Quantitative Trading

Part 1 sets the frame. It defines quantitative trading as a research science, lays out the workflow every quant follows, and teaches just enough probability and statistics that the reader can tell a real result from a random one, which is the entire game.

• 1The scientific method applied to marketsQuantitative trading is the use of data, statistics, and code to find and test trading edges systematically, treating each idea as a hypothesis to be proven or, far more often, rejected. This chapter contrasts it with discretionary trading, frames an edge as a statistical claim, and sets the honest spine that most claimed edges are noise and that the professional world is largely closed to retail. 10 min
• 2How a quant actually worksA quant follows a disciplined funnel: form a hypothesis grounded in a reason, gather clean data, test carefully, validate on unseen data, and only rarely deploy, with most ideas dying along the way. This chapter lays out that process and argues that the discipline of the process, not any single clever idea, is what keeps a researcher honest. 9 min
• 3Enough statistics to not fool yourselfYou cannot judge an edge without the basic tools of probability: distributions, the mean and variance, sampling, and the standard error that tells you how much a result might be luck. This chapter teaches exactly that much, in plain terms with code, so the rest of the course has the vocabulary to separate a finding from a fluke. 10 min
• 4Telling signal from noiseThe central problem of quant research is separating a real, repeatable signal from random noise, and markets are mostly noise. This chapter shows how much data it takes to be confident an effect is real, why a good-looking backtest is not the same as a statistically significant one, and how easily randomness imitates a pattern. 9 min

Part 2 is about the raw material of all quant work and the traps hidden in it. Getting the data right is the unglamorous majority of the job, and the biases that live in bad data, survivorship and look-ahead above all, are the quiet cause of most backtests that lie.

• 5The raw materialQuant strategies run on data: prices and volumes, fundamentals, corporate actions, and a defined universe of instruments, each with its own quality and pitfalls. This chapter surveys what data a quant needs, where Indian market data comes from, and why the quality of the data sets a hard ceiling on the quality of any result. 9 min
• 6The unglamorous eighty percentRaw data is full of splits, bonuses, dividends, gaps, and errors, and using it unadjusted produces nonsense; worse, using tomorrow's information in today's test produces a beautiful lie. This chapter covers adjusting prices for corporate actions and, crucially, point-in-time data that avoids look-ahead bias in fundamentals. 9 min
• 7The stocks that vanishedA backtest run only on the stocks that still exist today quietly excludes every company that failed, inflating results by ignoring the losers, which is survivorship bias. This chapter defines a tradable universe honestly, including delisted and dead stocks, and covers India-specific realities like liquidity filters and the futures-and-options ban period. 9 min
• 8Returns are not a bell curveQuant models often assume returns are normally distributed, and they are not: real returns have fat tails, cluster their volatility, and change character across regimes. This chapter shows those properties in real data and warns that a model built on the bell curve will underestimate exactly the rare, large losses that matter most. 10 min

Part 3 surveys the quantifiable phenomena that genuinely drive returns, taught as the factors behind India's own smart-beta indices. These are the closest thing to real, documented edges in the public literature, and the part treats each honestly: what it is, why it might work, how it is built, and how it fails.

• 9Ranking the whole marketThe core quant idea is the factor: a measurable characteristic by which you can rank every stock, buying the top and avoiding or shorting the bottom, betting on the characteristic rather than on any one company. This chapter introduces cross-sectional thinking and grounds it in the NSE factor indices that already package these ideas in India. 9 min
• 10Buying the winnersMomentum, the tendency of recent winners to keep winning for a while, is the most documented factor and the basis of the NSE momentum index. This chapter builds a simple momentum rank, explains the behavioural story behind it and the brutal crashes it suffers when trends reverse, and tests it honestly. 9 min
• 11Cheap and goodValue (buying cheap stocks by fundamental ratios) and quality (buying financially healthy, profitable ones) are two of the oldest factors, and India publishes indices for both. This chapter builds each from fundamentals, contrasts their opposite cyclicality, and shows why combining them can be steadier than either alone. 9 min
• 12Less risk, more returnThe low-volatility anomaly, that calmer stocks have delivered strong risk-adjusted returns, is one of the most surprising findings in finance, and the size factor (smaller companies) one of the oldest. This chapter builds both, explains why low volatility contradicts naive theory, and treats the defensive factors as the sane core of a factor portfolio. 9 min
• 13Trading the spread between two stocksBeyond cross-sectional factors sits statistical arbitrage: betting that a relationship between prices will revert to normal, most classically the pairs trade on two related stocks. This chapter introduces mean reversion, the ideas of stationarity and cointegration in plain terms, and builds a simple pairs trade, while being honest that this space is crowded and hard. 10 min

Part 4 is the scientific heart of the course and its hardest honesty. It builds a proper portfolio backtest, judges it by the metrics that matter, and then confronts the statistics that expose false edges: the multiple-testing problem, out-of-sample validation, and the deflated Sharpe ratio. This is the part that turns an enthusiastic backtester into a sceptical researcher.

