Risk Premia and the Dynamic Covariance between Stock and Bond Returns

We investigate two topics: (1) the nature of the dynamic covariance matrix of stock and bond returns, and (2) the intertemporal relation between risk and return. We estimate a conditional two-factor variant of Merton's ICAPM in which long-term government bond returns proxy for the second risk factor. Stock and bond risk premia are linear functions of an asymmetric dynamic covariance (ADC) matrix for stock and bond returns. We find that conditional bond variance responds symmetrically to bond return shocks but is virtually unaffected by stock return shocks, while conditional stock variance responds asymmetrically to both stock and bond return shocks. Models that impose a constant correlation restriction on the covariance matrix between stock and bond returns are strongly rejected. We find that intertemporal risk-return relationships are very sensitive to the specification of conditional first and second moments