A wavelet analysis of oil price volatility dynamic

In this the paper we investigate the oil price volatility, by studying the causal relationships between different volatilities captured at different time scales. We first decompose the oil price volatility at various scales of resolution or frequency ranges by using wavelet analysis. We then explore the causalities between absolute returns of oil prices at different time scales. As traditional Granger causality test, designed to detect linear causality, is ineffective in uncovering certain nonlinear causal relationships, we use the nonlinear causality test introduced by Péguin-Feissolle and Teräsvirta (1999) and Péguin-Feissolle, Strikholm and Teräsvirta (2008). Our results confirm the fact that the vertical dependence is a strong stylised fact of oil returns volatility. But, the main finding consists on the presence of a feed- back effect from high frequency traders to low frequency traders. In contrast to Gençay et al. (2010), we prove that high frequency shocks could have an impact outside their boundaries and reach the long term traders.