• 14Backtesting a whole portfolioA quant backtest is not one asset traded in and out but a whole portfolio rebalanced on a schedule, ranking a universe and holding many names at once. This chapter builds that portfolio backtest, with periodic rebalancing and realistic transaction costs, extending the single-asset backtest from Python for Trading to the cross-section. 9 min
• 15Return is not the number that mattersA raw return figure tells you almost nothing; a professional judges a strategy by risk-adjusted and behavioural measures: the Sharpe and Sortino ratios, the information ratio, maximum drawdown, turnover, and hit rate. This chapter computes each from a backtest and explains what a good and bad value of each really means. 9 min
• 16Torture the data and it confessesThis is the deepest honesty chapter in the catalogue. If you test enough strategies, the best one will look excellent purely by chance, so a great backtest found after many tries is probably luck, not skill. This chapter demonstrates the effect directly in code and names the trap that ruins most quant research. 10 min
• 17Testing on data you have never seenThe defence against a false edge is to judge a strategy only on data it was never tuned on, using held-out test sets and walk-forward analysis that repeatedly trains on the past and tests on the next unseen slice. This chapter builds that validation, extending the walk-forward idea from Algorithmic Trading, and names the leakage mistakes that quietly reintroduce cheating. 9 min
• 18The honest Sharpe ratioThere is a way to put a number on the multiple-testing problem: the deflated Sharpe ratio, which lowers a strategy's apparent Sharpe to account for how many strategies you tried before finding it. This chapter explains the idea in plain terms, computes it in code, and uses it to turn an impressive backtest into an honest expectation. 10 min

Part 5 assumes you have one or more edges that survived honest testing, a rare thing, and builds them into a portfolio. It covers combining signals, constructing and risk-modelling the portfolio, sizing positions, and managing the risk of the whole book, all at the portfolio level the earlier risk course could only gesture at.

• 19Many small edges beat one big betA single edge is fragile; a portfolio of several weak but uncorrelated edges is far steadier, because their ups and downs partly cancel. This chapter shows how to combine factors and strategies, why the correlation between edges matters more than the strength of any one, and how diversification of edges is the quant's real advantage. 9 min
• 20Building the portfolioTurning signals into positions is portfolio construction, and it rests on a risk model, an estimate of how assets move together. This chapter covers the covariance idea, the common weighting schemes from equal-weight to risk parity to mean-variance, and the honest warning that mean-variance optimisation is dangerously sensitive to its inputs. 9 min
• 21How much to betHow much to allocate to each edge is as important as the edge itself, and the mathematics of optimal sizing runs from volatility targeting to the Kelly criterion, which the Risk and Psychology course deliberately deferred to here. This chapter teaches sizing at the portfolio level and the hard lesson that full Kelly is too aggressive to survive, so professionals use a fraction of it. 10 min
• 22Managing the whole book's riskA quant manages risk at the level of the whole portfolio, not the single trade: total volatility, drawdown, exposure to each factor, and the tail risk that normal models miss. This chapter covers portfolio value-at-risk and its limits, factor-exposure control and neutralisation, stress testing, and the discipline of turning a strategy off when it breaks. 10 min

Part 6 is the catalogue's final reckoning. It explains why most quant strategies fail even when built well, measures honestly the gap between a retail trader and the professional quant world, and lands on the genuinely useful and accessible version of everything the course taught: disciplined factor investing through India's smart-beta indices. It closes the whole catalogue.

• 23The strategy graveyardEven a strategy that was real can die, and this chapter catalogues how: overfitting that was never caught, capacity limits that size destroys, crowding as others find the same edge, costs that grind it down, and regime change that ends it. Alpha decays, and a quant's job is as much retiring dead strategies as finding new ones. 9 min
• 24You versus the professionalsThe people who make quant trading pay have advantages a retail trader cannot match: better and cleaner data, teams of researchers, large-scale infrastructure, lower costs, and, at the fast end, co-located machines. This chapter measures that gap honestly, so the reader can see clearly which few quant approaches remain sensible for an individual and which are a fantasy. 9 min
• 25The realistic pathThe genuinely useful, accessible version of everything in this course is factor investing: gaining disciplined, rules-based exposure to documented factors through the smart-beta index funds and ETFs India's exchanges already offer, rather than running a solo strategy. This chapter presents it as the sane quant path for most people, low in turnover, evidence-based, and buildable today. 9 min
• 26The end of the road, and the start of the workThe final chapter of the final course recaps the research process and the statistics of honesty, points the reader to a lifelong sandbox practice of careful research, and closes the whole catalogue. The lasting message is the one every course built toward: data, code, and mathematics are tools for finding the truth about an idea and, far more often, for rejecting it, not edges in themselves. 9 min