. Gambling games have the probabilities being skewed against you, whilst rigorous trading has the probabilities skewed in your favour. Tip: Avoid over-explaining to laypeople who are not willing to listen; if they insist on re≠maining ignorant, let them. Being mentally enslaved by arguments only holds you back on your journey; it is a distraction. Main takeaway: Disengage . Research Privately . Improve. Now we must explore more pointed cognitive errors and measures to prevent them. 2 How to Detect Emotional Traps in Real Time Recognising is survival for serious traders. The difference between a consistent, persistent trader and a disorganised, erratic one comes down to awareness and active attempts to refine. Everyone feels fear, greed, frustration, and hope, but the best traders notice those feel≠ings early, before they alter decisions. Large drawdowns are painful; you can only mitigate the pain and its effect. The key is to position yourself in advance so you have the cognitive awareness to catch the dissonance before it drives your behaviour into tilt. Watch for the feeling of panic: If you feel this, pause. Be sentient. Breathe, trust your framework instead of your feelings. Once you overcome this, your sense of control is restored. Do. not. deviate. Panic is rarely logical. It is your bodyís nerves reacting to a sudden accumulation of risk or spikes in potential risk (volatility). Lower timeframe traders, especially scalpers, are most likely to experience this at least once whilst increasing their trading size. It is real money; itís natural, but it must be handled correctly in real time, or you stand to lose everything. The market doesnít care if you miss one move, but it will punish you for your deviations. Stick to your rules. Youíre better than this. Signs itís happening: High heart rate (you can feel it), tight chest, racing thoughts, freezing, feeling of helplessness, and a stress-induced flush (it feels hot). Post-session feeling (minutes to hours): Unsettled/Shaky, embarrassed or humble (depending on the outcome) Action: Youíve got to step away for seconds to minutes (depending on the time until close). If the setup is real, itíll still be valid after youíve regrouped your thoughts. You must! Note your triggers for strong emotions, e.g., three consecutive stop losses, and make it your goal to be extra vigilant when youíre on edge. Once you acknowledge it and avoid reacting several times, it fades away. Reflect post-session if this strategy is compatible with your nature, if you can actually manage the toll or if you can handle the stress it had induced. Itís okay to concede and trade a strategy on a lower timeframe or trade fewer session hours. Iíve traded 5m bars for 13 hours straight, no breaks on YM futures, and that made me go insane. No rule breaks, but I knew that it wasnít for me and re-backtested my strategy at different times, lowering the session time. Listen to your body instead of your ego. Watch for Narrative Building and other micro-coping mechanisms If you start constructing a story to explain what is happening, something like ìthe institutions/market makers are manipulating this,î that innate desire to feel in control is creeping in, which leads many traders to slip into emotional reasoning. The narrative fallacy feels comforting because it creates a story to explain randomness and chaos when your strategy isnít aligned with current price discovery. If you feel the need to self-soothe, you must address the feelings that make you insecure about your system, for example, how your data proving efficiency was collected, if at all, and whether the sample size is sufficient to make you more confident. Identifier: Over-analysis of the supposed îmotivesî behind price discovery instead of market microstruc≠ture or data, feeling unsure about your maximum drawdown thresholds, and not knowing what positive or negative returns to expect typically leave a trader feeling uncertain, which is the worst thing for a traderís psychology. Whilst these traders may have enough faith to press the button at first, taking what the system offers, it only works until thereís a large drawdown. Thatís where illogical thinking takes over. îThe difference between experiencing an 8R drawdown live without seeing it in a backtest versus seeing a 13R drawdown in a backtest is night and day. Suddenly, íWhat am I going to do? Whatís going on?í transitions into, îAh, itís just another drawdown; Iíve seen this plenty of times.î Thatís powerful. Suggested Action: Strip the stories from your trading and focus only on what is visible instead of whatís imaginable. Collect more backtest data; make your first-party data collection more rigorous, including maximum peak-to-trough drawdowns, average return per setup, long performance, short performance, average monthly return in R, etc. There are spreadsheets and backtesting platforms that will help aggregate this data for you. Main takeaway: A lot of your psychology is down to your loss-averse subconscious buying it. If you believe in it on the surface, your brain, which feels comfort in control, demands the patterns to connect the dots; data provides this clarity. Now we will revert to the less nuanced classics. 3 Vanilla but costly. Common Logical Fallacies to Avoid Cognitive errors such as revenge trading and hindsight are common in discussions. So we are focusing on other serious deviations that lead to poor performance. 1. The Dunning-Kruger Effect This is when beginners overestimate their market understanding or skill. This happens on almost everybodyís first profitable run. Example: ìIíve been profitable for a month, Iíve mastered trading.î Delusional. These traders often need a slap to wake up, and the market usually delivers it within the next 60 days during the post-beginnerís-luck shakedown, where the market grabs them by the ankles, shakes them upside down until all the coin and fluff falls out of their pockets, and theyíre back at square zero, not one. Itís humbling. Anyone can have a lucky profitable run, but sustained profitability is what truly matters. Start treating trading as a career instead of a lottery. 2. Loss Aversion: Loss aversion is the innate desire in people to avoid losses, which cause us pain. This is the reason people overhold losing trades or opt to use no stop loss. Everyoneís a genius in a bull market with no stops, but they will get liquidated in a bear market. 3. Survivorship Bias and Anecdotal Evidence He succeeded with this specific discretionary strategy so I can do it; I just need to learn. -The most common cognitive bias exploited by trading gurus. 4. Appeal to Ignorance This is when a trader believes something is true because it hasnít been proven false to them. Example: ìNo one has shown that this indicator doesnít work, so it must be reliable. Iíve seen it work for a couple of weeks.î A lack of evidence is not evidence of validity. 5. Confirmation Bias and Ad Hoc reasoning Confirmation kills objectivity in trading; it is when a trader seeks out data that sup≠ports their narrative, filtering out what contradicts it, eliminating balance. For example, a trader may seek out sources to confirm beliefs that their strategy works but dismiss data science flaws, such as the strategy being fitted to historical data in≠stead of having an edge. The reason this is dangerous is because it is easy to fall into the trap of believing youíre being analytical when all thatís taking place is reinforcement through biased sources that validate your ideas. The market doesnít care about your convictions; it cares about liquidity and probability. Ad hoc reasoning is when traders invent explanations for a movement after it has happened. Example: ìMan, see, I was right about the direction; I just didnít expect the price to interact with it today.î I wouldíve, couldíve, shouldíve. No P&L. Waffling. The reason these biases are dangerous: These biases make traders feel like market wizards who just canít map it out in their execution. They make a losing trader feel like a practitioner who just needs a little bit more digging to find their gold, but they never design something mechanical, never achieving that dream. Years wasted, ouch. An Honourable Mention: False Cause (casual oversimplification) Retail trader: To be successful, you have to trade in a style that matches your personality. This common false assertion implies that there is a tight cause-and-effect relationship, ìpersonalityî and îstrategyî when relating to trading success. The Reality: To be successful you need to trade in a way that provides a positive return on average (pro≠vides an edge) that fits within your constraints. The verifiable drivers are whether a trading strategy fits within a traderís constraints and realities, such as their time available, attention, ability to follow rules, risk tolerance vs goals, etc. Personality might correlate with those things coincidentally, but it is not the causal key on its own. That is what makes the reasoning flawed. 4 What to do with what youíve learnt Reinforcement exercises: First exposure Read over sections that intrigued you the most. Second exposure One week from now: read and re-immerse yourself. Any time you doubt yourself, revisit it, click, and stick. Set an alarm or reminder in advance to follow through. If youíre backtesting, forward testing or live testing, revisit the handbook every other week after reading it a couple of times over. I produced this handbook to give you something you can actually use in your trading rather than just another PDF to read, We need this to stick. Revisit it when you are backtesting, forward testing or live, and use it as a checkpoint to keep your thinking sharp and your behaviour honest with yourself. The value only shows up if you apply it, reflect on it, and make the small uncomfortable changes in your own process. I hope you actually apply this and make the uncomfortable but necessary changes that move your trading forward. Thanks for reading ñ Ron

Cognitive Conquest

Cognitive Conquest Sentient Trading Society Mental Frameworks & Discipline Ron Introduction Flawed reasoning and posture topple every empire. This document is built around this standard: if you break your rules, you do not explain it; you resolve it. When you see a group with weak hands, you do not comfort them; you disassociate. Feel above it, because you are. Those who treat financial markets as entertainment do not get paid. We must remain firm. Let us tackle one thing at a time, together. Crusade 1: Low-signal stimuli is optional, make the correct decision. Being self-critical turns into shame, which only worsens your psychology. It subconsciously wires us to conceal mistakes, which leads to deeper cuts to our wellbeing. To dominate psychology, you must stare down your flaws in isolation. Lead by self-analysis instead of self-criticism. When you make an error, you must own it, diagnose the flaw, identify the trigger(s), and redesign your decision-making process so the moment you are about to bend a rule it feels wrong early enough to stop. If you ever feel lost, revisit the Logical Fallacy Handbook. You need to be in the place mentally where someone saying, ìI lost $200 because I broke my rulesî, feels absurd. If you are in a space where people relate to poor reasoning, it is only an anchor. Acceptance and exposure to excuses make us naturally susceptible to poor reasoning. Relativism does not produce a positive P&L in trading ìWhy would you knowingly violate a process that protects you?î This is about disassociation from what you should not be doing rather than shame. Leave those groups, and stay where accountability and rigour are normal. 1 Crusade 2: Be responsible, attentive, and selfish with engagement. Avoid risking money you cannot afford to lose. Plan for consistency instead of impulse. Do not use forums as a substitute for genuine reflection and change. There is plenty of fruitless venting on social media, rewarded with engagement instead of reflection and growth. Many let the void created by poor trading decisions erode their lives. If you are not established, glance as you pass, but do not stop to watch the circus. Crusade 3: Maintain clean edges by not mixing systems. A trader who mixes systems in testing (especially in forward tests) will never know what works. Confusion feels like complexity; our subconscious senses this inadequacy, weakening our subconscious faith in data. In backtesting, one instrument or setup may hold up better than another, but in real time the added variance from mixing systems worsens our psychology through uncertainty. It is dangerous because of how hard it is to notice until itís too late. One strategy has one path which feels more predictable. One strategy, one logic, one set of rules per account. If you want to test an addition, test and run the strategies separately. Only introduce new ideas to an account if it is necessary. Clarity is the priority. Crusade 4: Do not worship the outcome and judge only the process. The worship of outcomes is the fastest way to corrupt discipline. A winning rule breach teaches the worst lesson: that you can be rewarded for weak reasoning. At the start must judge yourself by adherence rather than P&L. I know we are here to make money, but it must be accepted that the P&L is a by-product of discipline. This is why we urge traders to taste their profit post-accumulation. It is by far the most effective way to follow the plan. Whether it is a nice piece of clothing or a nice meal, when you experience your gains, you will laugh at the idea of deviating. My first purchase was a £200 pair of shoes with ìSPXî (for the S&P 500) embroidered on it, something small but memorable is enough to stick. Build pride around clean execution. If a trade loses, and it was executed correctly, it is acceptable. If a trade wins, and it was taken incorrectly, treat it as debt because if the spiral of flawed reasoning continues, it will be collected later. If a profit was unintentional withdraw and forget, do not reward sloppiness. If you are mentally anchored by individual outcomes, psychological issues will always be there. This profession is about hundreds of trades not a couple of dozen. Crusade 5: Reduce ego exposure, anonymity protects discipline. Ego needs an audience. Once you start performing, you start defending a persona, and the market will indirectly probes it for weakness. When people are watching, losing streaks start to feel like a growing threat to your status rather than just another drawdown. Avoid public predictions, public P&L, and public debates. They create pressure to be right instead of pressure to approach markets correctly. The best traders protect their process from social contamination. Your results are private. Let your behaviour validate itself; you do not need a crowd. The Most Dangerous Retail Trading Tropes ìPsychology fixes bad strategy.î Reality: Discipline cannot save negative expectancy if your strategy returns negative, it doesnít matter how disciplined you are; the market will break you. îIt takes years to find your first trading edge.î Nuance: It only takes years if you are aimless and without good resources and/or a solid mentor. Success doesnít occur overnight; it comes from a sequence of correct decisions, since trading outcomes are fundamentally path-dependent. The biggest red flag is when a so-called educator says it will up to take years to learn their way of operating. They are subtly communicating that their way or results are not reproducible. People waste years of their lives chasing discretion or intuition-based strategies. The Three Trojan Horse Statements 1. îThis path is very long.î 2. îYou just need to give this trading methodology time.î 3. îIt is really rewarding to those who can see it through. Iíve seen a guy make $100k using this strategy.î Anecdote. When unprofitable traders recommend a trading method, itís often a way to further rationalise the months or years they have already invested. It serves as a defence mechanism; itís for them, not for you. Traders can sacrifice a great deal of time and effort; for many, it hurts less to keep believing in their method than to move on. The markets do not care how long a trader has persisted with flawed logic; substance is rewarded and aimless perseverance is dismissed. Those who ignore this remain anchored by the sunk cost fallacy for years. Flawed reasoning is the enemy. ìEvery losing trader says psychology matters more than a real trading edge. Sit in the silence and wonder why.î -Ali

Trading Psychology: Mental Frameworks and Discipline

Real Market Psychology Sentient Trading Society Ron & Ali 1 Introduction -Ali Most traders do not fail because of bad psychology, they fail because they never had a verified edge to begin with. We cannot fix uncertainty with mindset books or motivational quotes. You can only fix it with data, testing, and knowing your numbers so well that your emotions have no bearing on real decision-making (discussed later). This quick read will help you shift your trading psychology from the anchor that weighs you down into the reliable life jacket that keeps you afloat. Over the years I have unfortunately witnessed people capable of trading struggle with this idea of market psychology, while my results improved after placing full trust in rigorously tested and analysed, rule-based systems. I concluded, from this experience, that deviations due to poor psychology are attached to unverified edges. It is not a factor that exists once we perform proper testing and know what to expect from our strategy (the good and bad). After understanding the numbers deeply is when it clicks. Psychology matters, but it should not be studied heavily, as the solution is simple for most people. There are multiple studies showing human preference for feeling in control and data provides that. Your long-term success is not based on a single strategy. It is based on your ability to adjust. This is why we teach strategy design and not an individual strategy for people to copy. If you cannot create additional systems when your strategy is exhibiting performance drag (long recovery time) or abnormal amounts of losses never ever seen before, your trading career is over. Strategy design skill is key. Markets are dynamic. If we do not adapt, the market will eventually force the lesson. This is what we talk about in other write-ups. The difference is that we adjust mechanically. We refer to the decline of strategy effectiveness as ìedge decayî, and we change our strategy if the current market regime is weighing it down. Psychology is the issue. Having an edge is the solution. Fantastic psychology and discipline are linked to edge and real-time execution experience. There is no secret mindset. Many profitable traders who are discretionary have similar foundations. It is all about your subconscious buying the strategyís effectiveness, and it is easiest to do that with first-party collected data. If a trader has an edge and they experience poor periods and rebounds in real time it is training they will never forget and they will come out stronger after every drawdown. I will explain my reasoning concisely. The message becomes clearer the further along you read. These guidelines, combined with experience is how you can conquer your psy≠chology and develop that confidence you need in real time. Permanently. 2 The Impact of Psychology on Trading Traders may succumb to emotional decisions and intervene with an already built and tested strategy due to some unforeseen event. They may end up going against their testing by closing a position prematurely or changing parameters such as the location of a limit order in order to feel safer. Figure 1: Spiral of Flawed Reasoning A live position, which could have been profitable, was interrupted and changed, which caused it to become a loser or caused it to profit less. This throws off the entire system as this error cascades through the strategies traded timeline. Namely, the profitability will be removed, the edge will be diminished, and the calculations and analysis performed on the backtest will no longer have predictive power. These manual interventions by traders who feel emotional are destined to lead to a failed strategy over time. I would assume you agree that if emotions intervened just once, then they are most likely going to intervene again. Once emotional decision-making enters the process, it becomes a game of chas≠ing outcomes rather than trading; gambling. The maths stops working and the traderís edge fades. Unfortunately, the moment emotional decision-making is introduced within someoneís trad≠ing, the results typically degrade. If you trade emotionally, you undercut your own edge. Working in a systematic way provides you with more objective and informed decisions, in≠creasing your chances of sustained profitability. 2.1 Raw Interpretation of Trading Industry Psychology Traders who have been programmed will insist that îyour aim in trading is about aiming to survive, it is not about making money,î and similar phrases. This is a distraction. Trading is about making money. P&L. You should not be ashamed or humble about it psychologically. Trading development is amazing when the process is introverted, as it is a personal mission, but do not confuse this for humility. Humility and complacency do not make money. Precise boldness does. People distracting themselves with popular trading psychology content tend not to make it. Although I do suspect the reason some ìeducatorsî say these phrases is to keep people pinned down with poor reasoning, many of these ìeducatorsî unknowingly reinforce these ideas because they were taught the same flawed reasoning. The reality is, the more time traders commit to a flawed ideology, the harder it is to pull away from. Even if you know it does not add up, it is hard to escape because of sunk cost fallacies. One can try to medicate it, but without a foundation of rigour, the inevitable small storms will sweep traders off their feet, destroying their accomplishments. People are manipulated into thinking that their poor performance for years, when an edge is promised, is normal. Poor psychology is scapegoated instead of having a verified, replicable edge, whilst traders are dismissed or ignored when they question the narrative. Education regarding logical fallacies is conveniently never apart of their curriculum. This can feel uncomfortable to accept, but it is how it is. Do not let them distract you; the more you churn, the more the trading industry earns. -Ron 3 An Averaging Machine Figure 2: Illustration of the stabilisation of returns over larger trade samples. A 40% win rate 1:3 RRR strategy is used in this example (extremely effective) The market is an averaging machine. A few trades can seem profitable, or even unprofitable, but this is not enough information to deduce the correct outcome. A wide range of trades over a few months will determine the profitability of a strategy; this is because all of the trades are averaged out. Suppose we flip a coin a few times. It will not show a 50% probability distribution immedi≠ately. A coin does not flip to heads then tails then heads then tails and so on forever. It may land on heads a few times and then tails, etc. This means that with a few flips we may have 7 heads out of 10 flips, meaning the apparent probability of getting heads is 70% and tails is 30%. We know that this is not right. In fact, in order to obtain the true distribution, we will need to flip many, many times. This applies to trading too. Each new trade is independent of the previous, just as each coin flip is independent of the previous. An emotional trader will allow all trades to play out as the strategy pleases in the backtest but will not in live trading due to emotions. This prevents the strategy from reaching its full potential. As an example, notice that you cannot deduce the win rate of a strategy from a few trades; many trades are required in order to find the accurate win rate. After many trades in a backtest, we will know what win rate the strategy tends to take on. This averaging effect of the market applies directly to trading psychology. A few trades altered due to bad psychology can throw off the whole system, and the market will average these mistakes out throughout the strategiesí traded timeline. Over time, this will lead to a lot of disappointment. Long losing streaks are inevitable in trading and should not be confused with ruin. They are tough, but they are the iron that sharpens you to parry the market with deeper cuts next time. Let us move on to Figures 3-4. Figure 3: There is a 10 percent chance you will experience a losing streak = with a 50% winrate over 200 trades and 50% chance of a streak =7 Figure 4: This shows the average maximum drawdown (peak to trough) of a strategy with an average trading outcome of >0.3R (including gains and losses), especially at a higher trading frequency (e.g., day trading). This figure assumes the strategy maintains its edge. 4 The Solution From the context provided so far, we should be able to conclude something important. Emotional intervention will never improve your profitability. Realising this will make you emotional in the opposite way. Now, you will be scared to intervene with the strat≠egy, worrying that it will affect the profitability. So test your robust systematic strategies correctly. Ensure that you know what to ex≠pect from a strategy based on your backtest. With this information at hand, know that intervening will lead to less money entering your pocket. There should exist no factor which will lead a trader to make decisions based on their emotions. If there is, then the trader does not know their strategy. They have not tested it properly. They are unaware of the effects that intervening has, and hence they allow their emotions to take control. 5 Fear I am scared to intervene with my strategy. I have tested it and analysed the data to the point where I would not even dare to change the location of a limit order by even the smallest amount. This is because I know that my strategy on its own will generate me money if I follow it precisely. A strategy must be formed correctly in order for us to not want to intervene. Just know that the market does not care about how we feel, and if you do make a decision based on intuition or emotions, then you are only losing money for yourself, not for the market. The only person you are letting down is you. The market is already hard to trade as it is. We already require beautiful strategies to take advantage of the sliver of an edge that exists. Anything you do outside of your strategy just means that you are losing that small edge... for what? In reality, traders will always feel emotions when trading. You may feel excited over a big trade, bored over a few losses, or optimistic for the next few days. It is the ability to simply not act on these emotions which will make you follow your strategy perfectly. You cannot eliminate yourself from feeling them, but you can eliminate painting the chart with them. They do not matter. 5.1 The Trick Question of Consistency Most traders waste years chasing îconsistencyî, but it is often a mental trap. Market changes and edge decay actively work against any form of long term stability. Figure 5: Distribution plot based on retail profitability data. (12-month outcomes, harsh) Performance will always decline eventually, drawdowns happen, and the vagueness of ìconsis≠tencyî amplifies performance anxiety. The question worth asking is, ìAm I making money in a structured, intentional way?