Quant trading in India

What "quant" means

A quantitative trader turns an idea about markets into a rule, measures what the rule would have done on past data, and trades it only if the evidence holds up.

It is a method, not a speed. High-frequency firms are quants, and so is an investor who rebalances a portfolio once a month by a tested formula.

Quant and algo are different things. Quant is how the decision is reached. Algo is how the order is placed. Many algos automate a hunch that was never tested, and plenty of quantitative research is traded by hand.

Who does it in India

• Proprietary trading firms in Mumbai, Gurugram and Bengaluru, some among the largest by volume on NSE, trading at high frequency from co-location.
• Asset managers running quant and factor funds, and the factor indices NSE publishes (momentum, quality, value, low volatility).
• Portfolio managers, alternative funds and family offices running systematic books in index and stock derivatives.
• Individuals, now that broker APIs and daily data are within anyone's reach.

The data India gives away

NSE publishes, free and every day: the price of every stock and every derivative contract (the bhavcopy), open interest by category of participant, index values with their P/E and dividend yield, bulk and block deals, and company results in machine-readable form.

That is enough to study market regimes, valuation, breadth, positioning, the option chain and company fundamentals without buying anything. Intraday and tick data are the part that costs money.

Aurevian Schema is built on exactly this public data: it is what a small quant team can do with it.

What tends to be traded

• Index option selling and volatility: the largest systematic activity in Indian markets, built on implied volatility usually exceeding what follows.
• Intraday momentum and breakout rules on index futures and options.
• Pairs and statistical arbitrage between related stocks, or between a stock and its future.
• Factor portfolios: ranking stocks on momentum, value, quality or low volatility and rebalancing on a schedule.
• Event and flow studies: results, index changes, expiry effects, institutional positioning.

None of these is a secret. The edge is in the details: the exact rule, the costs, the sizing, and the discipline to stop when the evidence changes.

The honest part

Most apparent patterns are noise. A quant's job is mostly to find out which ones.

Three questions do most of the work. How many separate times has this been seen? Fifty days in a row inside one rally is one case, not fifty. What did it cost to trade? Does it hold on data it was not found in?

Schema applies the first of those to everything it shows: it counts separate episodes, and says so when there are too few. Backtesting Indian markets covers the rest.

How to begin

• Learn to handle data before learning strategies: a history with unadjusted splits will fool any model.
• Start with daily data and slow rules. They are cheaper to test and far more forgiving of slippage.
• Write down the rule and the test before looking at the result.
• Keep a record of every idea tried. Ten failures behind one success is a warning about the success.
• Trade small, and compare what happened with what the test said would happen.

Keep reading