--- title: "Machine Learning for Alpha | RupeeCase Learn" description: "What machine learning can and can't do in systematic equity investing. Signal-to-noise, effective sample size, which ML tools actually work. By Tanmay Kurtkoti." source_url: "https://www.rupeecase.com/learn/path-5/module-5-3-machine-learning-for-alpha" --- [Path 05 3 of 5](/learn/path-5/)Machine Learning for Alpha[Next 5.4](/learn/path-5/module-5-4-alternative-data-india) Every second fund manager now says AI or machine learning. Most of them mean a gradient boosting model on financial ratios, relabelled as ML driven [alpha](/learn/glossary#alpha). A few mean something more. The problem is that equity return prediction is one of the hardest places to apply ML, for reasons that have nothing to do with how clever the model is. Machine learning is part of how RupeeCase research is done, so this is not a sceptic's essay. It is an honest map of where ML helps in systematic equity investing, where it does not, and why that line matters more than any headline. 1. **Untouched hold out**A final period nobody looked at during research. 2. **Untouched hold out**A final period nobody looked at during research. 3. **Untouched hold out**A final period nobody looked at during research. 4. **Untouched hold out**A final period nobody looked at during research. 5. **Untouched hold out**A final period nobody looked at during research. Most ML projects that fail in production fail at step 2. Shuffle financial data at random and the model trains on information from the future. ## The core problem: signal to noise ratio In image recognition a good model can be right more than 99% of the time. In equity return prediction, a model that calls the direction of monthly stock returns correctly 55% of the time is genuinely strong. The reason is noise. The predictable part of a stock's return is a small fraction of its total variance. 0.05 An information coefficient (correlation between forecast and next period return) many quant researchers would be happy with −0.02 Lag 1 autocorrelation of Nifty 50 daily returns, 7 Oct 2024 to 9 Oct 2026, computed from NSE closes: yesterday says almost nothing about today That has a sharp consequence. Models built for high signal problems, deep networks with millions of parameters, are hugely over specified for equity returns. With so little signal, extra capacity goes into memorising noise. ### Worked example: why a small edge is enough The fundamental law of active management, from Grinold, links skill and breadth: information ratio ≈ IC × √(independent bets per year). Suppose your model has an IC of 0.05, barely visible on any single stock. Rank 200 stocks a month and treat about 200 of the 2,400 rankings a year as independent. IR ≈ 0.05 × √200 ≈ 0.71. That is a strong information ratio from a forecast that is wrong almost half the time. Now the trap. Double the breadth by trading weekly and, on paper, IR rises to about 1.0. In practice weekly trading quadruples turnover, and at 0.1% STT on each side of a delivery trade plus charges, the extra costs can easily exceed the extra edge. Breadth is only free on a spreadsheet. The size of moves repeats; their direction does not 1. Absolute returns, lag 20 days**0.12** 2. Absolute returns, lag 20 days**0.12** 3. Absolute returns, lag 20 days**0.12** 4. Absolute returns, lag 20 days**0.12** 5. Absolute returns, lag 20 days**0.12** [Autocorrelation](/learn/glossary#autocorrelation) of Nifty 50 daily log returns, 7 Oct 2024 to 9 Oct 2026, computed by RupeeCase from NSE closes. ## The sample size problem Take the [Nifty 500](/learn/glossary#nifty-500) and 20 years of monthly returns: 500 × 12 × 20 = 120,000 observations. It sounds like plenty. It is not, because the observations are not independent. Stocks move together on the same day, positions persist from month to month, and the market changes regime. The **effective sample size** is a small fraction of the raw count. In Indian data the problem is sharper, because clean, survivorship free histories for the broader market are shorter than in the US. So complex models are almost always over fitted on financial data. The right answer is not to avoid ML. It is to use regularised, parsimonious models and validate them on data they never touched. ## What actually works in practice Searching rule designs Works well Using ML and large scale search to choose which signals rank a universe, how many names to hold and how often to rebalance, then testing each choice out of sample. The output is a simple, fixed rule a human can read, not a model that changes itself every day. Feature engineering Works well Combining raw inputs into robust composites, detecting interactions between factors, extracting structured information from filings. ML as a feature builder, not as the return predictor. Tree ensembles Works with care Random forests and gradient boosted trees handle non linear effects and give feature importance. Gu, Kelly and Xiu (2020) found tree models and shallow neural networks beat linear models on a large US sample. Keep them shallow and regularised on Indian data. Text from filings Conditional Under SEBI's listing rules, listed companies publish earnings call transcripts on their websites