î As long as the gains are accumulated through sound research and testing, your gains are valid. Do not be enslaved by the outcome Retail consistency is comfort-driven. Institutional consistency is probability-driven. The professional idea of consistency is not emotional; they care about the consistent execu≠tion of their edge and exposure management instead. Retail wants certainty, but serious traders accept uncertainty as the cost of the edge. Serious traders ask ìHow do I ensure the edge plays out over time without blowing up?î instead of ìHow do I make money every day?î. For example, a +$10,000 month followed by a +$5,000 month followed by a -$3,000 drawdown month is 4k average per month, expect positive returns to be uneven. Losing traders attempt to optimise for the frequency of reward, while profitable traders optimise for survival and scalability. References [1] Discussion Paper DP25/3 Expanding Consumer Access to Investments -FCA December 2025 Do not chase consistency. It feels rewarding because of the grandeur the retail industry gives it, but when the occasional, inevitable periods of underperformance occur, the feeling of lost reliability and perceived control often takes its toll. 5.2 Priority Re-arrangement To stay sane when establishing yourself in trading, never prioritise trading over your academics or career. The lack of certainty in profitable return distributions, combined with the pressure of sacrifice, makes variance over short samples feel far more cutting. It often puts people in an unnecessarily insecure position, which breeds loss aversion and flawed thinking. The rumination and pain are not worth it. Those university lectures and labs are some of the most sentient work you can do. Do not fold to social mediaís sensationalism, 98% of posters do not show basic trading statements. When you see the nonsense lifestyle, swipe off it. 5.3 Biologically Attack Poor Psychology Strip post-noon caffeine, get more sleep, and eat fewer snacks. Prepare your own food where you can. If caffeine-dependent, taper it off. A lot of people in our age group, especially, are hooked on îslopî. Donít see this as me talking from above, we have both abused caffeine in university, but the crash and how it tampers with sleep finishes most people off. Many people like to gloss over the fact that these chemicals in our food slow us down, espe≠cially in higher doses. Depressants and stimulants, including alcohol and caffeine, amplify anxiety and should be reduced to see baseline improvement. I know not everyone can afford to consistently eat îcleanî, but try your best, especially before your trading days (Sunday to Thursday). Try to exercise regularly too, as it helps relieve stress. Amplified anxiety mixed with stress can quickly turn into fear, mitigate it naturally where you can. References [1] Caffeine intake and anxiety: a meta-analysis This study shows that caffeine doses, espe≠cially 400 mg or more, were associated with a noticeable increase in anxiety risk compared to lower doses. 6 Revisiting Real Time Discretion -Ron Discretionary traders who rely on intuition tend to have more psychological issues, regardless of if they are profitable or not. By being intuitive, you are forced to rely on yourself. This leads to drawdowns and overall poor performance/return drag being taken personally. Systematic approaches nullify this problem. Suddenly, it is your systems underperforming which you would seek to optimise or replace. We do feel pain; we are not robots. I just do not let it influence my real-time trading decisions when I am active. I have been through it and had to fight it, just like you. Once you have the evidence that your system works, psychology management becomes a hundred times easier. All of my future possible decisions have been made before the strategy is deployed (it does not change from session to session). No intuitive real-time decisions, limited decision fatigue. Once you see and once you taste the P&L (Withdrawals). It is tough to relapse. When you enjoy the money, it sticks. When you start taking money out of your trading, consider spending some, so you experience what your discipline can provide. Through experience, the discipline becomes difficult to break; The longer your success persists, the more you will resist folding to flawed reasoning in real time. Discretion can be a part of rules. An individualís specific success from having a ìfeelî for the market cannot be replicated by traders, so it is a suboptimal pathway to success for most traders. Remember, this is why you want mechanical trading strategies. These guidelines, combined with experience, will help a lot of people develop that confidence they need in real time. Discretion itself is not the enemy. Intuition is. 7 The Hidden Edge in Trading Is Removing Decisions The secret is to stop making real-time decisions. Make a profitable system grounded in logic with predefined rules regarding the entry logic for every entry, risk management and trade management ahead of time; with that, you will not need to think about anything when pressing the button besides executing your setup. Design your strategy so specifically that you have at least a good idea where your target, stop and entry are before your entry criteria is complete. This mitigates the chance of scrambling in real time. It is so much easier to be disciplined if you know the exact thing you are look≠ing for in every trading session; people underestimate this. Ditch grandiose frameworks and trade something predefined, repeatable and real. For example, when you do not have to think about where your target is because you know the specific rules and sequence for every possible target, trading within your means becomes ultra-relaxed. Before my entry technique has finished forming, I know where my entry price, stop, and target are on most iterations. That is real freedom. Zero decision fatigue. The only decision I leave myself with is putting that trade on. Even that decision is made in advance. Freedom in trading is not about having a constant fight or hustle, instead it is best to create space for clarity and control, allowing your trading edge to take over. People talk about îfreedomî in trading but opt for methods that make trading draining. When everything is planned ahead and your trading behaviour is consistent, not only are your chances of profitability higher, but your mental stability will also be improved, especially under stress, e.g., drawdowns. Efficient trading becomes boring when you first settle into it. The next step is to become creative with how laid-back you are, so execution becomes as close to effortless as possible. On most days, we cruise through our lives, adjusting and waiting for alerts to trade. We donít rise to trade; we wire it to be the other way around, and that gives us the ultimate time freedom. Strategy Automation Nuances: A trader cannot automate their strategy if their rules are not ultra-precise and clear. That is common for most îmechanical tradersî; they think it is purely systematic until they are forced to code their strategy and see discretion or hard-to-define elements. Automation is secondary if they have got a real edge in trading. Most do not. Common consequences: When traders try to find their edge through automation, most will end up with an overfitted system, which is when their strategy looks fantastic on a backtest but does not perform well in real time. Why: When a fitted to work well on noise that will not repeat 1:1 in real time. Many manual traders often tweak for the sake of better data, too, but with automation, it is sped up. What looks like casual optimisation easily turns into overfitting, which the market does not respect. For those who want to automate: We suggest you create your edge(s) manually to avoid overfitting; if you need or want to automate later, it is up to you. 8 Why we do not automate testing -Ali We have yet to find a need to automate our backtesting for three main reasons: 1. Backtesting yourself allows you to get used to the strategy so that when you trade live, you are able to do so with fewer errors. 2. We put quality time into designing each strategy, meticulously defining the rules logi≠cally, so we do not need to create 20 new strategies at a time and test them until one clicks. This supports self-preservation, saves time, and mitigates the risk of overfitting. If you automate 1000 tests on loose logic, a few may appear to be extremely effective, which is misleading (false positives). Quality>Quantity. 3. When you backtest manually, it helps you gain experience and provides more insight into potential flaws in your strategies, and it may inspire other strategies. Having a computer do it for you can take that ability away from you. Manual work forces you to observe and address nuances up close. 4. We have done extensive manual backtesting ourselves; it only sharpens the mind. Automated backtesting requires you to double-check the data for discrepancies anyway. If one mistake slips through the cracks in automation, you may not see it until you realise in real time, when it is too late. 5. Part of it comes down to preference. If we were in a position where it would be extremely beneficial, we would do it, but backtesting is something you will enjoy doing, especially when you are rewarded with results for your participation. If you want to automate and have the skills, go for it; people under STS have succeeded with and without it. 8.1 Why we do not automate testing -Ron The Reality There are 1000s of potential strategies that would be effective, but there are just as many that can be based on hope. Through failure, you can detect this. If you only automate, selection bias risk increases by extreme amounts. The longer you are in your building phase, the more you understand market logic. Everyone has different ideas, and people will naturally remember idea structures that are likely a waste of time. An average automation environment: ï High sample ï High weakness Benefit: More frequent dopamine hits. A user feels like they have found something effective sooner, although it is less likely to survive stress tests or a live environment (a major problem). Consequence: Elevated selection bias risk (outliers are statistically more likely). Automation may feel more enjoyable in the moment but often fast-tracks inefficient strate≠gies. An average STS Environment (manual or automation-assisted) ï Lower sample ï Noticeably lower weakness Benefit: More robust systems, as lower sample sizes are encouraged by our framework: strategy rules, timeframes, markets and other parameters have to be intentional and logic-driven for each strategy. This increases the integrity of each strategy indirectly by reducing selection bias risk. Two amazing strategies out of 10 intentional, logic-first builds are much better than outliers out of a large sample of builds with, e.g., tweaked parameters or other changes seeking superior testing results before logic. Consequence: More time commitment per strategy, as the underlying logic is structured by the trader before testing instead of an automated script. For our traders who choose the automation path, the testing itself can still be automated after building. Now let us move on to the dark spots nobody covers. The Secret to Sticking to Your Trading Strategy (Even When It Hurts) Sentient Trading Society (Public) 1 Introduction A lot of traders are gaslit into thinking they have a discipline or psychology problem by educators. But in reality? They have a data problem. 9/10 of the time*. If you cannot confidently answer questions like: ï ìWhat would my results be if I traded this strategyís exact rules over the past 12+ months?î ï ìWould I have been profitable?î The more guesswork (people call this intuition) that is required for your trading, the more likely you are to deviate from your set of rules. When losing streaks come, which are inevitable (they will), most traders are left guessing, not knowing stats like maximum drawdown and maximum drawdown peak to trough. That guessing kills confidence. Most traders (especially short-term) fall apart not because their system is bad, but because they have not put in the work to trust it. They are trading on hope rather than on data and mechanical proof. 2 A Personal Example Last week (April 21st 2025, time of writing) I lost the equivalent of over $70,000, my biggest monetary drawdown ever. But I did not flinch. I kept trading my strategy. See Figure 1 Why? Because I have done the work. I have the data. I have tested the rules. I know my edge is real. That gives me the mental resilience to hold the line even when the market floors me like it did. Knowing peak-to-trough maximum drawdown is very helpful. Once you know your strategy has been in a drawdown equivalent to, for example, 13 consecutive losing trades (for example, -13% on 1% risk) these drawdowns are no longer scary. They become normalised. Figure 1: Stop Losses From Intra-day Trading Dow Jones Q2 2025 -Ron 3 Data Over Discipline So if you are constantly second-guessing yourself or jumping from one method to another, feeling uncertainty, you do not need more discipline. You need more data. *The other 1/10 have a capitalisation problem. They cannot afford to lose their deposited funds. Prop firms are useful for filling that void if you have a profitable system. You would not run a business without knowing the costs, margins, and profit expectations, so you should not do that with trading. Trading is a business. A lot of people would not buy a product online without reviews but would trade a strategy with inadequate data to suggest the system is what you believe it is. 4 Practical Guidance Backtest honestly without overfitting or curve fitting, preferably with bar replay software, for example, TradingView (a personal favourite). Track results properly, preferably in a spreadsheet or text form, with stats or isolated trade setup details that can be grouped together and processed so you know how effective your system is, the returns to expect, and the maximum peak-to-trough drawdowns to expect. Utilise it to build evidence and create a word document with screenshots explaining the rules if you need it. 5 Summary Confidence does not come from wins. It comes from knowing your system has a legitimate edge. Of course, you will deviate or feel unsure if you do not have evidence to back what you are doing. Think of data as a case study to proceed with your plans. Sentient Trading Societyô What Is Back Testing Burnout? Sentient Trading Society Ron & Ali Introduction Through conversations in Q1-Q2 2025 with several traders who are deep into backtesting and are complaining of feeling burnt out, fatigued, low energy, and an inability to push through their work. They are quick to rush into comfort and complacency, thinking their 30-sample size back test is somehow enough. 1 Trader Disillusionment The experience these traders are having is disillusionment. Backtesting forces one to confront reality; a traderís emotional investment in their systemís success causes them to feel synthetic fatigue when they encounter data that fails to validate their prior beliefs of the strategyís profitability. In essence, when you find out that the system which you worked hard to build is not working in the backtest, then you will feel upset, and feelings of time being wasted creep in. It makes you wonder if it is worth continuing to build strategies and test. It also makes you likely to just ignore backtesting entirely and trade live with this unproven strategy. But these feelings are only detrimental to your success. It is just not possible to achieve profitability instantly; it indeed takes a few tries even armed with the knowledge of how to design systems correctly. Your mind is just îpretending" to be tired to avoid the rigorous work required to validate a strategy. For your sake, do not let this tiredness keep you from working hard. We have tested countless systems and it is hard to continue, but we kept going. Nowadays, it is far easier because of experience, but if you are not used to backtesting thoroughly, then it is hard to get used to it. It feels hard at first, but most traders do adapt if they keep going. We did. 2 Being Persistent In our own work, we have found that approx. 8 out of 10 systems are not desirable enough for us to run. This conclusion only comes about after thoroughly testing and analysing each one You can idealise a strategy in your head all day. But when you start collecting the hard evidence in a backtest, psychology takes a toll. By ípsychologyí we are referring to the fact that traders will flinch; they will stall. They will tell themselves that they can finish the backtest later or that the minuscule amount of data gathered is enough, and the strategy is suitable enough to be executed live on their hard-earned money [1]. 1 The truth is, it can be tiring to repeatedly commit time and not get what you want, but you must keep going. Every pushback saves time, as you get better at developing, and money, as you avoid a system with poor performance. You must persist, not yield, and obtain the robust systems you require. 3 Psychological Barriers A lot of the time, it is not the work that is hard; it is what the results might say. [2]. The core issue arises from a trader subconsciously recognising that their dataset (backtest data) is too small or the results are inconclusive or bad, leading to a drop in discipline. Of course, it is tempting to avoid that fear, but the true test of a strategy happens once this period of discomfort is ignored. Honest, thorough backtesting can give insight into whether a system is possibly going to lose you money or make you money. You just do not know which is the outcome until the test is complete. So, push through and get it done. 4 Conclusion Burnout happens due to disillusionment regarding your backtest. Recognise that for two well-trained traders, the difference between the one who finds a working strategy and one who does not is the back test. Persist through the uncomfortable, boring, long periods, as this is the only way to ultimately develop a strategy that is likely to withstand the intensity of the live markets. One thing to note is that similar feelings may occur when designing a strategy too, but that is a whole other story. Thank you, Ron & Ali References [1] Odean, T. (1999). Do Investors Trade Too Much?. [2] PMCID. (2021). Quantifying the Cost of Decision Fatigue. 2 Discretionary trading & Its Psychology Sentient Trading Society (Public) Discretion isnít the enemy; intuition is. Discretion can be okay as long as you run a fixed, consistent, logical procedure thatís been tested; itís okay to run. (something most discretionary traders donít do) Personally, I and Ali are purely systematic traders, but if you want to apply discretion to be more flexible, hereís how to do things the right way. 1 Examples of acceptable discretionary elements: 1. Trader A: A day trader Ignores trade setups during news releases. He is selectively not applying his low-timeframe strategy during news for a logical reason (avoiding slippage). This is accounted for in his backtest ahead of time. 2. Trader B: A swing trader using a specific economic report or financial release to support his trade direction for the day or week consistently ex. Interest rate changes (Economic) or COT Reports (Financials) He uses it the exact same way. Every single time. Trader B in this example is using COT Reports (Financials) in a way thatís consistent; if institutions are increasing long exposure, he wants to buy; vice versa. He might increase his risk for buys exclusively instead of eliminating shorts completely. There are multiple ways trader B could do this. He has it all backtested ahead of time. 1.1 Examples of common unacceptable discretionary elements: 1.2 Intuition / Gut feel often veiled as ëExperienceí 1. Trader C Feels like the price has dipped or spiked ìtoo fastî towards his entry so he decides not to enter because recently these trades seem to hit the stop loss often. Trader C suffers from a nasty cocktail of Recency bias paired with Ad hoc reasoning by default, followed by a tragic mix of Hindsight bias + Confirmation bias if he was randomly ìcorrectî on the occasion he deviated from his strategyís rules. Thatís how you get smoked. 1.2.1 The reason this is dangerous: These deviations are untested so it adds noise to the personís trading, randomising real-time trading results. & in a backtest environment, it causes inconsistent results. The confirmation bias is terrible, as it tricks the trader into believing deviating from their strategy was a good idea. Also, if the traderís deviation backfires, theyíll likely absorb it personally and feel unnecessary pain. The worst part. If deviating actually ìworksî for you a couple of times in a row, you might stick with it even if it begins to backfire, leading to unnecessary erasure of potential gains & amplified pain. Why does this happen? Humans seek certainty and want to feel in control. These biases help the person feel safe; instead, it randomizes the traderís results. Itís not a conspiracy or a theory; this is human biology. You must set yourself to not fold. Here is an example of a paper discussing it: https://pubmed.ncbi.nlm.nih.gov/20817592/ 2 The Repeated one-off event change Trader D changes his trading behaviour risk based on events (The source doesnít matter) Itís not tested and accounted for in testing for example Trader D could think to himself after reciprocal tariffs that heís going to ignore all of his long setups because people believe the market heís trading will continue to decline. Result: He misses out on buy setups during small pullbacks. Why this is dangerous: Even if it ìworkedî, the confirmation bias & hindsight bias would likely fuel Trader D to further sabotage his future trades, trying to randomly fit his day trading behaviour to random economic news events. 3 Still unconvinced about the limits of intuition? Ask yourself this ï How much value comes from following your profitable strategy as designed? ï How much value comes from the intuitive exceptions that you make in execution? ï Which one is more important? If you imagine a graph with two lines, you will immediately notice that the systemís value is always higher than that of real-time exceptions because the strategy is the process behind your trades. Whether your positions are based on price, indicators, or even fundamentals, the value (if any) provided by intuitive actions never surpasses the strategy, as the strategy is the foundation of all decisions made. For there to be convergence between the strategyís value and intuitionís value, the win rate of the strategy would have to be doubled over a large sample. This would be like increasing a 1:2 RRR systemís 40% win rate to an 80% win rate, which is very unrealistic and unheard of. Intuition can also damage your average RRR per trade significantly, often even more than it damages win rate as seen in Figure 1. Figure 1: The accumulated underperformance from intuition over 100 trades. This is the potential damage intuition could have in this scenario accumulated over 100,000 unrelated simulations for a smooth, accurate representation. There is equal probability statistically that gut feel has a positive or negative imapact on your trading performance The point is that intuition cannot surpass the trading idea if you have a profitable system. Intuition might turn a 1:2 RRR systemís win rate from 50% to 55%, but that modest increase in edge does not measure up to what the system provides. Whether intuition is beneficial to an individual is largely random and person-specific bound by statistical laws regarding variance and the law of large numbers. Replace intuition with discipline by converting it into testable rules that apply discretion mechanically to your trading instead. 4 Summary So discretion in trading isnít inherently bad; the lack of structure is. What makes intuition so destructive for most traderís P&L is the aimlessness in trading causing inconsistent execution patterns this leads to random results because of decision noise. If discretion is used with discipline, pre-defined logic, and is consistently applied, it can absolutely enhance performance instead of eroding it. An example of this would be avoiding news trading consistently. Instead of generic rules like ìclose out all positions before newsî, there could be a straightforward rule like ìclose out all positions exactly 15 minutes before newsî, which can be factored into testing rigorously. 4.1 The key things that every trader should apply: ï Define Every Discretionary Rule: If youíre allowing yourself flexibility, write it down. A rule only becomes valid once itís defined. ï Backtest or Forward Test: Every discretionary element should have a liquidity-related reason, e.g., News avoidance for slippage, Overnight bid-ask spread spikes etc. or historical data supporting its inclusion. ï Apply Consistency: Use your discretionary filters the same way each time. Random changes in decision-making destroy your edge. ï Separate Emotion from Adaptation: Adapting to new market conditions is logical; reacting emotionally will get you humbled. Document why youíre making changes, if any, and test them. ï Stay Aware of Biases: Your cognitive biases, such as recency bias, confirmation bias, and the illusion of control, can be your real enemies that weigh you down if you allow them to. If you feel the insentience creeping in, make sure you note it down. End note: A traderís development at the start can often be a battle between emotion and structure, as they are not yet used to the hunt. Those who learn to tame discretion can build scalable consistency. Discretionary elements, when framed within logic, can become a useful tool but must be sequenced into rules. Additional Reading (Peer-reviewed papers): Trends in Cognitive Sciences, Behavioural Decision Making Thanks for reading -Ron Read This Before You Tell Anyone You Are a Trader Sentient Trading Society (Public) Ali & Ron In This Document We Go Over ï Prevention and the mistakes to avoid, with personal examples and solutions. ï Your environment, trading groups, and the unseen consequences. ï Ego. This document discusses social and psychological mistakes around talking about trading, with examples and ways to avoid them. It is intended for traders who want to reduce unnecessary pressure and setbacks caused by how they present their trading to others. 