Oil price shocks and stock–bond correlation

Abstract

This paper investigates the role of oil as a determinant of the US stock–bond correlation. The analysis uses monthly data over the period from February 1990 to July 2021. We examine the impact of oil shocks, using the Ready (2018) method, alongside a range of macroeconomic variables on the nature of stock–bond dynamic correlation. Our main findings demonstrate that during recessionary periods, the stock–bond correlation is adversely and statistically explained by oil supply shocks, and that correlation tends to statistically diverge from that of it is counterpart during expansionary periods. In addition, demand shocks are more pronounced in periods of pessimistic investor sentiment, whereas supply and risk shocks appear during optimistic periods. From a risk management standpoint, a backtesting exercise shows that the incorporation of supply and demand shocks generally improves the forecast of portfolio volatility under various portfolio weighting schemes and market conditions. A time-varying hedging exercise also reveals that accounting for both demand and supply shocks reduces the cost of hedging, mainly following major crisis periods. The contribution of shocks appears most with a long position in the bond market after the 2014 oil crisis period. Our main results remain the same after performing a set of the robustness checks.