Diverging Dynamics of Global Oil Market Volatilities : A Generalized Structural Analysis

The global oil market has experienced a steady decline in oil production volatility and a concurrent increase in the volatility of real oil prices since the mid-1980s. To study the underlying causes of the diverging volatility dynamics, we estimate a generalized time varying SVAR model that allows for disturbances of observation and transition equations to be correlated. This approach is particularly suitable for capturing the effects of changes in oil levels on oil volatilities. The Bayesian model comparisons decisively prefer our generalized model against traditional models. Our model shows that changes in the speed of adjustment of oil price to its equilibrium accounts for much of the observed diverging dynamics of volatilities. The generalized model also attributes the volatility divergence to decreased variability of flow supply shocks (responsible for almost 32% of the reduction in the oil production volatility) and increased variability of ow demand and speculative demand shocks (responsible for about 57% of the increase in the oil price volatility). Relative to traditional models, we find that the enriched dynamics between the first and second moments of endogenous variables in the generalized model are required to reveal these new stylized facts for the global oil market