1 Social Mistakes Every Trader Should Avoid Trading is a game of numbers and averages. It relies on the outcome of many, many trades. ìWhy tell someone how your trades are going when you have only placed 30? One of the biggest mistakes traders make socially is letting people know that they are planning on making this absurd amount of money and they have this trading strategy that will do it for them. Fast forward 3 years, and they continue to let people know." -Ali As much as some traders try to appear humble about their intent, we all trade for money. There is nothing wrong with having that attitude. Veiling it with ìI do not trade for the moneyî or ìdo not trade for the moneyî is not helpful without context. -Ron 1.1 Personal Experiences with This Mistake (We Have Made Them) I used to talk about trading and was overt with my small successes when younger. When I was a teenager in school (around 16), I made close to £200 in one night on GBPUSD whilst sleeping out of luck. I was ecstatic. I told friends, my parents, and even my teachers. Luckily I was too young for it to be consequential. People probably did not take me seriously, so it was not so consequential (luckily). For most people they have done something similar. Doing this optimistic talk, even with parents or your significant other, creates performance anxiety. People will start to ask you, ìHow is your trading going?î or get concerned for you when they do not understand it is just another drawdown. When I was 17ñ18, I had small wins here and there, short periods of profitability followed by devastation. When I was profitable, it was met with scepticism. When I was struggling, I was looked down upon. By the time I was 18ñ19, I fully understood this. I learnt these lessons the hard way. Many do not. When I made my first life changing money in my trading, my family did not know for over a year. By then it was permanent. I did not flaunt. I set social boundaries on asking about my trading and explained why. I have made and lost thousands in university lectures and labs and did not say a thing. Do not subject yourself to this. I remember my grandmother saying to me, ìMake sure you do not lose it all.î ìAre you still gambling?î My mother said something along the lines of, ìI see you are trying a lot with this. I do not want to be mean, but why do you not get a real job?î Ouch And my father cornering me and questioning my goals: ìWhat are you going to do? What is your goal? What job are you going to get?î This is the easiest way to get performance anxiety induced stress or feel demoralised, and it was all preventable by keeping my mouth shut, but I did not. It is not that your family does not like you; it is that they do not understand trading. They do not understand it or believe in it. A lot just see trading as gambling. They do it because they care, not to destroy you. These effects are often subconscious and still weigh you down. I am telling you now, if my family (especially my mother) knew I was in a £100k peak-to≠trough drawdown (over 130k USD) when I was, she would freak out, and things would be tough even post recovery. Do not even make me imagine it. It would be beyond unpleasant. It is not about hiding what you do; it is about having peace in your environment whilst performing. -Ron 1.2 Telling or Encouraging Family or Friends to Trade -Ron I want to keep this short. If you do this, you will create unrealistic expectations because of stimulation from your successes or retail social media BS. In UK college (16ñ18), when someone approached me with TradingView open on my phone, I was silly enough to tell him that I was trading profitably at the time and talked about it. They will get clingy, fast. Especially when you are ìprofitable.î This is noise that you cannot afford to have in your development stage. It slows you down. You can still run while wearing a 10 kg/20 lb weighted vest if you are conditioned; even then it is still noticeably harder. Even when full time. If my distant family or friends ask me, I still play it down and pretend like I do something else. It is not worth your sanity. When you buy luxury goods such as a Rolex watch, you do not need to explain, flaunt, or post it. Do not get pushed into revealing what you do not want to; some are persistent or manipulative. If you want a trading partner, make sure you both work efficiently and are both traders. Have high standards for traders you talk to. It matters. The thing to do instead is play down what you do or, at most, say you are just looking into it. I know it feels good to talk about making money, but the best answer is no. If they witness you trading live, just say it is a demo account. If a person in public approaches you (has happened countless times to me and Ali), play it down and say invest in the S&P 500. Do not waste your time. Listen to me. We have made these mistakes. Do not fold. It is best to be quiet as it is far easier to succeed under these conditions. If people do not know you trade, it is even better. Keep quiet and set boundaries. The short term dopamine spikes are not worth it. Do not confuse this for humility. You have a positive P&L to register. 2 Your Environment for Growth and Why It Matters: Trading Groups 2.1 Sharpen Your Edge by Choosing Your Circle These are not opinions. This is behavioural psychology innate in us. If you want a genuine long-term trading career, your environment is one of the main levers you can actually control. One of the biggest reasons smart people do not make it in trading is not a lack of skill. It is the lack of environmental control. Most of you are still hanging around Discords, Telegrams, and subreddits full of entertaining but directionless chatter, i.e., waffle. You are soaking in the opinions of people who have not built anything, have not refined anything, and have not proven anything beyond a few lucky MT4/MT5 or journal screenshots. Now, I need you to think about this logically. If someone spends a lot of their time with smokers, even if they tell themselves, ìI will never smoke; that is not me,î odds are eventually they will, even if it is just trying one. Or testing that one logically flawed trading concept. It is not always a direct influence. It is subtle. It shifts your baseline without you even noticing. It is innate in us; we are human beings. You can like the cigarette; you could get positive backtest data on something baseless. Is that noise worth it if you do not want to smoke? Is that noise worth it if you want to succeed in trading? You might not copy trades. But you will start absorbing the bad reasoning. The loose discipline or approach. Shallow risk thinking. And broken retail logic. And without realising it, you have let noise interfere with your trading once again. That gets expensive. The Uncomfortable Truth About Most Trading Communities The moment you decide to get serious about trading. Truly serious. You have to choose to block out negative influences. That means talking to and surrounding yourself with sharp, structured thinkers. People with systems, data, and discipline. Not just good vibes and memes. I know it is comfortable, but that was never enough. If you are serious about having a real trading career, you will take action on this. You owe yourself more. We want you to win. That is what we built the Sentient Trading Society for. We are not here to entertain foolishness, the Discord #rules channel reflects that. We are here to win and to help you become the type of trader who earns their profitability with structure and self awareness, not ego, logical flaws, and emotions. Now we get it. You probably like some of the people in these groups, and that is completely fine. Stay in touch with the smart ones if you want. That is what we did. During your development phase while you are still building your initial edges and refining your trading psychology, it is important to shelter your eyes from laymen. You need to cut the noise. You need direction, not noise and stimulation. The more you expose yourself to logical flaws, emotional responses, and low effort posts and thinking, the more that standard becomes yours psychologically. Remember that this is not an opinion; it is behavioural psychology. 3 Ego There is both a social and a personal consequence of ego. The focus will be placed on the social aspect, as that forms the basis of this document. Traders have egos. We have seen this in almost everything. Sometimes traders struggle to accept other tradersí ideas, and they will always want to impose their thought processes even if they have yet to show any sort of system or returns. This is a problem because most traders lose. Even based on probability, what are the chances that any one trader with an unchecked ego has managed to crack trading? What are the chances that one trader has figured out everything to do with the market? Likely less than a few percent. A traderís emotion is linked to their ego as well. They feel an emotional attachment to their current trading style, which they have probably developed over the years. Unfortunately, most traders lose money (easily near or above 90%), so the question a trader must ask themselves is if their trading style is the way it should be done. Just based on statistics, almost everyone is not doing it correctly. Yes, they may have the underlying understanding, but their application is lacking, and there are things they have not yet defined explicitly. For instance, traders think psychology is detrimental to trading, but that is not trading. That is gambling. The attachment to any existing style of trading has to be removed. We feel no attachment to any specific strategy or style. "Without this lack of an attachment, we would have been stuck on trendlines and crayons on a chart for the past few years. We realised early on that the only way to do this is to do what nobody else does properly and to have our knowledge so precise that it can be compiled without logical flaws, as we have done with the documents." -Ali ìBeing too humble will not work. Delusion will not either." -Ron ìThe only way to make proper returns is by understanding and being able to apply everything from the chart itself to the analysis behind the data, properly, with no logical flaws." ñ Ali 3.1 As Clear as I Can Be If you want a real shot at this, leave these other trading groups. For now. Cut ties with trading environments that dilute your sentience, even if it feels uncomfortable at first. Immerse yourself where precision is normal. Where deep work is respected, not dismissed. Where sharp thinking is the minimum expectation of you. Even we kept each other in check. When you go full time one day, and many of you will likely lose that desire for old noise anyway. Entertainment stops being the goal when real mastery in strategy design and trading becomes the pursuit. Until then, give yourself the best chance and cut the retail nonsense. Choose your environment the way you would choose your mentor. Intentionally. We wish we had something like Sentient Trading Society. Now it exists. Do not dilute the advantages. End note: Let this piece serve as a push to steer your course towards the intention and consistency required. The way to develop both is through reading and refinement over time, with continued re-reading of the psychology section. Each structured emergence from drawdown increases your confidence, and each material reinforcement makes you more sentient. You have the knowledge now, so the only thing left to do is fill the void with experience. Reading matters, but it must be matched with action. If you have active or persistent issues with your trading psychology, we suggest revisiting STSí psychology literature every 1 to 2 weeks, with plans to gather real-time experience. If you are active and issues are arising in real time, do the same with the Logical Fallacy Handbook to develop the real-time cognitive dissonance required to avoid deviations. Thanks for reading, and it will be a pleasure to see you succeed, both mentally and financially. ó The Sentient Trading Society

Mechanism Validation and LLMs

A Sentient Framework For Reasoning Sentient Trading Society Mechanism Validation & LLMs Ron Introduction: Regardless of a successful traderís style discretionary, or mechanical (like us). If you want sound interpretations and logically grounded trading ideas, you need to treat each ìsmart ideaî like a tree you have to reverse-engineer. We believe that reasoning needs to be deeply rooted to allow for objective interpretation. If you have an idea that seems smart try to work backwards. Trying to understand everything from the very first node leads to a purer, more complete understanding. The purpose of leading you through this thinking process is so you can actively avoid being caught in a cycle of circular reasoning which leads to poor trading performance. For example, statements like ìthe market moves this way because it doesî remove the genuine cause, which creates uncertainty. That uncertainty spills straight into execution: second-guessing, amplified stress, and temptations to deviate, which turns trading into something anxious instead of calm and controlled. Let us begin. The mindset is simple: 1. Start from what you can actually point at. 2. Walk backwards in clean steps. 3. Separate what you know from what youíre assuming. 4. Do not rely on convenience. Stop only when you hit something structural. 5. Use it for or against your idea(s). Yes, this happened because of [A Logical Reason]. 1. Here is where it stemmed from. 2. Is it logical? Why does it happen? 3. Is there objective evidence supporting it? 4. Is this something that will realistically continue happening? 5. What would interfere with the behaviour? Remember, when approaching a trading idea, first-party due diligence is always the correct approach, but at the same time we must ensure that cognitive biases do not blur our vision. There are things that we want to believe and reality often conflicts with it. The Backwards Chain Node 1: The observed behaviour Write as a blank statement free from narrative. I will provide a simple example that illustrates the extremes of reasoning with markets. ï My research shows that price typically mean reverts between on YM Futures intraday more intensely compared to other liquid US equity indices. (example) Node 1 is your firm anchor to sentience. If the initial ínodeí feels fuzzy or vague, everything downstream becomes nonsense, so it is important that things are straight forward and objective from the beginning. Node 2: The immediate mechanism What directly causes Node 1, in physical terms which have basis, the market cares about the substance. îTraderís get taken out on the highs and lows by market makers before reverting to each side, that is what causes mean reversion.î -A claim without evidence + an absolute. îThe institutions accumulate and manipulate prices to go up they engineer overextension to fade retail (less than 15% of order flow)î -Another baseless claim. The Reality (What it is) Intraday mean reversion is the repeated undoing of short-term market moves, often following imbalances in order flow. Mean reversion is often caused by market takers (traders using market orders) and facilitated by liquidity providers (often market makers), but the range behaviour can also be driven by arbitrage, hedging, and other liquidity events which will be discussed later such algorithms reducing positions to limit directional exposure. The Real Mechanism Aggressive buyers or sellers can push price away by taking liquidity from limit orders, and when the pressure from aggressive participants fade out (exhaustion), price pauses, the order book refills with orders and the liquidity providers and other participants make attempts to rebalance exposure. This offsetting of size often pulls the price back into prior value (the reversion). In ideal scenarios it presents as minutes to hours of one sided price movement and imbalances and mean reversion often follows later to rebalance at the end. ï We Start At Permanent Component(s): Information needs to be factored into prices for efficient market pricing a purely efficient market often resembles stationarity. To be clear, the market is not perfectly efficient, but it tends to move towards efficiency when conditions allow for it. And our trading styles aims to benefit from it. ï Temporary Components: A liquidity shock is a temporary phenomenon where market orders have more impact than usual due to lower immediate liquidity this causes price to keep on rising or falling until a wall of liquidity (passive liquidity via limit orders) is found to meet the demands of aggressive buyers or sellers. This is referred to as îrecoveryî, îrebalancingî or îabsorption of the shockî. When new value is found the same thing will repeat on the next liquidity shock causing price to reverse. Mechanisms are supposed to be concrete and verifiable. If you cannot imagine measuring it, it is a narrative. What cannot be measured is unfalsifiable, and its effectiveness, or lack of it, cannot be proven, making it entirely subjective. The negative consequences of limiting to no evidence is reflected psychologically first. Having proof that what you are doing is objectively sound is reassuring while you are in the red. Sound reasoning and evidence are the keys to sanity during drawdowns. We tend to regard claims that cannot be proven as useless when dealing with financial markets. Node 3: The enabling conditions We start by logical foundations and market understand led by verifiable market first principles and facts; what the market respects. You want your thinking to be aligned with the marketís true mechanics. Logical Foundation 1: Short term imbalances are often ìliquidity eventsî, not ìinformation/news based eventsî A burst of aggressive buys (market orders taking offers from sell limits) can move the Dow Jones up simply because it blasts through the offers (sell limit orders) available on the order book . That does not automatically mean there is informed buying from institutions, it often means someone had to execute their positions now (Large buyer(s)/crowds, MMs rebalancing, hedg≠ing, large liquidations, common stop placements, etc). In this case when the írecoveryí from the íshockí takes place (Assuming the price movement caused from the shock was positive) ï The order book replenishes. New sell limit orders appear at and above the prior closest available ask price (best offer). ï The buyers no longer needs to pay up for immediate exposure, so price drifts back towards the pre-shock level or beyond it which over a time in it shows up as a range and appear like a îstatationaryî set. That is mean reversion driven by an imbalance between the liquidity offered not being enough for the exposure demanded with market orders. Logical Foundation 2: Liquidity providers manage inventory, and that creates a pullback When a dealer, such as an investment bank, or a liquidity provider, such as a market maker, sells into aggressive buying, they end up short. Remaining net long or net short (especially) for too long is risky because their role is not to speculate on direction, but to collect the bid-ask spread (and, for dealers, commissions). As a result, they periodically offload inventory to return to a risk-neutral (delta-neutral) position. That ìsell into impulse (fading aggressive buy flow), buy back afterî behaviour tends to make the next move go the other way after a one-sided push, which presents itself as mean reversion. These foundations start to turn a generalisation or that ípatterní you saw in data into something real, but there is more work to do. You need to know why it would exist and persist. Basis (The Why) 1. Dow Jones Future are less liquid than other related markets In my own research, I came to the conclusion that YM may show larger or more frequent short-lived dislocations than ES because it is typically less liquid, trading many thousands of contracts per day while ES (S&P 500) often has over 10 times the volume contributing to this rangy effect. 2. Liquidity Provision Operations led by MMs often stabilise pricing At short intraday timescales (many minutes at a time), liquid stocks, especially those present in the Dow show mild reversal tendencies intraday which is linked to imbalances and limit order placement by MMs that make individual price movements require more units on aver≠age. For example, if price moves up 50 ticks in 30 minutes, liquidity providers (MMs) may pull back during the move, often widening spreads repeatedly, while offering more liquidity to sellers via buy limits to neutralise their short exposure. 3. Index futures are influenced by arbitrage, limiting directional pulls and pushes (drift) YM is an index future so when it stretches away from the group of stocks (a basket) which is used to price the index, or deviates from where its related products imply the market should trade, arbitrage and hedgers actively lean against the price dislocation. A price dislocation is a mispricing between two related assets. Price dislocations happen when oneís assetís value is higher or lower than it should be momentarily when compared directly with a related asset. E.g., or example S&P 500 ETFs and S&P 500 Futures (Both are S&P 500) An example of a Price Dislocation [PD] and Statistical Arbritrage[ARB] $DIA a Dow Jones ETF goes up before YM does [PD] HFT algorithms will buy YM futures just before there is chance for a correction (discrepancy arbitrage) they pocket the discrep≠ancy as profit[ARB]. To help remember this phenomena remember it is as Cross-market fading, algorithms fade the price across multiple markets. An additional conclusion I had came to is: Because the weight of each stock is more pronounced (the Dow is more concentrated and price-weighted compared to S&P), a handful of higher price-weighted stocks in the Dow, such as Goldman Sachs or Apple, can influence the index more. Then I discovered basket level arbitrage where institutions can buy and sell a group of stocks at a time effectively making their own synthetic ETF. This allows them to hedge or arbitrage mispricing between YM and the underlying basket of stockís value. With this institutions take advantage of dislocations within a group of stocks vs the index further contributing to mean reversion/rangy behaviour on the Dow Jones (YM Futures) An example of a Price Dislocation [PD] and Latency Arbritrage[ARB] For example, if the implied Dow level (where it should be value wise) from the underly≠ing stocks move first, but the indexís value or related markets update later due to latency (lag)[PD], a HFT algorithm may trade YM against the íbasketí to hedge that short-lived dislocation pocketing the difference as profit [ARB]. Effect (The What) Micro-structurally, that means fewer price movements at the tick level are permanent because algorithms actively front run or fade mispricing instantly across multiple markets, which pulls prices back and undoes the movement, resulting in many price ranges on a microscopic level which shows up over longer timescales intraday e.g., minutes to hours (The behaviour which you can take advantage of backed by evidence). Node 4: Structural roots and deviation What underlying things at play support that support or produce these conditions and what are the incentives for its persistence? For this specific behaviour it is: ï Fundamental market design (continuous auction, order queueing mechanics, tick size, exchange rules and market participant incentives) ï Market structuring (index linkage to other assets, correlations, arbitrage) ï Regulation, margin rules, clearing constraints (E.g., Consolidated Audit Trail and enforcement e.g., bans, suspensions and issuing of fines for rule breaking) ï How transactions are handled for the asset, whoís doing it and how and why it is traded in this or that way (especially for commodities). This is where you get your ìpurerî understanding. Attention should be given to the mech≠anisms that consistently generate the defined effect and support the strategy. Each ideaís roots will look different. The discipline is what makes it objective For each link between nodes, write three lines: ï Evidence: You need to define what you accept as proof this link? Without a goalpost you cannot score. Donít shift them, that is cheating (overfitting). ï Alternative: What else could explain it? This must be explored multiple times before settling, ï What would disprove the idea: What would make you abandon the idea? What makes it collapse? This stops the chain turning into a baseless conspiracy and something respectable. It turns it into a model with rigour. Remember that in liquid markets you generally have negligible price impact as a smaller trader, so you can take advantage of these behaviours in ways larger participants cannot, as long as you approach it in a genuinely unique way. The reasons I have named will not go away That is what makes it powerful. These underlying causes exist in the present and they will realistically persist unless there are public structural changes that threaten the underlying supports, for example, intense regimes that directly hinder the operation(s) such as, regulation, or other fundamental mar≠ket changes that weaken the behaviour(s) in question. Why before what, logic first. There is no vague îin theory this and that should happenî just clear verifiable facts why each idea you have is logical and should continue in real time. (Not a guarantee but probability is skewed in your favour with evidence). If the reasons why this behaviour just like any other chance, we re-adjust. This is what makes you adaptable. Reminder: Avoid sharing or publishing your exact strategyís rules Even a couple of dozen traders, combined, trading decent contract sizes on the same exchange (stocks, futures) or venue (FX or CFDs) using the exact same strategy can influence the path of short-term price discovery momentarily, which could influence other participants and later influence others (path dependence). Because of the sequential nature of price for≠mation, that impact (noise) can sometimes be the difference between a trade succeeding or hitting the stop loss on individual positions, degrading the effectiveness (especially in short-term trading). If not, a poorer position (from everyone placing the same trade) in the queue can lead to missed order fills, and if market orders are used, it increases the risk of slippage against your trades. Serious traders are aware of this, which is why a profitable trading strategyís specific rules and parameters are often kept private. This is serious Even if you share your strategy with a small group, they can share it further, and so on. The potential for alpha decay through sharing is relatively low, but the risk of degradation is not worth it, and if you publish it as a video, even worse on YouTube for example, transcripts can be harvested and may even influence AI or LLM outputs. The potential adverse effects in each case cannot be accurately quantified, so it is best to avoid sharing entirely. Node 5: The final question you must ask yourself is, what is the reason it does not happen 100% of the time? You must be specific. Node 5 is not solely for validity but also for psychology, as it restores the feeling of agency during inevitable drawdowns. You know, before deployment, the conditions under which your setup is likely to work, and the reasons it fails when it does, which limits real-time stress, second-guessing, or panic. Running sound ideas with verifiable logic results in calmer, more profitable trading. Why it stops mean reverting on trend days (A bullish example) On genuine information days that have news induced liquidity shocks e.g., from CPI releases NFP etc. Aggressive order flow from news that cause that initial spike is not a one-off liquidity event. It keeps flowing in, and market makers cannot neutralise their exposure without moving price further. When they close their short exposure, it turns into buy volume as they exit their position, so mean reversion fails and you get continuation (price moves higher). Other participants also exit, and that order flow (from buys) can push prices higher and higher until price meets a passive wall of sell limit orders (liquidity), where the sell volume offered via sell limits exceeds the market buy volume. Conclusion In practice, the aim is not to predict every move 1:1, the aim is to build your own repeatable processes. Each process you create has a job to identify when your edge is present in a way that deploys risk consistently to benefit. Remember on a 1:2 RRR system with a 50% success rate it takes 4 individual positions on average to get a 2R return, if a trader risks $500 per trade (example) it would take them 4 trades to make $1000 on average before costs and so on. Traders can only get that with consistent strategy execution assuming their edge survives. Main takeaway: If you can explain the mechanism objectively, you can define what would disprove it, and execute it with discipline, you move from opinions to an exploitable idea that the market respects. That is how you achieve true clarity in trading, and how you stay in the game long enough for probabilities to work in your favour. The primary steps Sound idea -> Sequence into rules -> Testing Good? -> Test further. 