Asset allocation models and market volatility

Asset allocation and risk management models assume at least short-term stab ility of the covariance structure of asset returns, but actual covariance a nd correlation relationships fluctuate dramatically. Moreover, correlations tend to increase in volatile periods, which reduce the power of diversific ation when it might most be desired. We propose a framework to both explain these phenomena and to predict changes in correlation structure. We model correlations between assets as resulting from the common dependence of retu rns on a marketwide factor. Through this link, an increase in market volati lity increases the relative importance of systematic risk compared with the unsystematic component of returns. The increase in the importance of syste matic risk results, in turn, in an increase in asset correlations. We repor t that a large portion of the variation in correlation structures can be at tributed to variation in market volatility. Moreover, market volatility con tains enough predictability to construct useful forecasts of covariance.