Shrinkage Estimation in Risk Parity Portfolios

Abstract

We investigate the impact of shrinkage estimation techniques for the moments of asset returns on risk-parity portfolios. In contrast to mean-variance portfolios, the risk contributions of individual assets in risk-parity portfolios are fixed a priori. This additional restriction stabilizes empirical portfolio weights in time. We show that the marginal risk budget for each portfolio asset indeed serves as a natural shrinkage target. Hence, we provide a new perspective on risk-parity portfolios. In an extensive empirical application, we compare and combine the various shrinkage strategies to popular risk-based approaches from the literature. We find that while using shrinkage estimators in risk-parity portfolios enhances out-of-sample performance based on several criteria, traditional covariance shrinkage estimators dominate all other strategies in high-dimensional settings.