3+ systems, run them separately. Overthinking does not lead to a positive P&L; instead, it breeds hesitation and procrastination. Some do this for years. For some, this can develop into a long cycle of îwhat about this?î, îwhat about that?î. All think, zero action. As soon as you have verified your edges, it is best to escalate into action as soon as you get the chance. We recommend revisiting this document within your strategy development phase. Thanks for reading -Ron Artificial Intelligence x Trading Sentient Trading Society STS AI Usage Guidelines AI and Trading: Why You Should Think Twice Introduction I have seen some traders jump on the AI trend, expecting that layperson-friendly LLMs like ChatGPT and other simple tools will shortcut their path to a trading edge. In practice, that expectation often leads to disappointment. Here is why reliance on AI platforms often does more harm than good when your goal is to build and validate profitable trading strategies. But AIís deductive reasoning is weaker than the average humanís, yet a lot of people try to cut corners with it, so let us break down what tends to go wrong and what to do instead. 1 Artificial Intelligence: The Consensus Problem Majority opinion does not equate to edge. Large language models such as GPT digest a motherlode of public data from several sources derived from all around the entire internet. Sure, there is a lot of high-quality, but a lot of garbage too. When talking trading, answers provided are based on LLM induced group-think instead of what actually works. I asked GPT for a solid strategy for YM futures. GPT outputs were of opening range breakouts, VWAP and MAs, or worse, ìliquidity sweepî retail narrative backed systems. These are overused retail strategies that have little to no edge. If you press GPT, it exposes its mistakes. If you ask it to optimise your strategy, it can add these flaws to your system. On the surface, consulting AI for strategy building guidance seems smart, using LLMs for strategy generation tends to push you towards crowded templates. It is extremely saturated, too. In markets, following the crowd usually means buying high and selling low (loss of edge). 1.1 The Yes Man and Echo Chamber Effect These chatbots are built to be friendly and helpful, that is, agree with you. You feed them your ideas, and they cannot help but nod along. A consultant that only tells you what you want to hear is useless. Treat your GPT the exact same way. 2 The Misalignment in Perception and Overfitting General purpose chatbots versus institutional systems. Free to use and easily accessed chatbots like GPT are nowhere close to systems institutions build and/or have access to. Do not confuse simple chatbots with proprietary machine learning models. Using AI to build strategies using historical data is also the easiest way to create overfitted systems. AI has biases we could not even fathom. Artificial Intelligence > Machine Learning? Machine learning has been used for data mining (finding patterns within data) for many years, long before todayís public AI tools. Institutions also have data, proprietary protocols and hardware (worth millions) dedicated to making it useful, which is inaccessible to normal traders. Implying AI and Institutional ML are the same is false equivalence; they are not the same. Data mining without research regarding the cause of edge often results in useless strategies with amazing test results. 3 The Alternative for Those Struggling Look at how markets actually work. Order flow mechanics, OHLC candlesticks, and market microstructure to create your own ideas. Form your own ideas, convert them into clear, concise, testable rules, then backtest and forward test them honestly. No intuitive gaps. Spell out everything from entry to exit, including position management. 4 Where AI Can Help (With Caution) AI can be useful, but the end user must use their brain to filter out the nonsense (there is a lot and mistakes). Most people just will not be able to filter out poor information because of the required knowledge ahead of time. This makes AI more harmful than good for the average person using it for trading. AI often misquotes studies or books (you know if you have used it to study). GPT does a lot of paraphrasing, too. Example of AI being useful: Helping the end user create an indicator or automate their trading system. You still need to be careful with mistakes. You can outsource research and coding to AI, but not your thinking. All research and code must be triple-checked or debugged, which requires understanding regardless. (Written in Q4 2025) Instructions on how to do so are provided towards the end. 4.1 Why AI Wonít Save You Deductive reasoning in trading starts with general rules or assumptions and works down to specific, testable outcomes, and AI is not as efficient as a regular human when performing such reasoning tasks. AI is not as intelligent as most people think. Deductive reasoning, the branch of reasoning required for real, robust strategy building, is noticeably weaker in modern îState Of The Artî LLMs. Their ability to reason is below the human average. This was shown when testing LLMs with logic puzzles. This measure of efficiency depends on raw thinking ability. AI, such as GPT, underperforms the average human by many percentage points when asked directly with a single prompt and no worked examples (0-shot). Models still tend to underperform the average human even when the model is instructed to write out its reasoning steps or is prompted in a way that encourages step-by-step reasoning before outputting the final answer (Chain-of-Thought CoT). It is comparable to the average human being, and I believe that if you have read this far, you have intelligence beyond that threshold. Figure 1: Source: JustLogic: A Comprehensive Benchmark for Evaluating Deductive Rea≠soning in Large Language Models (2025) Speaking as of Q1 2026, there are professional jobs that specialise in optimising prompts for large companies. These people are called ìprompt engineersî. Yes, I am serious. The secondary problem is that AIís reasoning component has inefficient problem-solving abilities without human hand-holding in prompts, with many models producing success rates below 30% as seen on Figure 2. Context: Direct ask prompting: My problem is X. Give me the answer. This requires higher deductive power. Agentic prompting: My problem is X. First, plan how to solve it, break it into steps, execute the steps, and check the result. Additional Guidance is provided in the prompting, and most LLMs still answered incorrectly >50% of the time. The reason why this is a serious problem is that AI can give you the wrong answer confidently. If asked to solve specific problems (the very component required for profitable trading strategy development), there is a decent probability that the AI may get it wrong, and these mistakes cost money. Figure 2: Source: Pencil Puzzle Bench: A Benchmark for Multi-Step Verifiable Reasoning (March 2026) Figure 3: Same source with both agentic and direct ask results visible. Our point is Deductive reasoning, the branch of reasoning required for real, robust strategy building, is notice≠ably weaker in modern SOTA LLMs (State Of The Art). This ability to reason is below the human average. So while some people may be able to use AI to enhance their workflow, it is unlikely to reason better than you or your custom projects. In other words, AI will not hand you your trading edges or skills. After researching, I see AI as being like Microsoft Excel. I believe it will be used to optimise tasks and save time, but instead of being just for an accountantís time-saving, it will be used for many tasks. However, I doubt it will completely replace the operator or ìprompterî. The concept of artificial general intelligence is different, of course, but we are not objectively close to that. It is debatable. 5 Main Points ï Be your own critic. ï Avoid the passive role AI places you in. Stay curious and sceptical at all times. ï Treat your trading like a science experiment. Run fair tests and question every assumption. ï AI can be useful, but take every output with a tablespoon of salt. Using AI and LLMs Safely (Basic examples) If you do not have time to research, you can ask AI or LLMs to save time with these templates. I want you to learn how to use AI safely, so pay attention to the structure of these inputs and the warnings regarding the outputs. Now We Will Explore My Favourite Prompt Layering Protocol. AI x Trading: Prompt Layering (Universal) Guidelines on AI/LLMs use when learning Introduction The layering presented in this post are guidelines; they are not absolutes. When tasks become more nuanced, this will naturally grow your own unique process regarding AI use (the goal). Automating Tasks Using AI to outsource work, e.g., research, is fine as long as you triple check the information manually before accepting anything. What is accepted must be understood. Use paid models. Automating Reasoning Using AI to outsource your thinking, especially in trading, can be harmful as these models have been trained on a lot of poor trading-related information such as random articles, trading book opinions, video transcripts and media (noise) which dilutes serious research and valuable literature e.g., financial market books from accredited authors e.g., Maureen OíHara. Common issues AI often names techniques loosely, with little to no basis, and will be quick to correct itself or apologise when lightly questioned. You can ask for a peer-reviewed academic source priority, which helps but doesnít guarantee. Flaws often go unnoticed You donít need to overtly condemn AI/LLM for it to admit its mistakes. If you insist the LLM is wrong, it is more likely to agree with you even if its output was initially correct. We must remain aware of our incentives to believe (confirmation bias) and other cognitive biases. How we approach AI The secret is to ask the LLM to use proof in advance for nuanced questions and to verify, instead of falling into the trap of confirmation bias by asking AI to prove an assertion. AI isnít guaranteed to push back. Never say ìprove this is correctî or ìprove this is incorrectî; instead, frame it as ìcorrect or incorrectî and ask the LLM to provide evidence, direct citations, links, and context. If you are unable to understand the sources at first it is okay, as some sources use esoteric talk such as ítransientí instead of simpler language such as ítemporaryí (one of many examples) ask the AI for a simplified breakdown so it clicks. To use AI properly, I suggest layering your prompts and paying for premium models for better processing and less friction. When layering, it is important to avoid giving AI a highly nuanced task within a single prompt, as it can increase variance which can increase the chance of users missing misinterpretations by the LLM or mistakes within the outputs. Layering (Multiple checks, Simplified) 1. Open a new chat for this topic if it deviates from the current conversation for accuracy and speed. 2. Ask AI about something ultra-specific and make it clear what you want to know regarding the subject in focus, or ask AI to research what you are asking about. 3. Ask the model to perform the main objective or answer the question. If the problem is very specific, you should state your problem with a clear plan for how you want it solved. 4. Ask the AI if it is certain it is correct. (Check 1). 5. Ask for the sources or research to confirm itís correct, and review them manually to come to your own conclusion (Check 2). After that, search for evidence for and against the idea before accepting it (Check 3). Remember, AI is as a collaborative tool to enhance your personal perception. Do not replace your perception with AI. LLMs, especially GPT in my experience, can output many paragraphs whilst revealing so little value. It increases the average personís perception of value, but be sceptical of the outputs you receive. Do not accept the nonsense when your chat says something profound; press it for more substance. Occasionally, you have to interrogate because AI can some of the most grandiose waffle and the only way it sticks is if you allow it to cruise in your mind unchecked. LLMs often posture, do not let AI assume authority. Remember, deductive reasoning, as shown, is noticeably weaker in modern LLMs (below the human average), and deductive reasoning is what provides the edge. Question Structures (Universal) When asking AI something, always attempt to structure the question so it does not solicit the confirmation of your beliefs. Instead of asking the LLM, ìShow me how X is correct,î ask, ìIs X correct or incorrect? Explain how, why, and include your sourcesî within the prompt layering process. Do not accept the output at face value; maintain scepticism until reputable sources outside of layperson articles and mainstream media confirm it with mechanisms and universal, quantitative evidence. AI Personalisation Settings I avoid LLMs when I can, but if ChatGPT is the AI you choose (most common), here is how to make it less annoying to listen to. Custom Instructions (Copy and paste from PDF in your browser): Use British English (remove if you are not British) Make the paragraphs and headings like a human would Never use em dashes, they make me extremely uncomfortable Replace em dashes with commas, colons, or parentheses. Use standard keyboard punctuation only. Avoid sentence structures that require em dashes Other settings (Default) Impact: It will still talk like GPT, but it tones down that annoying waffling component AI typically has when trying to have an insightful talk. If your AI chat forgets and starts talking nonsense, input îRemember your custom instructions for all future promptsî in chat. If you have issues copying the settings successfully, re-open the PDF in a browser. Thanks for reading -Ron

Broker Choice, Venues and Regulation

The Industry: Broker Choice, Venues, and Regulation Sentient Trading Society OTC Market Microstructure Ron Disclaimer and Risk Warning This material reflects personal opinion only. It is provided solely for education and research about CFDs, derivatives and broker selection. It does not constitute financial, investment, legal, accounting or tax advice, nor a recommendation to trade, to use leverage, to open an account with any specific firm, or to operate in any particular jurisdiction or product. Trading leveraged derivatives is high risk. CFDs, futures and other derivatives are complex instruments that carry a high risk of rapid loss of capital. You can lose all of your deposited funds. In some structures you may incur losses beyond your initial investment. You should carefully assess your objectives, experience and risk tolerance, and, if needed, seek advice from a suitably regulated financial adviser. No guarantee of accuracy or completeness. Information about regulation, broker practices, platforms, spreads, execution quality and jurisdictions is provided ìas isî, without warranties of completeness, accuracy or timeli≠ness. Regulatory status, policies and commercial terms can change without notice. Ex≠amples and case references are illustrative only. Past trading experience or performance is not a reliable indicator of future results. No endorsement of any party or product. Mentions of firms, platforms, regulators or jurisdictions (including, for example, LMAX Group, Spotware, MetaTrader, cTrader, the UK Financial Conduct Authority, European regulators and others) are for identification and commentary only. No affiliation, endorse≠ment, approval or sponsorship is implied. All trademarks remain the property of their respective owners. By reading or using this material, you accept that the author and any contributors are not liable for any loss, cost or damage arising from reliance on it. You alone are responsible for your decisions, including the choice of whether, where and how to trade. An important note before we start: A different derivative is simply another way to accumulate and manage directional risk. STSí principles can be applied to any liquid market. The same strategy that works on, for example, an equity or futures market would work well with a serious CFD broker, and vice versa, if strategy costs remain below 30%. The same applies to FX (For USA). Note: If you are based in the United States planning to trade spot forex the document is worth reading, but if you have no interest in spot FX you can feel free to skip this one for now. If you are based in the UK, Europe, Asia, or most places outside the USA, you must read this document regardless of which asset class you trade. It is necessary. By the end of this document, you will possess the rare and noble ability to screen a non-Direct Market Access broker in minutes, a latte in your left hand, a mouse in the other, and a Cohiba Robusto clenched confidently between your teeth. In a skybar. Let me have fun. Allow it. Contents 1 Why I Trade CFDs Instead of Futures 3 2 Importance of cTrader and Visible Market Depth 6 2.1 CFDsdonothave írealíliquidity right? . . . . . . . . . . . . . . . . . . . 7 2.1.1 Reality: ................................ 7 2.2 CFDsareSerious Instruments ........................ 9 3 Consider your other options before CFDs: 9 3.1 Do not believe the no-dealing desk (NDD) or ECN/STP marketing. ................................... 10 3.1.1 If you want to see how your FX/CFD broker really fills your orders, thisiswhatyou needtodo ..................... 10 4 CFD Broker Quality Assessment 12 4.1 To Truly Understand A FX or CFD Firm You Need to Learn The Language 12 5 The Basics: A book vs B book Classification 12 5.1 îBestexecutionîisoftenmisleading. . . . . . . . . . . . . . . . . . . . . 13 5.2 îA bookî brokers and learning the key agreement language . . . . . . . . 14 5.2.1 Why most brokers do not operate under the agent model . . . . . 16 6 íA-bookí/íB-bookí Hybrid Execution 18 6.1 KeyHybrid Tells ............................... 18 6.1.1 Key nuance regarding hybrid classification: . . . . . . . . . . . . . 19 6.2 Why We Avoid SpreadBetting intheUK . . . . . . . . . . . . . . . . . 19 7 How to use FX and CFDs with Rigour 21 7.1 Ever wondered why a wick was longer on one broker comparedtoanother onFXorCFDs? ................... 21 8 Do regulated forex brokers manipulate prices? 21 8.1 How can I get filled where I want consistently with these price feed incon≠ sistencies?................................... 24 9 The Sentient Way & References 25 1 Why I Trade CFDs Instead of Futures There is no centralised time and sales/tape or book in CFDs. Each broker or CFDís Liquidity provider (LP) runs their own version(s). Sometimes the order bookís behaviour is static, sometimes dynamic. Depends on your trading size(s). CFD brokers often show static books up to a point. On IC Markets (A hybrid broker I used to work with before Q3 2025), for example, the first 15 units of US30/Dow Jones/15 Lots are usually easy to fill in both directions for Dow Jones. After that, things change, liquidity thins out, and you either need to iceberg or manipulate your fill price to get favourable slippage. For context, the highest Iíve had open on a Dow Jones CFD is 530 units short, which is equivalent to 106 YM contracts, which is $530 made or lost per point/handle (low timeframe trading). CFD Depth of Market Example (Dow Jones CFDs) If I want to buy 30 units (6 YM Contracts Equivalent), I might set a limit just above the ask to get filled on the drift. You learn how the book behaves. Not ideal, but manageable. Example of an order not being filled on an CFD Sell Limit Order position (This would be impossible on futures) Figure 1: The bid exceeded my order by over 3 ticks and I did not get filled on my sell limit order. (100 YM Contract equivalent in size). -Captioned in 2025 for context Comparing that to futures. There, you are the market once you get to size. You either get filled clean or you shift price. That can work for or against you. Some like the feedback loop. I donít with CFDs, I donít move anything. I can offload slowly at my target zones, within a few points, and the market doesnít notice. That is key. I get to scale without disrupting price. It is synthetic and slippage can actually work in your favour if you know how to manage it. Refined CFD platforms like Quantower and cTrader help, they let you automate partials and spread fills, all while keeping your average clean. Futures canít give you that without pushing the market at large, making my strategies more ìstealthyî and potentially more scalable up to a certain point. In my experience, the CFD broker I currently work with has liquidity providers that offer more than five times the instantaneous liquidity at the best quotes compared with the previous example, with bidñask spreads varying between 0.5 and 2.2 intraday during liquid sessions and zero commissions (Q1 2026). Summary: So why the CFDs? 1. The costs works out better for me if I use limit orders for execution. (All my systems do) 2. A key thing we get to avoid is currency conversion fees and unnecessary exchange rate fluctuation risk. I am British, my monetary holdings are in Great British Pounds, and I do not want prolonged directional exposure to US Dollars or any foreign currency unless it is intentional. When a Direct Market Access (DMA) asset is quoted in USD, I am forced to convert my GBP into USD to hold my positions, as the exchange only accepts US Dollars. These costs add up over time, and I would have to subject myself to prolonged foreign currency exposure (conversion on deposit); if the USD decreases in value relative to the GBP, my money loses value when it is time to withdraw. What about hedging our currency risk in response? Yes, it could be hedged, but I deposit and withdraw frequently, and I hold positions for many minutes to days at a time depending on the strategy, not months to years. Hedging would complicate our trading operations to unnecessary levels while still incurring losses due to rates. What regulated CFDs bring to the table (a) On most licensed CFD brokers, there is zero need to convert our accountís currency to trade any market, regardless of whether it is quoted in USD, EUR, or JPY. If I deposit £100,000.00 in January, it will still be £100,000.00 in December if no trades are executed (assuming there are no dormant fees or other charges). Sterling (£) remains sterling. (b) With CFDs, I get directional exposure and the currency exposure of the quoted instrument without needing to pay fees to convert my currency. The broker handles it all and nets the difference of the trade in my account, acting as matched principal in the transaction. This means my UK licensed broker, specifically, accepts my position and offsets the tradeís directional risk else≠where on my behalf and the result of the trade is reflected in the realised profit or loss. 3. More accurate position sizing: Futures have a minimum number of units as exposure to participate, and the market operates in steps. This is normal, but with futures, one contract can equal, for example, five units. There is no such thing as 1.2 contracts (six units); you have to step up or down by ±1 contract. Six units may be optimal for your set risk %, but as a smaller trader you are forced to choose between one or two contracts in this example, which can lead to under-risking or over-risking relative to oneís desired percentage. This benefits both large and smaller traders. 4. On licensed brokers, holding leveraged positions overnight is largely discrepancy and hassle-free when planned for and well-managed. 5. If I get the same or similar fills I would rather not be seen taking liquidity from the real market regardless of the strategy deployed. This has strategic advantages at higher sizes, minimal now, perhaps more later. In short form Because I stay mostly invisible. I can scale and manage exits over a range and not become liquid. CFDs often beat futures for costs on a good broker if limit orders are used properly & I can hold large positions overnight without tight margin requirements or spread spikes. My brokers charge $0 commissions an 0% currency conversion fees on CFDs + there are no exchange or data fees. Spreads are typically consistent on CFDs for example, IC markets had a 1.3 point spread maximum (usually a 1.1 point spread during the day) on the Dow Jones, whilst YM futures (the real underlying market) had spreads spiking over 10 points whilst reciprocal tariff volatility was in. I was saving over 90% in costs compared to the underlying asset. Final Note: Despite the regulations and measures in place, it is important to acknowledge that regu≠lated CFDs are not toys; they can be a suitable solution for retail traders, just like total return swaps (similar to CFDs) can be useful to institutions. It is all about context and use. If you are in the USA, it has been illegal to trade CFDs since the Dodd-Frank Act was passed in 2010. A backtest would need to be redone using the futures feed in order to trade the futures market of the CFD of interest. This approach requires the most effort and is generally not necessary for most markets, particularly for a European or British citizen. Do not use an unlicensed, unregulated broker to trade FX or CFDs -Ron 2 Importance of cTrader and Visible Market Depth If you operate with a regulated firm in the UK, one that is actually serious, that offers a transparent L2 book from liquidity providers or multilateral trading facilities that offer liquidity, e.g., LMAX Group, this reduces the risk of direct conflicts. This protects clients from abuse and manipulation, while reducing harmful incentives. Policies and questioning clarify how they operate as a regulated CFD firm. Transparency is required. Even if your broker does not operate as an agent (you do not directly engage with the book), it is still useful to see the bids and offers available so that, when your A-book broker matches you with the LP, you have an idea of which price levels you will be filled at and what price you will likely get if your order is offset with the liquidity providerís depth, assuming the LP is good. This should be checked in the execution policy and via support, followed by due diligence (researching the firm that will be your counterparty). I only operate on brokers that have a transparent market depth and reveal risk manage≠ment practice or liquidity providerís execution behaviour before depositing meaningful sums. Regulated brokers that have cTrader and similar platforms are forced to be more transparent because it shows trade fill information on each order unlike typical FX or CFD platforms. cTrader raises the transparency bar in practice (better order fill info reporting and market depth in many deployments). If cTrader is available to you, as a trader you still need to verify the brokerís execution model via the execution policy and venue disclosures, it is the only way to know. Always read the execution policy and conflicts of interest documents before joining. Key Information Documents are just as insightful. We suggest European traders who decide to operate with a CFD firm avoid firms that are not regulated by the Financial Conduct Authority in the UK (they do not play around). For legal reasons it would be unwise to name and shame, but there was a CFD firm that was regulated by a less serious regulator (opinion) which failed to enforce after a scandal that many would frame as a fraudulent operation. FX and CFD Brokers usually operate under multiple jurisdictions, which can be confusing at first, but make sure the UK/FCA jurisdiction is the one you are trading under. The FCA is an active supervisor and often takes enforcement action where firms breach rules, but outcomes vary. Treat FCA authorisation as a meaningful baseline for safety, while still carrying out your own due diligence. Personally, the 8-9 figure fines I am seeing are very reassuring, but it is your choice. 