Arbitraje estadístico
Statistical arbitrage is a quantitative approach that trades many small, temporary mispricings between related securities, based on statistical models, with positions usually hedged against the market.
Explained
It grew from pairs trading at Morgan Stanley in the 1980s into large portfolio strategies at quant funds. Profits per trade are small, so execution costs, risk models and diversification across hundreds of positions are essential.
Example
A model ranks 500 stocks daily by short-term deviation from their sector peers.
It buys the most oversold and shorts the most overbought, keeping net exposure near zero.
Common mistakes
Assuming 'arbitrage' means no risk.
Ignoring crowding: many funds may hold similar trades.
Underestimating costs at high turnover.
Questions traders ask
Is statistical arbitrage risk-free?
No. It relies on historical relationships that can break, as in the August 2007 quant crisis.
Can individuals do stat arb?
Simple pairs strategies are possible, but large-scale stat arb needs infrastructure and low costs.