The Drivers of Predicted Beta

Abstract

The authors present a methodology for attributing an asset’s or portfolio’s predicted beta to an underlying set of factors. The approach’s foundation rests on the concept of factor beta, which represents the beta of a factor portfolio with respect to the market portfolio. This analysis gives investors important insights into the drivers of predicted beta for a particular portfolio. The authors also study the cross-sectional dispersion of stock-level predicted betas. They show that the dispersion can be decomposed into contributions from individual factors. This framework provides valuable insight into how factor volatilities and correlations affect observed differences in stocks’ predicted betas.