2.1 CFDs do not have írealí liquidity right? Common generalisations retail traders make about CFDs îThere is no liquidity with a CFD? There is no order book. The instrumentís movement is loosely based on the underlying market, you operate as an entity with the CFD broker directly. Slippage or an unfilled order is a simulation and is effectively ífakeí.î îI believe the position should open/close exactly as you choose it to the fact that it may or may not, is completely falsified by the CFD broker/platform themselves. (Referencing last look)î 2.1.1 Reality: CFDs are over-the-counter contracts with the provider. Some providers source prices from external liquidity providers and may display ìmarket depthî derived from those quotes, but this is not the same as a central exchange order book. This is why having an A-book broker is important, because you want them to offset risk to avoid direct conflicts that can arise from internalising all flow. If all flow is internalised, the broker can simply refuse to fill an order, and slippage, partial fills, or no fill at all can follow. If you are with a broker that internalises all flow and you are not experiencing this yet, as soon as you start trading larger size, the slippage can appear, and it is ugly. I have experienced it and post-reading you will learn to dodge it. The problem is that most CFD brokers purposely do not show transparency or list their venues. For most regulated ìECNî brokers, transparency can be demonstrated through broker execution policies and the naming of liquidity providers, combined with cTrader environments this which forces the degree of transparency I require before depositing. Every CFD has liquidity; it is just provided in different ways. Some are more robust than others, and the liquidity model used by most (less serious) CFD brokers tends to be selective and fragile. The problem with most CFD brokers is that they are not transparent about their oper≠ations, so you donít know what youíre getting, and you canít see a level-two book. I never engage unless I know what I am getting and it is a regulated operation. What is a dealing desk? A dealing desk is where the order submitted is filled, delayed or rejected/requoted for an inferior price, typically through last look. What is last look? Last look is when a CFD broker decides whether to execute your order or not or to delay the execution, which almost always works against the traderís best interest; this is what most CFD brokers participate in which is why the industry has a poor reputation. How last look works against traders 1. Conflicts of interests: Traders get awful slippage due to selective execution delays or order rejection. 2. Other participants or the sell-side firm(s) involved can see your trade size and intent if you arenít filled on a market order that was rejected. This information can be used to front-run you if you trade larger size. Dodgy LP: ìOh, he just tried to get filled on a 500 contract marker buy and failed? He must know something. Letís fade him or reduce our ask size/sell limit volumeî (increases slippage on attempt 2). My broker does not participate in this: My broker is regulated in the United Kingdom. If they or their liquidity provider were to go against the policies, they would be fined heavily by UK regulators. Brokers that participate in these practices have to include it in their execution policy if they are market-making without LPs. If they are using liquidity providers and those LPs use last look, it will be stated in the LPsí policies. If your broker refuses to name their liquidity providers, it is not possible to know what practices you are subject to. Do not trade with them. I do not care about claims regarding îECN,î îraw spreads,î or îdeep liquidityî talk; it is marketing fluff. Show your LPs. No declaration, no deposit. My broker uses liquidity providers with limited to no last look for example LMAX Group Note: This means I and my broker do not have direct conflicts when the tradeís risk is offset at this liquidity provider/execution venue. Nuance: LMAX Exchange operates an FCA-regulated MTF with a central limit order book (important), LMAX states its venue prohibits last look so there is no last second discretion from LMAX regarding if the order should executed on request or delayed at market (classic FX/CFD dealing desk behaviour). It is important to know that, regardless of how you gain exposure in markets, some≠one else must take the other side regardless of the product. Order routing to liquidity providers (LPs) mainly prevents direct conflicts of interest with your broker. LMAX may be the principal counterparty to your trades (the hedger), but a market maker would also be the initial counterparty when you deal on an exchange. LMAX manages its directional risk too, just as brokers do. The market is a continuous auction, liquidity is being provided to you, which provides a market for someone else, which pro≠vides a market for someone else, it is a constant shift of risk regardless of the product you trade. That is important to accept. Note that LMAXís broker entity may still be principal counterparty (the other side of the trade) to clients, even when execution references LMAX Exchange/MTF pricing. Before trading meaningful size in a CFD broker, you must verify your condi≠tions in five ways: 1. Execution policy via the broker 2. Liquidity provider review 3. Discussion over writing with account managers 4. Trade fill quality and latency information (from cTrader) or other serious platforms. 5. Only proceed if costs and conditions are superior to what you get on the underlying market and the trading of CFDs is legal in your jurisdiction. 2.2 CFDs are Serious Instruments CFDs are not toys, they are serious derivatives. CFDs are essentially the retail-facing total return swap contracts that institutions use but modelled for retail traders with shorter-term positioning; the payoffs and structuring really arenít too different. TRS and CFDs are both over the counter and not central to one exchange; the prices are based on the underlying market(s), like futures that are centralised on an exchange. The reason people scoff at CFDs is because they got îbannedî indirectly in 2010 in the USA to reduce risk-taking after the 2008-2009 financial crisis; this doesnít mean CFDs are not serious. It was not an outright ban, it was a consequence of not allowing products structured like CFDs being allowed for US retail clients. Institutions still use similar products. 3 Consider your other options before CFDs: Most traders should be operating using futures, instead of CFDs. It only makes sense to use CFDs in Europe, the UK and Asia if it saves money to operate under a CFD; for example, as I operate with GBP and only use limit orders for trade fills, CFDs are cheaper under certain regulated brokers. With professional client status, you get lower margins/higher leverage than futures firms can offer with overnight holding options. For example, if the spreads on a CFD firm intraday fluctuate on average between 1.5 and 2 (common) with zero commission and zero data fees, with good execution with limits and order slicing, it actually works out cheaper to trade the CFD instead of the underlying market. I also get slippage in my favour often on fills where it is volatile; the slippage on stops averages out close to ± a small amount when factoring in positive slippage from limit orders placed on a quality, regulated broker. Just like for futures, do not assume neutrality. This isnít foolproof; data must be collected from trade fills. If theyíre no good, change brokers. For people using market orders, especially on Gold / XAUUSD / GC CFDs, they rarely make sense; for oil / CL traders there can be inconsistencies in feeds, but the liquidity can be superior if using limits on some firms. If you choose to go with a firm with more lax regulation, e.g., Seychelles (which I do not recommend), make sure the same firm is regulated somewhere else that is serious too, e.g., in Europe, and they have a strong reputation. For an ordinary person it is best to operate on futures. What every trader must accept; Regardless of whether you trade futures, CFDs with LPs, or A-book/B-book, market orders are absorbed by another trader or market maker one way or another. It is not as simple as trading with a broker that has îECNî liquidity providers to avoid B-booking; it is trading with a broker that passes your trading size on to another venue or is open to A-booking you through their execution policy or via an account manager relationship request as a high depositor (at a lot of brokers, you are auto-assigned to the A book as a winning trader without asking most of the time anyway.) I have used CFDs for over 8 years. 3.1 Do not believe the no-dealing desk (NDD) or ECN/STP marketing. Many brokers advertise themselves to be NDD or ECN/STP with íRawí spreads, but your orders get sent to a íliquidity providerí owned by them, or it is internalised to look good. Do not fall for that. Screen your broker(s) and individual venues before dealing. íLiquidity providersí and other vague talk does not cut it. Look at the websites and research how they work. 3.1.1 If you want to see how your FX/CFD broker really fills your orders, this is what you need to do Different licence, different treatment; The execution policy of a UK-regulated broker is not guaranteed to be the same under a different licence. So you must check the execution policies of the jurisdiction you plan to work with. Additional Note: Most B-book brokers that brag about ílow spreadsí or íhigh execution speedsí are posturing to mislead; depending on the broker, as soon as you trade more than >5-20 lots per position or 1-2 futures contracts equivalent for CFDs, you will feel the slippage or delays with standard execution protocols on B-book models. What I look for in a serious CFD broker 1. A License. 2. A transparent execution policy 3. Transparent, responsive support. 4. Searchable venues with high-quality execution and a good reputation. 5. A serious third-party cloud platform offering that shows transparent market depth, e.g., cTrader. These connect feeds directly to LPs, which mitigates opportunity to tampering with trading conditions. Figure 2: Pepperstone UK: an A-book execution policy example. Futures Brokers: Tradovate AMP Futures Stock Brokers: Interactive Brokers Webull Examples of licensed CFD brokers and their execution models (Q1 2026 for context) UK-regulated CFD Brokers (If you qualify for professional) Pepperstone UK (A book, matched principal. Flow is offset with Liquidity Providers or with other clients) IG UK Forex Direct (FX) (A book, agent model, orders interact directly with the market on currencies only.) Tickmill UK (A book, matched principal.) Interactive Brokers (IBKR) (A book, agent model) CMC Markets UK -Hybrid (Mixed, risk is internalised but selectively based on risk management priorities etc.) There are many brokers I have not named because they do not meet my standards. I am not affiliated with any of these brokers. This is not a recommendation, either. A reputable lower margin CFD broker I have worked with (Regulated, but not by the FCA) IC Markets Global / Raw Trading LTD (Hybrid) I do not condone the use of IC Markets, but I have used them on and off for years. (2021-2025) -Written in 2025 for context. If you work with SPY or QQQ, you can get higher leverage/lower margins on ES or NQ futures. To be clear, I am not affiliated with the brokers listed; again, this is not financial advice or a recommendation. It is your responsibility to do due diligence. Some brokers have better costs than others with regard to specific instruments; for example, one may have low bid-ask spreads for FX but high spreads for indices and energies, and vice versa. Remember, serious traders only work with transparent, regulated firms. Remain firm with your standards FCA supervision and enforcement (UK). The Financial Conduct Authority has recently stated it will act against firms and individuals that fail to meet required standards in the CFD industry (2025) [4], with a focus on Consumer Duty [1] outcomes such as clear communication around fair value for retail clients. The FCA, across multiple sectors, has been consistent with its word, and it does follow through. The FCA has fined firms over >100,000,000 USD throughout 2025 alone for breaches. The FCA has been very clear about its standards for firms operating under its licence. The FCA has published a multi-firm review [3] setting out examples of good practice and areas for improvement in how CFD providers assess and provide good pricing and value (including spreads, commissions, and overnight funding charges). When dealing with risks, the FCA uses language such as îswift and assertive actionî.[2] They set the standard, and CFD firms must follow suit. I do not see other EU regulators issuing 7 to 9 figure fines, 5 to 6 figure fines are a slap on the wrist in the world of finance. If wronged under an FCA licence, a UK client with a legitimate claim and supporting evidence poses a serious risk to the brokerís finances and FCA licence. This firm got fined over £1m for misconduct surrounding client classification (not even price manipulation), which put the firm into financial hardship. The regulator (FCA) reduced the fine out of mercy. 4 CFD Broker Quality Assessment Who Is This Section for? This document is for people who are interested in being able to identify a good FX or CFDs broker. It is something that is good enough for a single read, but it must be revisited while reading policies to understand the brokerís offering and regulators. There are interesting things to learn. All traders could benefit, as we go in depth on how non exchange-traded markets and their participants operate. This is useful information to know, as it helps you understand more nuanced situations and learn to interpret the surface-level dress-up found in modern legal documents (not legal advice). 4.1 To Truly Understand A FX or CFD Firm You Need to Learn The Language I call it legal kung-fu, agreements written to outwit retail through wording. It is almost impressive how esoteric some of these documents have become. Sentient Traders will know FX and CFD Muay Thai in the next seven pages. The best thing about these policies is that, if you rarely switch brokers, you will not need to read them often, but when you do, you have to do it properly. Let us begin 5 The Basics: A book vs B book Classification Retail traders often oversimplify this, insisting that if a broker hedges your trade internally at any point, they are a B-book broker that absorbs all client losses and pays out wins as losses. That is not true. I understand the mindset, though, as many brokers are not transparent. Let us define it properly, starting with where traders get it wrong. B-booking is when a broker internalises risk against its clients for financial gain. Since most traders lose money, the broker can make extra profit by internalising their clientís directional exposure without offsetting it. They are directly making a market for their client to trade against. The reason this is worse than an exchange market maker is that they can decide whether to fill the trader or not, solely at their discretion when the order arrives, regardless of any liquidity they claim is streamed on their ìECNî or ìdeep liquidityî marketing. On an exchange, when a market maker has a quote (sell limit) at the best ask (the closest available price to buy) and someone submits a market buy, the trader is filled and the exchange handles it, the deal is done. With a B-book broker (dealing desk), they can automatically delay the order and decide not to fill you at all, solely in their own interests, which costs traders money through artificially manufactured slippage. The main issue with B-book brokers is not the conflict of interest itself, it is how they behave because that conflict exists. B-book brokers are financially incentivised for you to lose, which means 1. They are incentivised to give you poor trade execution and slow withdrawals (delays, slippage, higher costs). If they give your trades negative slippage, you lose more money on each transaction, which goes directly into their pocket. If they delay a withdrawal, a trader may cancel it, continue trading, and lose more. 2. They are incentivised for each client to lose as much money as possible in the shortest time possible. B-book brokers often offer extreme leverage and ìfree educationî to make people confident. 3. They are incentivised to offer casino-like promotions such as deposit bonuses, and they often deploy aggressive advertising that showcases large wins. Modern retail prop firms operate similarly to a B-book broker, but instead of deposits, they take an evaluation fee and monetise the relationship later. The main worry is counterparty risk and payout obligation failure with less regulated, less reputable firms. If a trader makes 2000% in gains (rare on average) and the firm does not let them withdraw, it is over. To prevent this, anyone dealing with FX or CFDs should use a well regulated firm with a solid execution policy. For many trading styles, futures would be cheaper when compared to CFDs anyway, but I explain how to decide later in the document (assuming you are British, European, or Asian). 5.1 îBest executionî is often misleading What most retail traders do not understand about ìbest executionî, as presented in legal writing and marketing, is that brokers can purposefully offer a poor setup, for example systematically poor execution and spreads, and still give you the ìbest executionî within this inferior environ≠ment. That is legal kung-fu, the best execution inside a cage they built. For most retail traders, this is like walking into a beautiful house (the brokerís brand) and being told they are about to be served the best tasting dish possible (the trading conditions), made with whatever ingredients and seasoning are in the kitchen. But when the trader peeks behind the curtain, all they have is cassava and water, no seasoning, no depth, no venues, nothing. They even offer to show them the ingredients first (the agreement), but the trader does not bother looking, they just tick the box. The trader sits there expecting a Michelin-star meal, and 45 minutes later they are handed a tasteless boiled cassava (illiquid CFDs dressed up with ìraw spreadsî and îdeep liquidityî marketing). The trader stares at the broker and asks, ìWhat is this, is there anything better?î And they shrug: ìI offered to show you the ingredients in the kitchen.î Sigh. Be careful of brokers who state that they act as an agent (passing your flow) to someone else (a liquidity provider), while also being the sole execution venue on the liquidity provider list. This is the gaslighting of the legal world, because it is the equivalent of saying someone is crazy for telling you that you ordered a parcel to your address, when you say you ordered it to a pickup point, only for that pickup point to deliver it automatically to your address. They say the parcel is not being delivered to them, but they are indirectly internalising it under a nice wrapper. Some broker agreements are arranged to look like A-book when they route the flow to themselves under the same or a different entity (owned by them). I have seen at least two CFD brokers do this without listing any other liquidity providers. Do your due diligence. Let us move on to the broker risk management profile we want, and how to avoid tasteless execution environments. 5.2 îA bookî brokers and learning the key agreement language Retail essentially view A book like a futures broker. The retail idea is íA bookí brokers are supposed to directly pass on client order flow (acting as an agent) elsewhere to get filled for example, at a liquidity provider (such as a EU multilateral trading facility) or execution venue (such as an exchange) regardless of the trading size(s) submitted. In the institutional world this is called Direct Market Access (DMA) What matters: 1. ìThis broker does not keep directional risk against clientsí best interestsî, meaning they do not trade against clients with the primary intent of holding that risk to absorb a clientís loss as their gain, and pay out wins as losses (B-book). 2. The brokerís primary revenue comes from spreads, commissions, or mark-ups, rather than from clients losing money. 3. The broker list of serious transparent venues which they íhedgeí íoffsetí or íexecuteí with. 3 out of 3 boxes must be checked. How our view differs: 1. The retail framing is too absolute, because not every broker that does not operate im≠mediately as an agent intends to absorb client losses as profit. Each execution policy is what reveals how the broker actually operates. 2. Because it is a guaranteed conflict of interest for dealing. 3. A broker can use many buzzwords and retail jargon to make traders feel safe, but without a transparent agreement and a reputable regulator, there is no way to know. A broker can advertise îNo dealing desk / STPî which means nothing if they route orders to a venue owned by them which has a dealing desk. To keep things reasonable what a trader should define as A book is one of these three scenarios: 1. The broker acts as an íagentí and sends orders directly to a liquidity provider typically in exchange for commission unless client order flow is paid for e.g., Robinhood is paid by citadel to hedge retail flow. Upside: Little execution conflict of interest with the broker. Commissions or PFOF are the typical source(s) of an agentís revenue. Unique Downsides: 1. Execution quality solely depends on the exchange or execution venue. 2. Typically, order size minimums are less flexible. For example, an agent brokerís mini≠mum lot size may be 0.1 (set by the venue), which restricts them from executing 3.64 lots but allows 3.6 lots. This leads to less accurate risk management, which adds up over time. 2. The broker acts as ímatched principalí where they act as your principal/counterparty only in the beginning just for your broker to handle the order briefly to find a hedge for your trade with a venue/LP to neutralise exposure. Upside: Reliable execution and low conflicts of interest, they collect spread markups, commissions or both. A Unique Downside: Execution may be unreliable during fast movements as the broker has more directional risk on execution if no hedge is found immediately. 3. The broker íback-to-backí or íriskless principalí hedging, offsetting or execution with≠out discretionary language such as íwe mayí Almost the same as matched principal; the language often overlaps. back-to-back is more immediate hedging practices, with minimal delays after your broker has filled you to find a counterparty via a liquidity provider or exchange. The difference here is more about the brokerís risk management preferences not your order execution qual≠ity. Matched principal is: ìwe aim to offset your risk when we have it on our sheetî, whereas back-to-back is îwe want to get rid of this directional risk immediately, as soon as it is on their balance sheet. íA Bookí Nuances: Some brokers say they act as ìmatched principalî or as an îagentî to fill your flow with a company affiliated with, or owned by, them. This is B-book in disguise, as risk is internalised (an indirect conflict of interest and another form of legal dress-up). Some ìA-bookî firms also offset or execute on a venue that lacks transparency. Regardless, due diligence should be done to identify the banks, liquidity providers, and multilateral trading facilities your orders are being sent to. The best thing is that you typically only need to do this once per broker that shows its venues and LPs, and most do not, because most are internalising client flow exclusively, as it is the most profitable model for unsuspecting traders, but that is not us. 5.2.1 Why most brokers do not operate under the agent model 1. The agent model is often far less profitable than B-booking because most traders lose. For example, if a trader loses $1,000 on a B-book broker, the broker gains approximately $1,000. A B-book brokerís gain is your loss, and vice versa. If the same trader makes the same loss on an A-book firm, the A-book firm only gains what was earned through spreads and commissions. This is why A book brokers typically want higher wealth, income or deposit amounts. In my experience they request information related to this on client sign up. 2. Some liquidity providers have minimum position sizes, which makes trading inflexible for many under a raw agent DMA model Wholesale liquidity providers and venues often require high minimum position value for example, >1m euros position value per trade (10 lots). Most Retail FX traders do not trade sizes this high. So, instead of a DMA agent model, it is sensible for brokers to fill their clientsí trades quickly with a low-latency execution engine that routes and fills trades at the appropriate venues automatically. A trader with a Ä5,000 (0.05 lots) position value will be filled under different venues and conditions than a person trading a Ä5,000,000 (50 lots) position value, which is fair, required for good fills, and facilitated by the broker as a part of the service. The example figures provided are based on EUR/USD. Trade example with a íMatched principalí or íback-to-backí execution broker that íhedgesí or íoffsetsí with venues, Quotes are derived directly from the LPs/Venues and there is a $7 commission (industry standard) per lot traded (round trip) ï Venue 1 (wholesale venue) 10 lots minimum size ï Venue 2 (liquid) 0.1 lots minimum size ï Venue 3 (Retail, standard) 0.01 minimum size A broker could get a request to buy 53.44 lots at market from a client (a trader), the trader gets filled quickly and the broker is the counterparty (for milliseconds) The entry is 1.10000, the value gained or lost per pip on this position is approx 534.4 USD. The commissions are $7 per lot equal to 0.7 pips in movement (1 lot = $10 per pip.) 7◊53.44=374.08 Net in costs (round trip) ï Venue 1 fills 50.00 ï Venue 2 fills 3.00 ï Venue 3 fills 0.44 During this time, the broker filled the trader at 1.10000 (VWAP aggregated from venues in the feed), but by the time the hedge was completed the price was 1.10001 (0.1 pips difference). On 53.44 lots, the broker therefore experiences $53.44 in slippage as a loss, and earns $374.08 from the trader in commissions. The broker also pays the venues $2 on average per lot, costing: 2 ◊ 53.44 = 106.88 USD So the brokerís net profit on this trade (once closed, before any overnight charges) was approx≠imately: 374.08 - 106.88 - 53.44 = 213.76 USD The trader is now filled on their desired size quickly and efficiently, and the broker collects its spread mark-ups and or commissions based on volume, so everyone is happy. The trader gets the desired directional exposure at a good price, and the broker gets paid. Nuance: The trader is typically the one who experiences slippage, whether in their favour or against them, on an A-book broker, but some brokers operate this way to achieve faster execution speeds, so it is important to clarify this with an example. There are key definitions you must understand before we proceed. 