Equity Factor Timing With Kalman Filter

Abstract

Extensive research literature has shown that equity factor premia is not constant over time. With the broad adoption of factor investing among active managers, to generate differentiated market views and returns, factor or style timing remains as one attractive meaning. We present an adaptive linear factor model to capture time-varying factor risk premia and discuss its application in equity portfolio investment. The model framework that allows time-varying factor returns can potentially capture more returns and allows an investor to understand the changes in cross-sectional factor returns. We estimate model coefficients through a Kalman filter model and discuss the results with an empirical example.