Definition
Statistical Arbitrage
Statistical arbitrage is a class of quantitative strategies that exploit short-term, statistically predictable price relationships across many securities, holding diversified long and short positions to harvest small mispricings.
Stat-arb in India typically runs across baskets of correlated equities and their futures, using models that identify temporary mispricings and expecting them to revert. Because each edge is tiny and uncertain, the strategy spreads risk across hundreds of positions and relies on the law of large numbers.
Execution depends on low costs, efficient execution algorithms and the ability to short, which in Indian cash equities is constrained. As a result, much domestic stat-arb is expressed through index and stock futures, where leverage and two-sided positioning are straightforward.
Related terms
- Pairs TradingPairs trading is a market-neutral strategy that goes long one security and short a related one when their historical price relationship diverges, betting that the spread will revert to its mean.
- Mean Reversion StrategyA mean reversion strategy assumes that prices or spreads that deviate from a historical average will tend to return to it, so it sells what has risen sharply and buys what has fallen.
- CointegrationCointegration is a statistical property where two or more non-stationary price series move together over the long run such that a particular linear combination of them is stationary and mean-reverting.
- Market-Neutral StrategyA market-neutral strategy balances long and short positions so that the portfolio has little or no net exposure to broad market movements, isolating the manager's stock-selection skill.
Plain-English explainer from The Dispatch Investors Encyclopedia. General information, not financial advice.