1. îAgentî: The brokerís job is to route your order to another party or venue for execution. Your ìrealî counterparty is typically the venue participant you trade against (or the liquidity provider), and the broker earns from commission, spread mark-ups, or fees for arranging the trade. 2. ìWe act as principalî / ìprincipal basisî. You are trading against the broker as the legal counterparty to your trade (not an agent routing your order). Your contract is with them, even if they hedge elsewhere. If they hedge or offset the risk elsewhere, it is fine, as they have a risk-free relationship with the client (zero directional exposure). If the trade goes in the clientís favour or against the client, the brokerís P&L remains flat. 3. ìSole counterpartyî / ìyou contract with usî Stronger version of the principal language. It confirms the firm is the only counterparty you face for the specific markets in the policy are under that entity (the broker). It does not mean that they do not hedge risk, it means at the very start they take on the directional risk against your trade at least initially. (Often milliseconds max if A book, but for the entire trade potentially if B book/hybrid, a key conflict of interest.) 4. ìSole execution venueî / ìwe are the execution venueî. The firm is stating that the place your order is executed is, in effect, their own venue. This often implies internal execution first, with external hedging being optional. Classic íB bookí language. 5. ìDealing deskî / ìmanual executionî / ìdealer interventionî. Human or discretionary handling may occur last look (delays, reject, reprice/requote). This is a major transparency and execution-risk flag for active traders. This is why people find B-book brokers undesirable. Modern brokers often hide requotes in slippages. 6. ìInternalisationî / ìexecute against our own bookî / ìmatch internallyî. Orders may be filled in-house instead of being passed to external venues or liquidity providers. This can be good for order fills because it means that one client who buys can match another client who is selling speeding up average execution time but this language becomes dangerous when combined with the presence of îwe may reject, cancel, voidî, ìat our discretionî and ìwe are the sole execution venueî because that often shows that they are internalising flow and fill your orders exclusively at their discretion. So if a broker believes it is against their best interests they could delay your execution causing slippage or reject it entirely. 7. ìLast lookî / ìTrade Optimisationî / ìTrade Optimizationî. The executing venue (or LP) can accept or reject after seeing the order. This can show up as selective delays, worse fills, or higher reject rates in fast markets. îTrade Optimisationî is the legal kung fu language that some brokers use as a substitute for last look execution to dodge retail scrutiny. 8. ìBookî / ìRisk Bookî / îBook Riskí / ìInventoryî. This refers to the directional exposure risk the broker holds. For example, if a broker has inventory short Ä100 million, they have a book imbalance of -100m. They need directional exposure equivalent to Ä100 million to get back to zero (imbalanced book), to ìrebalanceî, ìneutraliseî, or ìflattenî their book so they have no directional risk. 6 íA-bookí/íB-bookí Hybrid Execution The broker is your counterparty, but it offsets your exposure elsewhere in a more nuanced man≠ner which is why it deserves its own section. In this situation a trader works with the broker as principal (they are the legal counterparty), then the broker hedges a group of tradersí risk internally (B book side) or offsets that exposure with a venue, bank, multilateral trading facility, or liquidity provider (A book side). You never become a direct participant on the venue, the broker does but only at their discretion (inconsistent) These brokers aim to provide trade fills that are close to instant (fast confirmation and direc≠tional risk accumulation). A-book/B-book mixed together (hybrid) works like this: Some flow is hedged (A-booked) and some risk is actively taken on and held internally (B≠booked), often based on size (small or large sizes), client profile (winning or losing), or risk rules (maximum directional risk breach = hedging). These traders want the best of both worlds they want to absorb the losses or losing traders and generating money from îwinningî trader spread markups and commissions. This is a very common model in industry: This internalisation that happens on the B-book side is often referred to îwarehousingî but not explicitly in agreements. This is the stereotypical B-booking scenario. The conditions are usually not great, spread markups are typically high and their grandiose sponsorship deals is one of the first things they show off for credibility signalling. Hybrids are often sneaky, have vague hand-wavy policies, and many posture as íECNí brokers etc. and some show their venues pretending to be íSTPí or professional-grade. 6.1 Key Hybrid Tells 1. ìWe may hedgeî / îwe mayî / înot obliged to offsetî. This is the clearest hybrid tell. Optional hedging means some if not all client order flow can be internalised/warehoused (B-book) and some can be hedged (A-book), depending on the brokerís discretion or desires. ìwe may hedge in whole or in partî. Any use of the word îmayî must be reviewed rigorously for intent concealment through perceived flexibility (excuses). Or was it necessary to keep say îmayî. The îmay hedgingî talk also allows internal matching (internalisation) and reduces external hedging costs that they face for hedging clients who trade larger sizes. 2. ìWe are the sole execution venueî paired with vague statements regarding ìliquidity providersî. This combination often means that the trade is directly against them unless stated other≠wise explicitly that they offset internally or match with venue(s) or liquidity provider(s) and external hedging is at their discretion. This makes hybrid risk modelling very plau≠sible. In my experience, hybrid brokers often use vague, liquidity-focused language to give the impression of operating an A-book model. Posturing. 3. ìRisk managementî clauses describing selective hedging triggers. Look for language like: ìwe may hedge in whole or in partî, ìwe may reduce exposureî, ìrisk limitsî, ìmaximum exposureî, ìbreach triggersî, ìconcentration limitsî. These are the mechanisms that determine when the broker passes flow out versus keeps it. 4. îImplications: Different client different treatment undertonesî ìwe may apply different treatmentî type statements, ìbased on order size, type, market conditionsî. size-based routing is a common hybrid rule (small flow internalised, large size hedged) to avoid excess inventory risk. 5. ìMarket makerî / ìwe may act as market makerî combined with LP routing. Market making language indicates the broker can be taking risk internally. If the same document also describes routing to Venues/Liquidity Providers, they are operating in a hybrid capacity. Most hybrid agreements are not this clear. 6. Venue list where the broker (or an affiliate) is also listed as a venue/LP. If the only ìvenueî is the broker or an affiliated entity, and others are unnamed or vague, it is legal kung-fu as it can easily be internalisation dressed as clean A book agent-style routing. I have caught over two brokers doing this. This structure also allows brokers to claim they are ëA-bookí while actually internalising trades through a different entity, a shady practice that even confuses AI/LLMs (Q1 2026 for context). Naming and shaming these firms could invite legal liability, but you now have the knowl≠edge to screen them out when choosing a broker, and that is something they cannot take away from us. 6.1.1 Key nuance regarding hybrid classification: Pause. This does not mean B book/Hybrid Read agreements carefully, as brokers may use language such as ìwe hedge with these venuesî with ìoffset the risk internallyî combined might sound like a hybrid model, but what the broker is saying in this case is that they will either match buyers and sellers internally to offset the risk, or pass it on to a liquidity provider or venue, which is A-book, not hybrid in that sense, but close. ìOffset risk internallyî is the key here, because to íoffsetí is to reduce or eliminate inventory risk. If they trade against you, it increases their risk rather than offsetting it. If they have an imbalance in their books they can use the order in conjunction with the hedging venues to eliminate inventory risk. In plain words it means they will not be invested in negative outcomes for your trade. 6.2 Why We Avoid Spread Betting in the UK Spread betting has no capital gains tax but tends to have worse order fill outcomes; this means on average you are less likely to get the price you want. (You will get worse prices than you see when you click íbuyí and vice versa). For intraday traders this really matters, regardless of the size. You also have direct conflicts of interests with your broker when you participate with spread betting, as you are always interacting with the prices the broker provides to you directly, and they are your counterparty for 100% of your transaction. For example, CFD Liquidity Providers are not quoting Dow Jones in GBP, but they do quote in USD. Even if the deposit currency is in USD you are subject to the same inferior conditions. We have not seen a single broker spread-betting broker that has superior liquidity (Level 2 bid and ask size) when compared to a CFD on the same platform; it is often an inferior offer systemically. The broker is your direct counterparty if you are spread betting and the UK tax-free advantage will not mitigate the elevated costs you will pay, e.g., slippage when trading larger sizes as these costs have a recurring impact on return. The taxman strikes once, but a constant 50% increase in costs pulls down your returns and compounding potential. It gets worse the longer you trade. Costs add up. Figure 3: 3% risk over 100 trades, 50% win rate, and a 1:2 RRR after costs on futures, CFDs, etc, with a 20% tax rate applied at trade 80 (example rate for illustration purposes) versus spread betting with zero taxes. Each line takes the same trades. Figure 3 shows that the added conflict of interest and counterparty risk a trader takes on with spread betting produces the same or worse profits, even if the trading is îtax freeî. In this scenario, the CFD firm has an average cost equivalent to 2 points (spread and slippage), while the spread betting firm has an average cost equivalent to 3 points (generous). The minimum average stop size is 20 points, so the costs are around 10% on a commission free CFD. However, on a spread betting broker, the costs are higher, resulting in costs close to 15%, which drag down the returns (1:2 to 1:1.9RRR) not including additional potential slippage on stop losses. That is the difference a low discrepancy can make in day trading situations. You should optimise for conditions, such as margin requirements and available instruments, and execution quality, including order handling, first. Most spread betting firms offer inferior spreads and conditions relative to the underlying market which it is based on. (Spread bets are completely synthetic markets) On a serious, regulated CFD firm, I could buy or sell hundreds of contracts on either side of the order book with agent or matched-principal execution and receive lower-conflict fills. But with spread betting products, it is often at their sole discretion which level they will fill me at, which puts me at the mercy of their order-filling engine(s) and the individual brokerís inventory risk biases paired with no market depth. The conflict of interest: When you îTradeî with a Spread betting firm when you lose money it is their revenue and when you gain profits it is their loss that they pay out to you. I do not want direct conflicts of interest with my broker. The key lessons here are: When something is framed as advantageous or cheaper on the surface for brokers or prop firms, always look into why, the truth lives in the fine print and conditions presented. The savings from day trading tax avoidance with spread betting instruments are scraped away by higher bid-ask spreads, last-look execution, and other conflicts over dozens to hundreds of trades. In our situation this results in overall costs which are comparable to cheaper, more serious derivatives while paying regular tax rates without the benefit of being able to offset future gains with incurred losses (unused losses) with added conflicts of interest at the point of execution. Educators who blindly recommend spread betting over CFDs or futures for ìtax benefits,î es≠pecially in the UK, do not trade live or with a high enough size to understand the consequences of operating with a firm that employs such structures. UK brokers offer inferior leverage to retail clients (non-professional) when compared to regulated futures broker or CFD firms regulated in different jurisdictions, so there is little to no real incentive to do spread betting for most participants. As a result, we operate either in DMA markets, such as futures or CFDs on serious, regulated brokers. Let us move on to how to operate with over the counter market quotes properly. 7 How to use FX and CFDs with Rigour This section goes over the truth about forex & CFD Pricing, Arbitrage, and Scalping -With Examples 7.1 Ever wondered why a wick was longer on one broker compared to another on FX or CFDs? In less than 5 minutes youíll know how to deal with it. How FX is priced, why forex brokers donít like scalpers, why they donít allow arbitrage and, most importantly, why regulated ones donít manipulate your trading conditions. Okay, letís go! 8 Do regulated forex brokers manipulate prices? No serious regulator tolerates this. Fines would be issued, and licences will be revoked. This is an offshore/unregulated broker issue. This is true for unregulated offshore brokers, and there are a lot of scammy unregulated FX brokers, but for regulated retail FX brokers, all pricing techniques must be declared and fair for clients. Regulated brokers were caught doing this in the late 2000s and 2010s and got destroyed for it. FXCM was banned completely from operating in the USA after a CFTC/NFA investigation re≠vealed excess conflicts of interest and key failures to disclose their dealing desk protocols. Firms get fined even for malfunctions; firm regulatory oversight like NFA (US) or FCA (UK/Europe) ensures this. Figure 4: Simple quote aggregation example from multiple liquidity providers To be clear, before we get into this, the same things I state also apply to CFDs like XAU/USD and US30 Brokers making a market is not the same as a market maker algorithm on an exchange. Forex brokers want to accept buy and sell flow, collecting spreads and commissions, if any, whilst maintaining net-neutral market risk. brokers aggregate prices from liquidity providers like ECNs and prime brokers to offset any risk there. This is also a reason why prices differ for FX and CFDs on brokers. I will address these nuances before continuing. Even honest, regulated brokers can disadvantage retail traders via wider spread markups, but they must be upfront and not quote with intent to harm or deceive; quote discrimination is also not allowed, and re-quotes via dealing desk brokers must be transparent, but those things can cost traders without being the criminal ìstop huntsî. While itís still a declared conflict of interest with the client, itís not the same as active predatory practices and quoting strategies. Basic FX broker example: EUR/USD 0.1 avg bid-ask spread clients ($7.5 comms per lot) $7.5 Comms * 2250 lots = $16875 earnings from comms $10 (P/L per pip per lot) * 0.1 spread * 2250 lots = $2250 earnings from spreads 1k lots long; clients: 1.2k short; same avg. price = broker goes long 200 lots at market to correct the imbalance. the reason is so they limit or neutralise market exposure. Serious FX brokers donít care if you lose; they care if you trade. Serious regulated retail brokers hedge imbalanced inventory at the market. The reason FX brokers donít like scalpers is because it makes it more costly for them to manage inventory risk (they have to rebalance more at market, eroding profit potential). Arbitrage trading is adverse selection for FX and CFD brokers, which is why CFD brokers do not allow it. Adverse selection occurs when a trader acts faster based on having better pricing information than the broker/MM to transact at a price that quickly becomes unfavourable. this can happen when traders hit lagging prices before the broker can re-hedge, When a trader does this success≠fully, the broker or market maker is highly likely to lose money from facilitating the transaction; this is why it is not permitted by most brokers. This is another core reason why many brokers offering OTC markets do not like scalping, as scalpers (especially bots) can accidentally become toxic flow in adverse selection scenarios. A simple FX market buy example: 200 lots are bought at market with lower spreads (sometimes negative) and commissions than offered to retail, and the broker pockets the difference. ex. $2000 offset cost ($16875+$2250) ≠$2000 = $17125 net earnings for the broker on this occasion. In terms of how retail FX is priced, these îmanipulationsî of ex 0.2 pips, for example, are just discrepancies between the feeds because of their pricing engine; retail FX brokers with serious regulation, like the FCA, arenít out to get clients. Thatís retail narrative. The reality is much less entertaining. For example, a broker uses 5 îLiquidity Providersî to price EUR/USD as seen in Figure 4. What causes the discrepancies? Is there a difference in feeds? Each Liquidity Provider prices forex differently for multiple neutral reasons. LPs can adjust prices by small amounts, similar to how MMs might adjust quotes on futures markets, but that is only to manage inventory risk or for other functional purposes. Serious îA-bookî firms do not do this to take out retail clients. Also, brokers donít always equalise the priority of LPs for their pricing calculations. Itís not always even. LPs with the best offers get pushed first. It constantly changes based on market depth and the conditions of the LPs. The best bid (shown on the chart) can be from LP1; the best bid in the same minute could be from LP3, constantly changing and flickering. This is how aggregation works and how pricing efficiency is achieved for spot FX. However, this is what causes small intraday discrepancies and differences between broker price feeds. Figure 5: Over time, these differences often mean-revert and tend to average close to zero, so the impact is usually negligible. That is why price action across brokers often looks very similar, even though small deviations still occur in the very short term. FX quotes are more comparable (efficient), but quotes for index CFDs are more variable, as the tick sizes differ, the LPs hedge differently, and the price mark-ups are different. For example, one LP could quote 45,000 for Dow Jones CFDs, while another LP quotes 45,010. The width between the quotes reduces the compatibility with aggregation. This is why youíll see the wick high and wick lows differ from broker to broker; it is not îmanipulationsî on a broker with a serious license in 99.9% of cases. Many forex traders complain about getting stopped out or not getting filled where they should be. The way to deal with discrepancies is to measure recent formations. 8.1 How can I get filled where I want consistently with these price feed inconsistencies? Retail FX Limit orders A trader wants to buy at a 5m resistance level breakout formed 1.5 hours ago (18 bars ago) using a Forex Com chart on TradingView but trades on a prop firmís price feed. The way to increase the probability of being filled at the exact price on the chart is to measure the distance of that level on the Forex com chart compared to the latest 5m bar high; letís say it was 10.0 pips lower on the TradingView chart. Figure 6: The Price Range or Date and price range tool on TradingView The shift for the magnet tool helps with these procedures a lot. The next step is to get a recently formed value, e.g., a 5m bar high. 1.17323 on the prop firm. The trader must subtract the distance and then add the maximum anticipated spread to get the limit order price to get filled on the prop firm at the same time as the TradingView feed. Formula (in this case) RecentBrokerHigh-TradingViewDistance+MaximumAnticipatedSpread The trader could know that itís abnormal for the intraday spread on his broker to exceed 0.3, so he could do (1.17323 -0.00100 + 0.00003) to get a Limit order price of 1.17227. Most of the time, the in-examples like this (RecentBrokerHigh-TradingViewDistance) will be equal to the same price of the level on your brokerís chart, but this method ensures youíll get filled at the correct price. Then the trader needs to go on the broker feed (the prop firm), get the bar high value ex. 1.17323 and subtract the distance and add the maximum anticipated spread, e.g., . +0.3 Figure 7: UK Regulated Broker example of factoring in bid-ask spreads Similar tactics can be done for making sure you get stopped out at the correct place and get your profits filled at the same price on your broker, assuming you port trades from one feed to another. For example, for a running short position you could measure the distance on the chart to update the accurate place. where you should get taken out; this prevents premature fills out of your trades. For example, a trader could be short EUR/USD at 1.17000 with a stop loss originally at 1.17000 + 10 pips + spread (because shorts fill on the ask price). the stop is at 1.17014, but because the measurement is now 0.1 pips off the trader must increase it to 1.17015. It seems small in hindsight, but those few times the price misses your stop by less than a pip, you could get taken out anyway if this isnít taken into account. 9 The Sentient Way & References As a trader using this approach, your goal is to make a market at favourable levels by tactically providing liquidity to enter and exit and by taking liquidity when conditions are unfavourable to get out. We aim to absorb/fade aggressive orders whether the market is DMA (e.g. futures or stocks) or OTC (e.g., CFDs or Swaps) 1. Superior entry prices compared to market orders 2. Superior order queuing Vs when your entry is equal to the best bid/ask For CFD Markets. We get rewards either way. We position ourselves to benefit by 1. Designing strategies that get accurate, superior entry prices compared to market orders 2. Mitigating vulnerabilities to delays and liquidity provider discrepancies by using limit orders exclusively. (fill at the requested price or better logic). 3. Scaling to size with order splitting techniques (Highest trade size ever: a 106 index futures contract size equivalent) 4. Get positive slippage from providing liquidity instead of absorbing negative slippage from taking liquidity from a synthetic book. 5. Operating with CFD firms that are regulated and show transparent market depth. Context (Important): On serious CFD firms, limit orders are filled at the requested price or better logic. Super rare non-fills can be handled manually by us with market execution resulting in small slippages, but non fill scenarios also happen on the underlying asset too e.g., futures but due to order queuing not synthetic liquidity nuances previously described. Everything is included within our costs before deployment. Order queuing: In this context, order queueing means that the first person to place a limit order on an exchangeís order book is the first to have their order filled. (First in, first out). We desire entries only where recent liquidity anomalies or inefficiencies are present, and want our profits to be taken where past inefficiencies are present. Limit in, limit out, and limit in, stop out for losing traders. End note: If CFDs turn out to be the lower-cost option for you, you must consider the discrepancies and manage them, if you are using market orders for execution, futures will almost always be cheaper. You should only prioritise CFDs if the costs work out cheaper, or it fits your sizing constraints better; it does for us, but it is your informed decision to make. I am happy to see you at the end. Remember. Choosing a good broker is like choosing the right location for a physical business. You have done the reading, but now you must apply it. Look at three different brokersí execution policies and classify them. Using CTRL+F will speed this up. Thanks for reading -Ron References [1] Implementing the Consumer Duty in the Contracts For Difference (CFD) Portfolio 2023 [2] Portfolio letter FCA strategy contracts difference 2024 [3] Multi-firm review of contracts for difference providersí provision of price and value 2025 [4] FCA review finds CFD providers may be failing to deliver fair value to consumers 2025 [5] FCA Handbook COBS 11 Dealing and managing [6] Simple definitions provided by European Securities and Markets Authority, mentions matched principal which aligns with FCAís definition. [7] Pepperstone UKís Matched Principal Type Execution Policy (Example purposes only.) [8] Algorithmic Trading and DMA: An introduction to direct access trading strategies by Barry Johnson Figure 8: A Visual Example Of A CFD Broker That Is Not Serious: An Opaque Bureaucratic Process Instead of Straight Answers.

