A statistical arbitrage trade is only as good as the costs and fills at its last step

A statistical arbitrage trade is only as good as the costs and fills at its last stepDiagram with 6 boxes. Pick a relation from data that was known before the trade to Measure how far today's price sits from that relation; Measure how far today's price sits from that relation to Enter only when the gap passes a rule fixed in advance; Enter only when the gap passes a rule fixed in advance to Hedge the market move so the position stays neutral; Hedge the market move so the position stays neutral to Exit on return, on a time limit or on a stop; Exit on return, on a time limit or on a stop to Charge costs, rolls, bans and fills, then read what is left.Pick a relation from datathat was known before thetradeMeasure how far today's pricesits from that relationEnter only when the gappasses a rule fixed inadvanceHedge the market move so theposition stays neutralExit on return, on a timelimit or on a stopCharge costs, rolls, bans andfills, then read what is left
The stages of a relative value rule, from the relation to the net result. A map of method, not a result. Source: RupeeCase method map; NSE contract specifications. As of 16 Oct 2026.

Statistical arbitrage is a family of methods, not one trade. A model finds prices that have moved away from a relation learned from history, such as a pair, a basket or a shared factor, and takes a hedged position that profits if the relation comes back. The word arbitrage is loose: nothing here is riskless, and the position can lose while the relation holds.

Three inputs do most of the work. There must be a relation that was stable in the past, a measure of how far today's price sits from it, and a rule for entering, exiting and sizing. Pair trading is the simplest case, and the pair pages on the half life and z-score show the standard measures.

In India the practical problems matter more than the theory. A retail account cannot generally hold a cash short overnight, so the hedge is usually a futures contract. Lot sizes make the hedge ratio lumpy, expiries force rolls, and F&O bans block new positions in one leg. Each of these changes the trade that was tested.

The flow below lists the stages of such a system. Read the last stage first: the cost and execution stage, where a paper edge meets brokerage, STT, slippage and the time it takes to fill both legs.

A relation that held in the sample can end after a rule change, a sector shock or a shift in who trades the stocks. Treat a relation as a hypothesis that needs fresh data, not as a fact about the market.

Before you trade. Derivatives trading carries a high risk of loss. SEBI's study "Profitability of Individual Traders in the Equity Derivatives Segment (FY25 to FY26)" (August 2026) found that 87.7% of individual traders in equity F&O made net losses in FY26, with aggregate net losses of Rs 91,685 crore. Read the SEBI study