within five working days of the call. Tone and guidance consistency can be extracted. The data is messy, often scanned PDFs, and edges decay as more people use them. Regime classification Conditional Clustering or hidden Markov models label markets as calm, trending or stressed. Regimes are clear in hindsight and noisy in real time, and a false switch is expensive. A simple trailing average rule often does as well. Deep sequence models on prices Mostly hype Networks that read raw price sequences need large data and strong signal; monthly equity returns have neither. Where deep learning does help is very high frequency order book data, not monthly factor strategies. Reinforcement learning Not practical yet An active research area for portfolio decisions. Training needs simulations that rarely match real markets, and policies are fragile when conditions shift. Worth following, not deploying. **The honest bottom line:** a well built rule with an economic reason, careful costs and disciplined [rebalancing](/learn/glossary#rebalancing) beats a complex model most of the time. ML earns its place in search, feature construction and text processing. It never replaces the question of *why* a signal should keep working. ## A minimal honest experiment If you want to try ML on Indian stocks yourself, start small and strict. Take NSE daily prices for a point in time universe, say the Nifty 200 as it stood on each date. Build five features with a reason behind them: 12 month return skipping the last month, 1 month return, 6 month [volatility](/learn/glossary#volatility), distance from the 52 week high, and average traded value. Label each stock by whether its next month return beat the universe median. Train a simple logistic regression and a shallow tree model on 2015 to 2020, tune on 2021 and 2022, and look at 2023 onwards exactly once. Rank the stocks each month, hold the top 20 equal weighted, and deduct STT, stamp duty, exchange charges and GST on every trade. If the shallow tree does not beat the logistic regression after costs on the final period, keep the simpler model. Most of the time, that is the answer. ## Leakage: the silent killer Leakage means information from the future sneaking into training. It produces spectacular back tests and live disasters. The common Indian forms: * **Overlapping labels.** Predict 21 day forward returns every day, and neighbouring labels share 20 days of the same future. A random train test split puts near copies of test rows in training. Purge the overlap and add a gap, an embargo, between training and test periods. * **Restated fundamentals.** Using today's version of a company's past results instead of what was filed on the day. Use data as of the filing date, which for quarterly results is within 45 days of quarter end under SEBI's listing rules, or 60 days for the year end. * **Survivorship.** Training only on stocks that are listed today drops every company that was delisted or went bust. The model learns that falling stocks recover. * **Index membership.** Using today's Nifty 500 list for a 2015 back test. Membership must be point in time; NSE reviews its indices twice a year. **Myth**A more complex model finds more alpha. **Fact**With weak signals and short, overlapping data, extra complexity mostly fits noise. Simple, regularised models usually hold up better on data they have not seen. ## Common mistake: reporting the best of many models The common mistake is not a bad model. It is a good looking one picked from hundreds. Every hyper parameter tried, every feature set tested, every random seed is another draw. Report the best and you have a multiple testing problem wearing an ML costume. Module 5.1 has the maths. The defence is bookkeeping. Write down every configuration you try. Judge the final choice on a hold out period you have never looked at. And discount the result for the number of tries, because the market will. ## How this shows up in RupeeCase Machine learning and RupeeCase Machine learning is used in RupeeCase research to search and optimise the model behind each strategy: which signals rank the universe, how many names to hold and how often to rebalance, tested out of sample. The rule that ships is then frozen and published, so the live strategy is a fixed, readable algorithm, not a black box that changes on its own. Every rebalance is published, and every return figure is labelled model, back tested. Read the methodology on any strategy page, for example [Large Midcap](/strategies/large-midcap), and the rest of the record on the [track record page](/track-record). ## Questions readers ask Can AI predict the stock market?Not the way headlines suggest. Models can find small, persistent tilts across many stocks, which is valuable at scale. No model reliably predicts the direction of a single stock or the index tomorrow, and anyone showing very high accuracy on Indian returns is almost certainly showing an over fitted back test. Is a machine learning fund better than a factor fund?Not by label. Judge any strategy on its rules, its out of sample evidence, its costs and its drawdowns. ML is a research method; it does not make a strategy safer or better by itself. Why freeze a model instead of letting it learn