Prop Firms

Modern Prop Firms arenít what you think. Sentient Trading Society (Public) Traditional Industry Prop Firms/Desks To keep this short, real prop industry firms look for talent and employ based on merit. The collected revenue is from real profit splits on real capital; traders are paid a base salary and paid a performance bonus or an agreed-upon percentage of gains. These firms had restrictions to protect investor capital. Rare since dodd-frank (limiting institution speculation) passed in 2010 after the 2008-9 financial crisis. Bucket shops/Scouting Prop Firms These ìfirmsî are pay-to-play and benefit from fees instead of trader success; these are not real traditional prop firms. It is just marketing. They collect revenue from failures (evaluation fees), most issue demo/simulated accounts to traders, and payouts are losses on their balance sheet. This model has direct conflicts against the traderís interest. Not all scouting prop firms operate this way but most do. These firms have restrictions to amplify evaluation failure rates generating more revenue. Itís a loss to most prop firms if you win The average retail trader is fooled by the multiple step process They ask the trader to make a 100% return on 10% risk (110%+ if you include trading costs), reset then a 50% return on 10% risk (55%+ including costs) then continued demo (live phase) trading up to weeks before having your first payout You are forced to operate with smaller risk % numbers which inhibits compounding; this is on purpose too. Opportunity Cost: The ideal optimistic scenario would be 8 weeks of demo trading before any financial reward (2.5% return per week avg after costs -quite high, most traders wonít be able to sustain it) On a live account you wouldíve multiplied your capital by your first payout. If you used the same risk profiling. Ex if 0.5% on FTMO 5% live. The more a trader chases and loses to prop, the less their potential edge will be when they get îfundedî itís a well-crafted seductive illusion. You bet $620 to get $10000 risk (80% after payout . $8000 . Income taxes . even less). Additional Input (analogy): Ali has simulated prop firms with similar models to FTMO and it wouldnít be profitable to issue real capital assuming every traderís p&l follows a random walk, The nature of retail firms is like a bucket shop/arcades. Any base salary even minimum wage would relinquish the main conflicts of interest in modern retail îprop firmsî because itís not pay to play itís youíre employed and get a bonus on top of your salary as incentive to perform. The Distinction Real firms earn based on profit -running costs ex. Salaries Retail firms earn based on revenue collected from challenges -payouts (Your gain is there loss) For example an MLM (Multi-level marketing company) could charge people upfront to close clients in sales for commissions (profit split) Or a normal company that has neutral or best interest at heart could employ sales people with a salary and pay a sales bonus This is the difference between real prop firms vs retail prop firms options. Prop Firm Mechanics Sentient Trading Society Ali It is important that you read my and Ronís previous documents to fully grasp this one. If you skip over them, then you are doing yourself an injustice and you will not understand some of the things explained in this one. 1 What Do I Know? Being obsessed with beating prop firms, I had developed a simulation of firms with simulated traders to see how firms maximise their income from traders. This gave vital insight into how traders can handle their firm accounts in order to make money from the firms. Prop firms do not share their performance metrics. This led me to making simulations from analysing countless firms and finding information through online sources and some contacts I have. Through these, coupled with my knowledge in building simulations, I formed one for prop firms. This is a proprietary simulation as we wanted to simulate our own prop firm model to make our own firm but then we decided not to open one for various reasons. Just to show that this was indeed done, here are some screenshots to show the develop≠ment of the model itself. Below is just to show proof of some of the work I have done. The rest of the document is text-based. There are many more pages of such content, but this is a short example, see Fig. (1). Also, a figure showing the resulting plots of simulations will be shown soon when I get time to run them! Figure 1: A few governing equations which describe a basic prop firm and the traders within it, there are a few pages more of these to describe the entirety of the firm and trader behaviour. In laymanís terms, it is a fact that the behaviour of traders follows a probability distribution, which allowed for this simulation to be realistic. For the people interested: These basic equations that I had derived were all coded and simulated with the assumption of the conservation of the number of customers. This means that a firm would conserve the property of having 1000 customers. When one leaves, another joins instantly. This gave simulation results that were more intuitively understood when I was analysing this a few years back. 2 Traders Do Not Understand Most traders do not understand what they are signing up for when joining a prop firm. Why do I say this? Because they do not understand the underlying costs of their strategy, the rules of the firm, and how to actually make money from firms to the best of their ability. If you do not understand how these firms operate and what their goals are, then you are walking into a trap. This document will break down what firms aim for, what they do to fulfil this aim, and how you can dodge these tactics to make money properly. 3 Types of Firms There are various types of prop firm such as scouting prop firms (retail) or actual industry prop firms (with a salary). Industry prop firm opportunity is rare since Dodd-frank (enacted in 2010). The difference: Industry Prop Firms/Desks To keep this short, real prop industry firms look for talent and employ based on merit. Bucket shops/scouting prop firms These ìfirmsî are pay-to-play and benefit from fees instead of trader success these are not real traditional prop firms. It is just marketing. We will discuss the main type of prop firm that traders recognise which are the regular online based prop firm such as Alpha Capital Group, FundedNext, FTMO and so on. 4 Conflict of Interest The firm makes most of its profits from failed evaluation stages. Their best customer is one which fails multiple evaluations and never gets a payout. From my simulations I could see that this was indeed the case. As there is a sheer discrepancy between the number of winning and losing traders, it is best to take advantage of the 99% of losers as they can consistently lose challenges and make the firms money. Firms are incentivised to have you believe that it is easy to pass and make money when the structure is actually designed for you to fail. But do not let this be the deterrent. Despite them having such rules, there are ways to still make money. We do however recommend trading live money if you have a good amount to trade, but for people who do not, firms are the next option. We can see that conflicts of interest show up here. The fees are fixed and the chances of success are capped structurally. There are amplified restrictions of maximum all time drawdown, daily drawdown, consistency rules, high profit targets, and trailing drawdown rules to even further maximise failure rates. Firms can even widen spreads, delay executions, and worst of all they restrict payouts. This is why care is required. 5 Poor Account Retention and Why It Happens Even if they manage to pass a challenge and become funded, a lot of traders still end up losing their funded accounts within weeks. The reason for this is simple. The strategy which allowed for them to pass the evaluation stage had high variance and was highly profitable within the period of evaluation and then its profitability dipped when it came to live. A strategy, of course, is judged based on its average performance. Average performance could be medium which means it performs highly profitable half the time and has low profitability the other half. This averages out to medium as I just stated. The firm knows that this is the case. They are so certain of this that firms have started to remove or increase the time limit required to pass evaluation. They know traders are so likely to fail that even given all of the time in the world, and they will still fail. But because we are capable traders, this can be taken advantage of. Most traders will go back after the blow up of their live account thinking they can get it back again and make profit. They pull their wallets out and pay for a new challenge. The cycle continues. 6 Understand the Mechanics (Important) I am going to describe some generic requirements that firms tend to have just to refresh your mind. Evaluation Phase 1 ï Profit target of 10% with no time limit. ï Do not exceed max all time drawdown say 10%. ï Do not exceed the daily loss limit of 5%. ï 5 Minimum trading days. ï Consistency or lot size rules may apply. Evaluation Phase 2 ï Profit target of 5% with no time limit. ï Same drawdown and daily loss rules. ï Consistency or lot size rules may apply. Funded Stage ï Biweekly payouts possible. ï 80/20 profit split. ï Weekend holding limits. No major news trading. ï Any rule violation means a terminated account. ï Account may still be simulated despite being ìfundedî. Here is the important thing to note and get in your mind (pulled from the Black Tesseract as you probably saw in my video.) ìA proprietary firm has options of account sizes at their respective costs. In order to extend the time alive within the market, there must be a system in place which will maximise the chance of large returns through continuous passing and failing of accounts. A pass of a firm account must subsequently be followed by a month of profits such that the money earned after passing exceeds the initial investment (the investment is the cost of the prop firm account). How much it should exceed is simply a value that is greater than the initial investment. Failing a firm account though requires maximum drawdown to be exceeded. It is now obvious that it is not a 50/50 chance of winning and failing. A win [in the context of a firm] is a return that is greater than the initial investment (if the amount that is greater than the initial investment covers the investment plus another account of the same size -then this is the principal balance) Definitions: Prop Firm Principal: How much it takes to buy a single prop firm, evaluation size that you are running if a 100k evaluations each eval costs $500, so your principal is $500. Principal Balance: How much you have accumulated from prop firm payouts net. Principal Multiples: Net Prop Income ˜ Prop Eval Cost A fail is a loss which exceeds the maximum drawdown. In order to win, first you must pass two stages of a challenge, normally 10% followed by 5%, then you must profit enough to accumulate the principal balance -this is a three stage process. A fail is a one stage process. So immediately the benefits of using a firm is not as great as it seems. This is why a strategy must be built such that within good market conditions it can make enough to reach profits that add to the principal balance immediately after the two stage passing sequence is complete.î In short, passing requires you to pass evaluation 1, evaluation 2, make enough money from the live stage (without exceeding drawdown) to buy an extra account in case of blow up. A Three stage process. Failing requires you to only break one rule, such as the maximum drawdown limit. A One stage process. Most prop firms are asymmetric, which skews probabilities against the trader; this is how retail prop firm evaluations are designed. As a result, traders should build and test multiple strategies to get positive expectancies in uncorrelated ways that fit within firm constraints. Variance plays a key role. Running three uncorrelated, robust strategies on different prop accounts with the intent of achieving pay≠outs (just as an example) is far more efficient than relying on a single strategy to produce a positive payoff. STS Prop Firm Guardrails & Examples. When you receive $5000 net in payouts and the evaluation fee is $500 you have a principal balance of $5000 (10x principal) the higher the multiplier the more edge you have over the firm. When in excess of 10◊, you can scale your account size in ways that align with your goals without breaching your predefined rules regarding principal management. Until then, any prop firm evaluation size-ups are premature and can set you up to give all your payouts back to the firm, which is a common trap. For example, a trader can make $2k with a $50k funded account, immediately spend it on $100k or $200k evaluations, and end up giving everything back if they fail. Many people who receive payouts suffer in this exact way and nobody hears their pain. The correct approach is to keep stacking different strategies on the same evaluation size (e.g., $50k evaluations for a $300 fee in this example) 1. Have at least 4◊ the principal saved up for a given account size (for firms exclusively) before attempting evaluations. > 10◊ is less aggressive (safer). The more, the better, do not rely on a single strategy. 2. Until you have = 10◊ your principal in net payout profit (3k in this example) do not escalate with higher evaluation sizes. 3. = 20◊ your principal (6k profit in this example) if you run >3 strategies simultaneously and want to scale up. The retention of new money requires discipline, do not break it. Pre-define your scaling plans, avoid planning specific financial goals and focus solely on principal accumulation. There is no room for folding. Gamify it. Example: An STS trader has tested three separate systems for prop firms that he is ready to run. This is what his strategy portfolio looks like: 1. Strategy 1 Example: A mean reversion system US30 or MYM Futures (5m OHLC) 1:3.5 Avg RRR (profitable outcome) and -1R (losing outcome) after costs 35% winrate. Surface Level Stats: Impressive performance, profitability exceeds a 1:2 riskñreward ratio with a 50% win rate after costs. 2-hour trading window. 2. Strategy 2 Example: A trend following system for Gold (hourly OHLC) 1:2.8 Avg RRR (profitable outcome) and -1R (losing outcome) after costs 40% winrate, Surface Level Stats:: Great performance. 8-hour trading window. 3. Strategy 3 Example: An aggressive reversal system for Natural Gas (15m OHLC) 4-hour trading window. Avg RRR 7.5R (profitable outcome) and -1R (losing outcome after costs with a 23% winrate. Surface-level stats: Amazing returns but high drawdowns. Strategy 3 trails the stop systematically by using predefined rules. The maximum target realised is a prop firm pass (none in the backtest). When the stop is hit, the position exits at its original loss or with the accumulated gain. The Danger [3] A less stable P&L, More return volatility (tail risk). Prop firms with îconsistency rulesî in payout rules will not accept such a strategy. The Benefit [3] Positions can be closed out automatically once the account equity passes the firmís requirements, potentially passing prop accounts prematurely and benefit beyond what the backtestís realised P&L would capture realistically (a small maximum favourable excursion advantage). This takes advantage of the prop firmís profit-target mechanic, as in a backtest the position can reverse instead of realising the profit required for the firm to pass. Why run two or more strategies instead of one? (a) Each strategy fits within the traderís constraints. All the strategies are uncorrelated and use different mechanisms for entries, tar≠gets, and risk management techniques. Each system is rooted in a basis, and most importantly, none of the strategies are curve-fitted to work. Additionally, each strategy executes at different times, which is a bonus, not a requirement. (b) Risk Isolation All are executed on different prop accounts, with each producing its own P&L path. No single strategyís performance can interfere with the others. It may seem like a lot of trading multiple strategies or instruments at first, but as long as you fit it within your constraints, it is entirely possible to manage while working or studying. We did. It is important to keep your execution realistic. 7 Solution In order to beat firms we must understand a few things. I will list them out and their descriptions one at a time. Read carefully. 1. A trading strategy has costs, which we have discussed in the other documents (VITAL). This means that to pass a 10% evaluation stage, you need to make the equivalent of say 12.5% return due to a cost of 20%. Nobody is exempt from this, It is important not to ignore this 2. Your strategy needs to win so consistently that it can last for a weeks on end without exceeding the drawdown limits of the firm. Your drawdown limits can be controlled by calculating your risk off of the maximum drawdown your strategy has witnessed in a backtest. This means that your strategy is effectively coupled to the firm account. A maximum drawdown of your strategy (which indicates blow up) will match the maximum drawdown of the firm. This simulates the conditions in a live account. 3. You must use a systematic logic-based strategy. This should take advantage of some≠thing real, like discussed within our Strategy Engineering Volumes. Your strategy should avoid intuition-based discretion and other poorly defined conditions, as that introduces human errors. Firm accounts are extremely sensitive to strategy fluc≠tuations, which links to the example I had earlier of a strategy that goes from highly profitable to low profitability with more chance of high drawdown. A systematic strat≠egy can maximise your chance of success. When I say maximise, I mean it. 4. You must analyse and backtest your strategy perfectly. Understand what your strate≠gies capabilities are and how it is expected to perform in the firms before strategy edge decay sets in. 5. Read the other documents and watch our videos to understand how psychology, costs, and so on tie into this document. 6. VITAL, THE GOAL: You must understand that it is fine to blow up accounts in prop firms as long as you on average make more money than the initial investment and withdraw it. This means that you will have a lot of principal balance to use to buy even more accounts. This is like ìexponentialî growth of the number of accounts you have. Doing this ensures that you can continue to trade accounts with the strategies you develop over time as the previous one faces edge decay. 7.1 A Simplified STS Principal Balance Accumulation Example; $3000 starting principal balance (Money at risk), 50k Accounts (Moderate, not aggressive). 10% target, 10% maximum drawdown and an 80% profit split ($300 per 50k evaluation) $300 x 3 strategies. Scenario: 1. 1 evaluation passes phase 1, blows up in phase 2 -$300 2. 1 evaluation passes phase 1, phase 2 and earns two payouts totalling $4500 and a $250 refund. 3. 1 evaluation blows up on phase 1 -$300 (2100 ◊ 80%) + (2400 ◊ 80%) - (300 ◊ 2) = $3000 Gain Changes: The principal balance now stands at $6000 (3000+3000) equivalent to 20 evaluations at a 50k account size. It is also worth noting that the trader has also accumulated 10% in drawdown as risk to deploy (5k USD or 4k USD (incl. -20% after payout)). Adjustments: In this situation, HE would continue trading the profitable account with the same setup and withdraw as often as possible to accumulate principal. If the trader wishes to expand, the amount funded ($50k in this case) should be chosen as the evaluation size, rather than jump≠ing prematurely to higher amounts, e.g., multiple $100k evaluations across multiple firms. Remember, this is how traders give all their gains back to the prop firm. Once enough prin≠cipal is accumulated, size can be increased proportionately. It is important not to avoid prematurely increasing prop firm exposure, as it can set us up to be forced to decrease it later (hinders recovery). This trader decides to move $3000 to his main bank account so he can operate with prop firms whilst only risking his profits. He develops two other strategies and buys 2◊ 50k prop firm evaluations for $500 altogether this is less than 10% of his principal per account (3000˜250) once he re-accumulates the withdrawn principal his plan is to step up to 100k evaluations+ and when his income tax bracket becomes absurd in his jurisdiction he plans to trade live instead with STS-Style Capital Partitions (six figures accumulated post-tax). When you trade with a prop firms, payouts are interpreted as îincomeî not îcapital gainsî resulting in higher taxes in many situations. This trader decides to move $3000 to his main bank account so he can operate with prop firms whilst only risking his accumulated profits. He develops two other strategies and buys $2◊ $50k prop firm evaluations for $600 altogether. This is only 10% of his principal per account $3000˜300). His plan, once he re-accumulates the withdrawn principal, is to step up to $100k evaluations and scale. His tax plans: When the income tax bracket becomes absurd in his jurisdiction e.g., =40%, he plans to trade live instead with STS-Style Capital Partitions (six figures accumulated post-tax) and pay capital gains tax or pay taxes under a corporation (whichever is cheaper) to pay a calmer rate of =20% whilst also getting >75% of his gains in cash after taxes instead of less than >50% net with the payout fee (-20% and taxes -40%) he gets tax advice to align to pay the least amount of tax possible, legally from a licensed tax advisor to avoid penalisation. 8 Opportunity Cost and Experience There is huge opportunity cost with firms. If you are to pass multiple stages and risk losing an entire account, then you may as well trade live money and always be able to use it to make profits. Unfortunately, like I said earlier, you cannot just do this with a few hundred dollars. You need a good amount of trading balance to make worthwhile returns. This is the reason why we do not engage with retail firms any more. For example, Ron had to pay around effectively 37% on his profits (UK Income tax and NI). There are ways to reduce this with limited companies but HMRC/IRS will make you pay up when you operate as a prop trader, they pay you out as a ìcontractorî. You do not pay capital gains on your profits, you pay income tax regardless of if you reduce it with a limited company, or operate as a sole trader, then the taxes on live trading is lower (subject to capital gains tax). In the UK, you can save over 50% in taxes if you realise your gains on your own live account. This is a statement, not tax advice. Another thing that we had to endure was the sheer amount of time it took to pass some accounts and collect principal balance. Sometimes, you have to trade a sim/demo account for up to months to get any financial reward. If you have an edge you need to take advantage of it now, not later! Deploy with multiple accounts of smaller size with different uncorrelated strategies instead of relying on one. Take advantage of variance, do not waste your time. If you have 2-3 profitable systems do not settle for íthe best oneí, run them all. You only need a pass and payout from one and you have principle which gives you space to churn and ífreeí risk from the prop firmís book. Strategies for passing firms can be more aggressive (your choice) but strategies for collecting payouts should be have stable performance. 9 Profiling a Firm -Ron Prop Firm Due Diligence and Red Flags: Is this prop firm serious? Are they suitable for me? Some of the most important questions to ask yourself, especially with the recent adjustments and rules. This is about both risk assessment (will the firm follow through) and costs (such as spreads and comms). ï Adverse Payout History Patterns (Public unresolved payout rejections) Some traders break rules and feel entitled it is always important to do due diligence to see if the trader was legitimately wronged by the prop firm before coming to any conclusion. ï Rule Change Behaviour; Some prop firms have capped the percentage risk taken on each position which they enforce on ífundedí accounts but not the evaluation which catch you out when it is payout time some will use stop loss distance to measure or the losing amount confusing and confining the traderís potential risk protocols. For example a swing trader could have 10% maximum drawdown, realise 5% in profit over multiple trades risking 0.8% per position. Good. But his trader wants to increase his risk to 1.2% to take advantage of compounding to get higher payouts (his strategy, not ours) he cannot do that because the prop firm says maximum risk per trade is 1%, if breached no payout. This needs to be known before any payment. ï Trade execution complaints (with evidence) When we worked with a prop firm years ago we had felt the slippage. Eventually it had got shut down by the CFTC (They did payout but with purposefully poor CFD execution). ï Server Issues and Resolution. I cannot name and shame for legal reasons but a futures íprop firmí in Q4 2025 was quite shocking with poor handling leading to blowups, rattled support and little to no compensation. Repeated server issues and freezes, skip them until they clear up their act. ï Couple risk to the firmís constraints You must adjust risk percentages based on your strategyís observed maximum peak≠to-trough drawdown so your system is deployed to survive within the firmís drawdown limits. Trading low risk in drawdown for months is wasted time. If your strategy fails. It is done. 9.1 Conflicts of Interest Part 2 Remember, some prop firms have rules against scaling in and frame it as îone-sided betsî or use other language against the traderís best interest. If you choose to work with a prop firm, make sure you pick one that does not have such rules when deploying strategies with scaling-in behaviour. Remember in professional trading environments, scaling is a standard practice; do not let retail prop firms manipulate you. If your strategy requires it, trade elsewhere. There are other conflicting rules we will list too. The reason we dislike them is because profitable strategies have uneven profit distributions, and these rules directly punish effective systems. There are other rules I will list too. Keywords to spot this conflict (Q2, 2025+) Risk limit / one-sided bet / one-sided bet / limits / overexposure Risk limits, e.g., 1% max risk per market or market type Keywords: Risk, limit Real consequences & Responses Swing traders who use larger stop sizes cannot risk as high a percentage on their trades, making prop firms with these rules less worthwhile. Suitable Responses Go ahead if your strategy rules are unaffected by the limits, or do not trade with that firm. The key is awareness. Do not over-adjust a strategy to fit into a prop firmís mould; if you do, the adjustments add noise to your potential results. Adverse Payout Rules: Indirect language used: On demand / On-demand / consistency In Action: îYour best trading day must not account for more than 40% of your total profit.î Instant consequence: if your strategy makes $4000 on its best day, you cannot withdraw until your profit reaches $10,000. If your strategy ends up making even more, e.g., $6000 in a day from multiple successful trades, you they force you to wait until >$15,000 is realised. The better you do, the longer they make you wait. They will often claim it is to prove consistency, but it is actually to increase the chance that you do not receive a payout (which is their loss). Return distributions over larger samples are more stable, but over small samples, there is more potential for unevenness. The Numbers 0.4 / 40% requires your payout to be at least 2.5x your best trading dayís realised profit, some prop firms have rules that are much worse such as 0.2 / 20% (5x) 1 ˜ Consistency% ◊ Bestday ' sP &L = Minimum amount withdrawable. 1 ˜ 40% ◊ $4, 000 = $10, 000 net profit is required to withdraw Here are some visuals. 40% Rule, 4000 best day, 10,100 profit net (39.6%, 2.525x), barely eligible for a payout as seen in Figure 2. If the best day was 4200, it would be rejected (example of interference). Figure 2: Illustrating how these rules actively work against strategies (Moderate, 40%) 20% 4000 best day, 10,100 profit net, withdrawal rejected; the prop firm requires 20,000 (minimum) to get a payout (5x), as seen in Figure 3. Figure 3: Illustrating how these rules actively work against strategies (Extreme, 20%) This level of conflict of interest is not acceptable; a lot of firms bait with cheaper prices, but in finance, when something is cheaper, there is almost always a reason, and you must find it before interaction. Avoid rules that are designed to punish natural variance; think in adversity and return distributions first. Focus on firms with biweekly payouts. Think about the many ways how your profitable strategies can produce returns, how it draws down and what succeeding can look like. Figure 4: 5 Simulated 1:2RRR 45% Winrate Equity Curves Input your backtest results into an online îequity curve simulatorî with a below ten lines (paths) to see the variance, click generate many times over 20 trades and get used to seeing it play out. The black line represents the average. Here is a link: Equity Curve Simulator Sections to look: Payout rules, policies Language used examples: îYour best trading day must not account for more than 40% of your total profit.î îThis rule states that no single trading day should contribute more than 40% of the total generated profits.î îThis rule promotes consistent trading practicesî îConsistency Ruleî The real consequences: Withdrawal rejection or no option to withdraw fairly earned principal. Increased vulnerability to drawdowns, edge decay and performance drag. Prop firm evaluation failure (if the rules apply to evaluation too). Unneeded psychological pressure to deviate because of perceived loss of agency. Automated strategies Many allow custom automated strategies, but you must double-check if you decided to code your strategy. Sections to look: Rules, practices, policies. Real consequences: Breaches or withdrawal rejection with no option to withdraw fairly earned principal. If prop firms do not fit your strategy or actively work against their traders, do not work with them. One strategy can work brilliantly with one prop firm and suffer on another. Check the leverage, trading costs and their reputation rigorously; seeing their spreads is not enough. We see plenty of futures prop firms with complicated rules, such as consistency rules and complex rule structures for payouts. Personally, we would not touch them. We prefer objective, clearly defined rules, set out in their support materials, FAQs, and legal documents. It is important to state that we prioritise live environments over prop firm accounts. We do not condone the use of retail prop firms. If you choose to interact with these firms, it should be done properly, but as of 2026, we are against their use. It is your choice if you want to interact with them; see these write-ups as insight into how we have beaten them in the past. We will not name and shame prop firms, as we do not want any potential legal liability (lawyerís advice), and we will not recommend, as their business practices do not align with our principles (conflicts of interest). than 0.6% per trade, with intermittent trailing (b) The prop firm reduced his leverage accross stops and mechanically scaling into favourable all of his prop firm accounts until he gets two positions (maximum of two). This is a moder-payouts or îtwo successful performance ate risk profile, not gambling. periodsî. Our main issue: Many prop firms deny payouts or lower available leverage at their sole discretion, which is a potent conflict of interest that is much rarer in regulated live environments. We have seen an overwhelming amount of repeated, clear evidence from multiple retail ìprop firmsî across multiple jurisdictions, especially in 2025 and early 2026. Making five figures or more from prop firms feels closer to fantasy than reality for most high-expectancy strategy builds. Now we move on to manipulation... Figure 6: This is not to help you. They frame it as something to help you when it is to help them dodge your payout. Remember, this firm only provides demo accounts, even at the îfundedî stage. These restrictions remain in place until after two payouts, which are scheduled accordingly. At 40% consistency, he would need to generate an additional 20k to reach his first payout using 1:10 leverage on a 100k account. Given that his strategy operates on a low timeframe, this could take several months. There is a potential conflict of interest, and the demo account operates with close to 0.5% risk per trade. The terms include unnecassary restrictive conditions in a simulated environment. This prop firm is based in a western european jurisdiction with strong-willed regulators present and they are still taking advantage of clients. People must remember that prop firms are not regulated. 10 Choosing a Firm If you choose to work with one. Ask these basic questions to yourself: 1. Are there daily drawdown rules what are they? Trailing drawdowns do not seem like a big deal until a random drawdown catches you off guard. Asymmetric targets or amplified trading costs can drag your potential returns lower. 2. What are the true costs of resetting, scaling, and withdrawing? This must be factored in to prop firm principal planning and management. 3. Are there consistency or hold time rules? Is the strategy compatible or a fantasty? 4. If I decide to accept a consistency rule, are my returns stable enough to qualify for payouts? For example, if a strategy averages 5% per month with 0.5% risk per position, and the best day after costs is 2% over a decent backtesting sample, this would be acceptable under a 40% consistency rule, since 5% ◊40%=2%. If the return distribution shows more extreme values the strategy is too vulnerable. 5. What are the legalities involved? Is the prop firm legal to deal with in your country and what are their policies? 6. How can I handle my taxes when it works out? If you do it incorrectly penalties will be issued. 7. What are the withdrawal rules like? Getting paid is the most important part. 8. Most importantly, am I capable of making systematic well back tested strategies, and am I able to do so such that when the current one fails, I have a new one ready? If the answer to most of these is unclear or unfavourable, do not trade with them. 10.1 How to Approach Prop Firms Seriously If you must Read our article first: https://medium.com/insiderfinance/how-to-approach-prop-firms≠seriously-007b9bb9cf8b This is an inside joke from something Ron said a while back, and we still laugh about it. He said... Thanks for reading, Sentient Trading Society.

