Tuesday, January 23rd @ 11:00-12:30 PM, 648 Evans Hall and Zoom
Factor risk models are used for three primary purposes: (1) predicting portfolio volatility, (2) portfolio optimization, and (3) decomposing risk and return into factor and idiosyncratic components. In this paper, we propose explicit tests or “horse races” to evaluate and compare the performance of competing risk models along these three basic dimensions. For evaluating the accuracy of volatility forecasts, we apply bias statistics and Q-statistics and stress the central importance of identifying the predicting horizon, which gives rise to the notion of the term structure of risk. For evaluating the efficacy of risk models for portfolio optimization, we argue strongly against the common practice of using realized information ratio as a measure of portfolio efficiency. Instead, we advocate using the realized volatility of optimized portfolios, which we demonstrate is capable of correctly identifying the more efficient portfolio in a tiny fraction of the time. Finally, we describe tests to directly compute the realized correlations between factor and idiosyncratic return contributions.