Statistical arbitrage (or StatArb) can be seen as a model-based investment (or speculation) process, by constructing simultaneously long and short positions of different assets. The selected assets’ current values differ (on a relative basis) from a theoretically or quantitatively predicted value. The constructed positions should represent neutrality, for instance regarding industry, sector, market and/or currency.

StatArb is widely used by hedge funds and other sophisticated market participants. Statistical arbitrageurs (also called ‘market-neutral managers’) are trying to profit from temporary deviations of market prices from their fundamental value. To implement statistical arbitrage, they combine skills and experience with modern science. This means they apply the value theory, statistical decision theory, game theory, statistical pattern recognition techniques, time series techniques and co-integration. Statistical arbitrageurs identify the arbitrage situation through mathematical modelling techniques and if performed well, a profit situation can arise from pricing inefficiencies between assets. Altogether, statistical arbitrage presumes statistical mispricing of price relationships that are true in expectation in the long run when repeating a strategy.

Statistical arbitrage is much wider than deterministic arbitrage and it is impossible to give a complete overview of the possible strategies. Statistical arbitrage is also described as an ‘art of association’. This refers to the construction of equivalence (in value) of portfolios across different assets.

Statistical arbitrage strategies have the following characteristics:

- Creates a lower volatility in portfolios.
- Executing transaction signals are systematic, or rules-based (not driven by fundamentals).
- The strategies have zero ‘Beta’ with the market. It implies that they are market-neutral. Hence, the yields of the strategies are independent of, and unrelated with, the market. Moreover, the strategies can even generate positive returns in case of economic downturn.

In a general sense, statistical arbitrage strategies perform ideal if the amount of transaction time approaches infinity, while market liquidity also approaches the ideal level. Over any finite period of time, a series of low probability events may occur that impose heavy short-term losses. If those short-term losses are greater than the liquidity available to the market participant, default may occur (see: the LTCM hedge fund debacle).