What to do after you finish the material

What to Do After You Finish the Material Your Path Forward The Sentient Trading Society 1 Start Building Re-read, work, research, test, repeat. Create your own trading ideas; research and develop your own strategies. Once you have 2ñ3 sound strategies that align with the materialís rigour and your own reasoning frameworks, run each one in isolation and evaluate them independently. One strategy per account. When you hit uncertainty, go back to the relevant sections and research the gap. Using AI for research is fine, but do not outsource your thinking to an LLM. 2 How do I stay sharp and avoid slipping back? Numerous readers report the feeling of seeing flaws they cannot unsee, which is good, as that is required for growth. However, a few have also expressed concern about drifting into flawed reasoning again after seeing a more desirable, structured way of reasoning, since our framework does not allow flaws to slip through. This applies to many demanding fields, as well as trading. Some procrastinate, freeze, or dissociate. Letting complacency dominate now is how you allow all the time committed to learning to go to waste. Remain firm. Remaining îdialled inî becomes natural At the end of it, what we want is for each readerís reasoning to align with ours, so that you emulate our thought processes subconsciously. Readers can leverage it not just in trading but in regular life as well. The more you revisit it, the harder it is to break, and profitability emerges naturally. You can feel the BS, you can even briefly fall into it, but when one tenth of your head is in the pool, you spring back up to breathe in the sentience. Flawed reasoning becomes like stumbling upon a foul smell and changing course to avoid it because, once you sense it, it becomes legitimately unbearable. You have the tools now, it is time to sharpen them through rereading and, when ready, use them. We recommend revisiting the materials at least once every month when you are operating profitably. Iím not going to speak for all people, but when I say this, Iím sure a lot will relate to being alone in a room, obsessing over trading, and seeing it as the way out. I relate to that, but 1 you must not ruminate over failure, as it is the number one accelerator of self-destructive habits such as self-criticism, which only inhibits your chances of success if you allow it to become your baseline. 3 Remember the Operating Standard Reason well and execute precisely. Before deploying any system, you must confirm with yourself: 1. If prompted, I could explain the mechanism(s) my strategy depends on objectively in plain language. (Anti-circular reasoning) 2. I have written entry, exit, invalidation, and risk rules. (Anti-noise) 3. I have executed a test plan with results provided by clear samples. (Rigour) (a) îI have modelled costs harshly but realistically.î 4. I have a sample size spanning at least 150 trades, and the backtest length is 12 months or longer (in-sample). Examples of application: (a) Your test spans = 12 months and includes 170 trades; this is acceptable. (b) Your test is 12 months long and includes 50 trades; you should continue testing until the trade count reaches = 150. (c) Your test is 9 months long and includes 160 trades. You must continue testing until the strategy data reaches at least 12 months. Note: Your out-of-sample data must fit the exact same conditions, and the out-of-sample data must retain at least 20% of the in-sample returns. So altogether, you need = 2 years of data, including a minimum of 300 trades per strategy. There can be adjustments to the sample for optimisation, but they are strategy-dependent. As a basic example, the period from January 2023 to June 2024 is the testing data for îTechnical Model Aî. June 2024 to January 2026 is the out-of-sample data, and the trader stress tests the model from January to March 2022 alongside other isolated periods. 5. I have objective conditions for pause, continuation, and strategy retirement. For example: (a) Maximum peak to trough drawdown in % Or (b) A fundamental change which invalidates the strategies dependency (System Class 2) If one of these are missing, the strategy is not ready for deployment. . Note for out-of-sample testing: There are multiple approaches for testing, and applying trading costs, this is the dependent on what your strategy seeks to gain from. For example, if you are looking at trading an oil crisis or a war policy type of shift mechan≠ically, in real time, then you could test what is happening closer to now and run stress tests and out-of-sample on similar events in the past. For systems that are closer to being purely technical, you can opt for an approach where you are using past data for testing, and then you do testing in the future outside of that window to see if the edge collapses. Remember, a key part is critical thinking. The materials exist for you to get a good idea of what needs to be done and how to think. You need to structure things logically, which aligns with your strategy. If a part has friction, you must restructure it. Do not feel trapped or it will inhibit your progress. Citation from Strategy Engineering Volume 1 PDF 2. îIf I do not get an in-sample expectancy / expected value reading beyond 0.2R, I stop testing, as I would believe the strategy is underfitted or something else has gone wrong. If a small, logic-based optimisation that complements the systemís logic cannot save it, I test on a few other markets that may be compatible. If that does not work, I toss the strategy away. If the strategy does not retain at least 20% of the in-sample return on out-of-sample tests, I also discard it, as this would reveal fragility such as poor logical integrity, overdependence on a specific market condition, or overfitting.î Note: These are heuristics; more robust sampling insights are shared with mentees. Do not fold. Good data or no deployment. 3.1 I want to look into other entry methods (not candlesticks). Is this applicable to the STS framework? Can I use non-candlestick entries with the STS framework? Of course, it is. It depends on data quality and the friction involved in real-time testing and execution. Candlesticks are a choice for simplicity. It is about what works for you and what you prefer within the logical framework that STS provides. If you choose to use something else, ask yourself this first ï Does this tool describe market price? Does it represent price or other market data? (Important.) Must be yes. ï Is it difficult for you to test your idea with this tool? ï Is it difficult to execute your strategy precisely with it in real time? These are the key friction points. We need good data and low friction. Do not settle for less. If the answers are negative or uncertain, you can continue to explore and research the tool(s), but do not deploy live until you can test and execute properly. If you cannot find a way to do that, that is fine; do not force it. Instead, revert to something feasible or direct your attention elsewhere. We must not get distracted by grandiose trading tools, frameworks, or marketing posture. If someone insists that one trading tool is the only way to trade efficiently, recognise the nonsense and depart. What are the best trading tools? It is best to avoid thinking like this; it is better to know which tools to avoid in trading. That is the perspective that saves time and money. ìBestî will always be biased and subjective. There are many valid interpretations, and many can be used to produce comparable results. Some tools are objectively baseless, such as technical analysis-based price cycles, and should be avoided at all costs. Valid tools will respect the questioning sequence provided above. 3.2 Keep Execution Data Private Operational Transparency and Data Security It is our policy to forbid the broadcast of our execution data on any monitoring websites. By broadcasting your trading behaviour through APIs (past or present), there is a significant risk of compromising your intellectual property, increased last-look risk with many non-DMA market environments provided by derivatives such as CFDs or FX or in rarer cases, order flow front-running. Professional quants never use their prime broker accounts for posting publicly on any re≠tail forum. We follow that rule very carefully ourselves. In the same way, we demand you to adhere to the same principle. You need to preserve your advantage when building consistency in your trading activity. The retail industry thrives off creating a culture of oversharing information, which serves only to compromise your psychological discipline and expose your execution models. The only exception to this strict policy of complete secrecy is your professional career. Should there be a need to produce an audited track record to land a position or get funded by an institution, doing that would be a mandatory professional move on your part. Outside of that specific institutional requirement, your trading data must remain entirely sealed. 3.3 What to do with your first payouts or withdrawal? If the gains were structured and intentional, treat yourself. Keep it quiet; your circle does not need to know. Mine didnít know until it was obvious. Celebrate modestly. Treat yourself to a nice experience, but aim to spend less than 20% of it. The less, the better. You only do this once; the point is to make it register what hard work can do. Our first expenditure was meeting in London in late 2021, buying designer SPX embroidered shoes, and eating at a Marco Pierre White steak restaurant in London. Nothing extreme, but enjoyable and memorable. Once you do it, you will never forget what this process can bring to the table; it is both leisure and a reinforcement exercise. If the gains were unintentional: Do not spend anything; keep what youíve accumulated as a safety net to grow in a structured and intentional way later. If you have made money by accident, withdraw it immediately. Avoid rewarding luck. 3.4 If You Feel Like You Are Drifting Return to the material, isolate the gap, and correct it. Do not add complexity to compensate for weak reasoning. Remember: îconsistencyî is not essential for successful trading. Regime changes and edge decay actively work against it. For competent traders like yourself, waiting for ëconsistencyí often wastes time. Markets are like waves; we know they change constantly. Waiting for the perfect wave every day will leave you stuck on the shore. Closing Statements Perseverance only pays on a sentient path. The market does not reward effort alone. You must show the market clarity, discipline, and continuous willingness to adapt when shown a different hand. You must Build your strategies carefully, test them honestly and operate professionally. And donít forget When you succeed, let us know. Reach out or post your progress to motivate others. Your progress will push the next reader forward. The Sentient Trading Society End of the Main STS Materials

Academic References and Supplementary Material

Serious Market Literature (Books) The STS Booklist 1 Introduction These Books Tell It How It Is To learn how market makers and markets really work. These books are written by people who are highly respected and who have overseen the establishment of electronic trading and its innovations. 2 Read in This Order 1. Trading and Exchanges: Market Microstructure for Practitioners Larry Harris 2. Market Microstructure Theory Maureen OíHara For Further Refinement 3. Algorithmic Trading and DMA: An Introduction to Direct Access Trading Strategies Barry Johnson After That Learn about market auctions. A great paper for that would be Continuous Auctions and Insider Trading by Albert S. Kyle. It is a long pathway to the truth, but even skimming will teach you what most people do not understand about markets in minutes to hours. Easily some of the best books going. 1 Favourite Additional Reading Opportunities Sentient Trading Society 1 Introduction Sentient Trading Society Strategy Design Model -Academic and Institutional Studies + Book References This document is reserved for those who are curious to see the backing behind our process and thinking or want deep trading knowledge. Pause and read when something piques your interest. Judge by citation. This is the literature and research that actually matters. This collaborative post by me and Ali combines institutional grade research with carefully selected citations. It will give you grounded insights into how mar≠kets work, market efficiencies, trading psychology, reasoning, the declining effectiveness of public strategies (alpha decay), and much more. This section lists literature and research that we have found useful when thinking about markets, edge, and trading psychology. It is not exhaustive, but it brings together many of the sources referenced throughout the material. 1.1 Use Guidelines: Whilst market academia is insightful, it typically doesnít include key nuances of real markets; it focuses on theory, so to find the truth, you have to look at multiple peer-reviewed publications and books to answer individual questions. An example of this would be price discovery and the concept of ífair value,í which are often described in a utopian fashion. Theories often assume that markets are efficient or speak as if they are, when in reality, they are not. Itís important to use these revelations as baselines for understanding markets rather than îhow to tradeî and an example of how I and Ali have done this is by modelling deviations from market efficiency to create market edges. I and Ali have read dozens of papers and have collectively skimmed hundreds of papers over the years to reach our conclusions. 2 Random Walk and Market Efficiency Eugene Fama -The Behavior of Stock-Market Prices Key Part: ìBy contrast the theory of random walks says that the future path of the price level of a security is no more predictable than the path of a series of cumulated random numbers. In statistical terms 1 the theory says that successive price changes are independent, identically distributed random variables. Most simply this implies that the series of price changes has no memory, that is, the past cannot be used to predict the future in any meaningful way.î Note: We are not suggesting the market is 100% efficient or random. We referenced this to show how randomness in a market is not good for your bottom line. The more efficient or random a market is, the harder it is to trade profitably. Eugene Fama -Random Walks in Stock Market Prices Key Takeaway: If the random walk theory is an accurate description of reality, then the various ìtechnicalî or ìchartistî procedures for predicting stock prices are completely without value. Burton Malkiel -A Random Walk Down Wall Street Key Lines: ìA random walk is one in which future steps or directions cannot be predicted on the basis of past history. When the term is applied to the stock market, it means that short-run changes in stock prices are unpredictable.î ìMathematicians call a sequence of numbers produced by a random process (such as those on our simulated stock chart) a random walk. The next move on the chart is completely unpredictable on the basis of what has happened before.î The core lesson of the random walk theory is that you cannot predict future market price movements by studying historical data if the market is 100% random. Alpha and Market Edge Decay Julien Penasse -Understanding Alpha Decay Read here. Highlights that alpha (edge over market) tends to diminish. Alpha decay is generally a non stationary phenomenon and inconsistent. Julien leverages studied anomalies for credibility. Key Parts: ìBecause alpha decay is generally a non-stationary phenomenon, asset pricing tests that impose stationarity may lead to biased inference. I illustrate the importance of alpha decay using the most commonly studied anomalies in the asset pricing literature.î ìAlpha decay refers to the reduction in abnormal expected returns (relative to an asset pricing model) in response to an anomaly becoming widely known among market participants.î Does Academic Research Destroy Stock Return Predictability? -Journal of Finance, R. David McLean Read here. Key Takeaway: ìPortfolio returns are 26% lower out-of-sample and 58% lower post-publication. The out-of≠sample decline is an upper bound estimate of data mining effects. We estimate a 32% (58% ≠26%) lower return from publication-informed trading.î On the Effect of Alpha Decay and Transaction Costs on the Multi-Period Optimal Trading Strategy by Chutian Ma and Paul Smith (2025): ìTo simulate alpha decay, we consider a case where not only the present value of a signal, but also past values, have predictive power.î High Frequency Market Making: The Role of Speed -Yacine A®it-Sahalia, Mehmet Sa.glam Read here. 4 Intraday Seasonality and Session-Based Rules Admati and Pfleiderer -A Theory of Intraday Patterns Key Parts: Documents intraday volume and volatility U-shape across NYSE hours. Table 1 shows how volume and volatility vary through NYSE hours. Essentially, the Volume and volatility are typically highest at the major market opens and closes (open and close shocks cause liquidity inefficiencies), and lowest in the middle of the day (more efficient). 5 Mean Reversion Versus Trending Characterisation Grant, Wolf, and Yu (2005) -Intraday Mean-Reversion After Open Shocks Key Lines: ìThis paper gives a long-term assessment of intraday price reversals in the US stock index fu≠tures market following large price changes at the market open... The significance of intraday price reversals is sharply reduced when gross trading results are adjusted by a bid-ask proxy for transactions costs.î Rama Cont -Empirical Properties of Asset Returns Table 1 lists ìVolatility Clusteringî and ìGain/Loss Asymmetry,î i.e., mean reversion characteristics for major indices. 6 Backtesting Bailey, L¥opez de Prado, and Zhu -Pseudo-Mathematics and Financial Charlatanism Key Lines: ìWe prove that high simulated performance is easily achievable after backtesting a relatively small number of alternative strategy configurations, a practice we denote ëbacktest overfitting.îí Read here. 7 Order Flow, Microstructure, and Market Mechanics Albert S. Kyle -Continuous Auctions and Insider Trading Key Lines: ìMarket liquidity encompasses a number of transactional properties of markets: tightness, depth, and resiliency. A liquid market has bid and asked prices always available, with small spreads.î Kyle shows that imbalances between buyers and sellers are what make prices move. Larry Harris -Trading and Exchange: Market Microstructure for Practitioners Parts 1.5, 3.10, and 3.11 are well written and direct. It provides accessible insight into order flow. Maureen OíHara -Market Microstructure Theory and High Frequency Market Microstructure Covers how liquidity and order flow mechanics underpin price formation. 8 How CFDs and DMA Markets Work How CFDs Work: IG CFD Agreement Algorithmic Trading and DMA: An Introduction to Direct Access Trading Strategies -Barry Johnson Chapters 2-4 cover market access, ECNs, LPs, and MMs in depth. 9 Trader Psychology Lo, Repin, and Steenbarger -Fear and Greed in Financial Markets Documents that day traders experiencing drawdowns suffer measurable stress responses. PubMed -Quantifying the Cost of Decision Fatigue: ìMaking decisions over extended periods is cognitively taxing and can lead to decision fatigue, linked to reduced performance.î Read more. ESMA 2018 Report: Shows that 74-89% of retail accounts lose money, with average losses ranging from Ä1,600 to Ä29,000. ESMA Report. Berisha and Asllanaj -The Role of Financial Instruments in Reducing Exchange Rate Risk: Explains how Total Return Swaps (TRS) and CFDs allow exposure to price changes without owning the underlying asset. 10 Additional Citations Lo and MacKinlay -Data-Snooping Biases in Tests of Financial Asset Pricing Models Read here. Hurst (1951): The original Hurst exponent paper on long-term storage in hydrology, adapted to finance by Mandelbrot. Jim Simons Interview: Watch here. Data Snooping (Common in Multi-Timeframe Analysis): Quant.Fish article. Turtle Trading Edge and Alpha Decay: Read here. Note: Turtle strategy returns declined after media exposure and structural electronic trading changes.