Economic value of modeling covariance asymmetry for mixed-asset portfolio diversifications

Mounting evidence from the literature points to the existence of covariance asymmetry for financial assets. That is, conditional volatility and correlation of financial returns tend to rise more after negative return shocks than after positive ones of the same size. This paper extends the literature by investigating whether investors could gain significant economic benefits from incorporating the feature into mixed-asset portfolio diversifications. We carry out the investigation for a portfolio consisting of US stock, REITs, and the risk-free asset, and find that covariance asymmetry is indeed a value-added feature for mixed-asset diversifications. This conclusion is robust to different portfolio performance metrics and asset allocation periods. More importantly, we demonstrate that the value added by modeling covariance asymmetry is unlikely to be offset by transaction costs. This leads credence to the implementability of a portfolio strategy which embeds the feature of covariance asymmetry. Our findings have important implications for fund managers and their clientele.