every day?A model that refits itself constantly is hard to audit and easy to over fit to recent noise. A frozen rule behaves the same way in the live market as in its tests, and every decision it makes can be checked against the published rule. What does Machine Learning for Alpha cover?ML has genuine applications in systematic equity investing, and genuine limitations that the industry consistently understates. This module separates what actually works from what looks impressive in a pitch deck. It works through The core problem: signal to noise ratio, The sample size problem, What actually works in practice, A minimal honest experiment and more. It is module 5.3 of Path 5, Advanced Quant Methods, about 8 minutes of reading, free. TTry it ## Machine Learning for Alpha, in numbers. Years per split given total history Your numbers, the module's formula, nothing stored. Calculator ### Train / Validation / Test Split Helper For time series, split must be chronological. Train on the oldest data, validate on the middle, test on the most recent. Never shuffle. Total history (years)Train share (%)Validation share (%) AApply it on RupeeCase ## 5Paisa Focused Alpha 5 Five stocks, three-week rotation. The most concentrated card on the lineup. An index built on the rules this path teaches, concentrated momentum book, highest octane on the lineup, held stock by stock in your own demat. Risk band: Very High. [See the strategy](/strategies/5paisa-focused-alpha-5)[Sign up for RupeeCase](https://invest.rupeecase.com/signup?strategy=5paisa-focused-alpha-5) [CAGR54.1%Sharpe1.60Max DD−24.2%Minimum22000Model, back tested, as of 27 Apr 2026. Not client returns, not a promise.](/strategies/5paisa-focused-alpha-5) QQuick check ## 10 questions, 7 to pass. Pick an answer and the explanation appears. Your result stays on this device and counts towards the Path 05 certificate. CCheck yourself ## Path 05 so far. Your quick check results on this device, stop by stop. 1. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) 2. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) 3. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) 4. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) 5. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) 6. [TPath test, 24 questionsNot yet](/learn/path-5/module-5-5-out-of-sample-strategies#pathTestSection) Explore next ## Put the module to work. * [Strategy5Paisa Focused Alpha 5 strategyNifty Alpha 50, 5 stocks, every 3 weeks](/strategies/5paisa-focused-alpha-5) * [Free module 5.4Alternative Data in Indian Markets9 min read](/learn/path-5/module-5-4-alternative-data-india) * [Learn guideSystematic Investing in India: RulesA pillar guide across the paths](/learn/systematic-investing) Community ## Join my WhatsApp group. Scan the code, fillFill one short form, and join my WhatsApp group. Slides, sources, factsheets and the rebalance logs live there, and members join a daily Google Meet with me. [Sign up for RupeeCase](https://invest.rupeecase.com/signup) [Join the WhatsApp group](https://forms.gle/C6HMtM1Z2GfeocQV7) [One form, one group](https://forms.gle/C6HMtM1Z2GfeocQV7)Scan with your phone camera SSources ## Where the figures come from. 1. [SEBI (LODR) Regulations: results timelines and earnings call transcripts](https://www.sebi.gov.in/legal/regulations/) 2. [SEBI (LODR) Regulations: results timelines and earnings call transcripts](https://www.sebi.gov.in/legal/regulations/) 3. [SEBI (LODR) Regulations: results timelines and earnings call transcripts](https://www.sebi.gov.in/legal/regulations/) 4. [SEBI (LODR) Regulations: results timelines and earnings call transcripts](https://www.sebi.gov.in/legal/regulations/) GTerms ## 5 terms from this module. Signal-to-noise ratioThe ratio of predictable (signal) to unpredictable (noise) variance in returns. Very low for equity monthly returns (~5%), limiting how much ML can help. Effective sample sizeThe number of truly independent observations in a dataset, accounting for correlation. Far smaller than raw observation count for financial data. Gradient boostingAn ensemble method that builds shallow decision trees one after another, each correcting the errors of the previous ones. Useful for equity factor research when kept shallow and regularised. RegularisationTechniques (L1/L2 penalty, dropout, early stopping) that constrain model complexity to prevent overfitting. Essential for all ML applied to financial data. Feature engineeringThe process of constructing new input variables from raw data, combining ratios, detecting interactions, processing text. Often where ML creates more value than in prediction itself. [All 91 terms in the glossary](/learn/glossary) [From the podcast archiveEP29: Pro Just Reset. Day One Of May Series. Buy Dips With HedgeThe Tanmay Edge, 29 Apr 2026, graded 3/5](/podcast/episodes/2026-04-29)[Pillar guideSystematic Investing18 modules across paths 2, 3, 5](/learn/systematic-investing) Education only, not investment advice. RupeeCase is a technology platform operated by QC Alpha Technologies Pvt Ltd; it is not a fund, a PMS or a SEBI registered investment adviser. Model returns are back tested and not a promise of future returns. Figures carry the date they were checked.