Correlation Trading
Correlation Trading involves taking explicit positions on the realized or implied correlation between two or more assets — separate from their individual volatilities — most commonly through variance swaps, correlation swaps, or structured products that pay off based on the degree to which assets move together rather than independently.
Correlation is a fundamental driver of portfolio risk and derivatives pricing, yet it is among the most difficult financial variables to estimate, forecast, or hedge. Traditional option pricing models handle correlation implicitly through the volatility surface, but correlation trading isolates correlation as a directly tradeable quantity. The development of variance swaps and correlation swaps as liquid over-the-counter instruments in the early 2000s made explicit correlation trading by institutional investors and hedge funds practical for the first time.
A correlation swap pays the buyer the difference between realized pairwise correlation among a basket of assets and a fixed strike level, multiplied by a notional amount. Buying a correlation swap is a bet that realized correlation will be higher than the strike; selling it is a bet that realized correlation will be lower. Alternatively, correlation can be traded synthetically by combining variance swaps on a basket and its components: the difference between a basket variance swap and the weighted sum of individual component variance swaps is mathematically equivalent to a correlation position.
The typical correlation trade in equity markets involves selling correlation — betting that stocks will move less in lockstep than the options market implies. This structural trade has historically been profitable because index options buyers create persistent demand for index volatility protection, elevating the implied correlation embedded in index options above subsequently realized correlation. This is essentially the same dynamic that makes dispersion trading attractive, and the two strategies are closely related.
Credit correlation trading is a distinct and more complex application. The pricing of CDO (collateralized debt obligation) tranches depends critically on the assumed default correlation among the underlying reference credits — specifically, how likely simultaneous defaults are. Senior tranches of a CDO are loss remote if defaults are spread out but highly vulnerable if correlations are high and many entities default simultaneously. The 2008 financial crisis dramatically illustrated the danger of mispricing credit correlation: the tranched mortgage CDO market imploded when realized default correlations proved far higher than the models had assumed, creating catastrophic losses for investors who had sold protection on senior tranches.
In currency markets, correlation trading involves positions on whether two currency pairs will continue to move together or diverge. Given that EUR/USD, GBP/USD, and other dollar pairs share dollar exposure, they tend to be highly correlated. Positions can be established through options structures that are long individual pair volatility and short basket volatility, or through direct spread trading in the OTC market. As with all correlation strategies, the primary risk is that correlation spikes during periods of risk-off deleveraging, punishing short correlation positions at the worst possible time.