Market variance risk premiums in Japan for asset predictability

This article evaluates the predictive performance of variance risk premiums (VRPs) in Japan on the Nikkei 225 returns, credit spreads, and the composite index of coincident indicators. Different monthly VRPs, such as expected and ex-post VRPs, are measured by using model-free implied and realized variances from option prices and high-frequency (HF) data, and their predictive ability is compared with that of VRPs using a realized measure based on coarser frequency return observations. The empirical results show that the VRPs in Japan with HF data are useful in predicting credit spreads and the composite index of coincident indicators, but lose their predictive ability for the Nikkei 225 returns. Such significant predictive power tends to be greater for the expected VRPs with HF data relative to the ex-post VRP with HF data and VRPs with daily data as well as for lower investment grade credit spreads.