Algorithmic Strategy Types
Algorithmic strategy types in equity and derivatives markets encompass distinct approaches including trend following, mean reversion, momentum, statistical arbitrage, and market making — each characterised by a different view of price behaviour and designed to perform well in specific market conditions.
Algorithmic trading in India grew substantially after SEBI introduced the regulatory framework for algorithmic trading in 2008 and allowed Direct Market Access (DMA) for institutional participants. By the mid-2020s, a significant portion of NSE exchange turnover was attributed to algorithmic execution. Understanding the major strategy archetypes helped both practitioners and informed observers contextualise market behaviour.
Trend following strategies assumed that price movements in one direction were more likely to continue than reverse in the short to medium term. They used signals such as moving average crossovers, channel breakouts, and momentum filters to enter in the direction of an established trend and exit when the trend showed signs of reversal. Trend following historically performed well during strong directional markets — the 2020 post-COVID rally and the 2014 Nifty bull run were environments where trend systems generated strong returns — but produced frequent small losses during sideways consolidations.
Mean reversion strategies assumed that prices deviated temporarily from a fair value level and would return to it. Pairs trading — trading correlated stock pairs (for example, HDFC Bank and ICICI Bank, or Nifty and Bank Nifty relative to historical spread) by shorting the outperformer and buying the underperformer — was a common mean reversion implementation. Statistical tests including cointegration analysis were used to identify pairs with stable long-run relationships. Mean reversion worked well in range-bound conditions but lost on trending days when divergences extended rather than converged.
Momentum strategies — distinct from trend following — typically operated on shorter timeframes and focused on the persistence of recent outperformance. Factor-based momentum strategies bought stocks that had significantly outperformed the broader index over the prior 12 months (excluding the most recent month) and shorted or avoided underperformers. NSE research and academic studies on Indian factor investing showed that a momentum factor had historically delivered excess returns over the broader market, though with periodic sharp reversals (momentum crashes) during volatile market transitions.
Statistical arbitrage expanded the pairs concept to portfolios of many instruments, using factor models to identify relative mispricing across sectors or within an index. In Indian markets, this required sophisticated quantitative infrastructure and low-latency execution, making it primarily an institutional strategy.
For retail traders building algorithmic strategies in India, trend following and mean reversion represented the most accessible starting points. Platforms such as Streak (integrated with Zerodha), Tradetron, and Python-based frameworks via broker APIs allowed retail participants to build, backtest, and deploy rule-based strategies — with SEBI's 2022 framework requiring all client-side algorithms to be approved and registered with the broker before live deployment.