Wavelet dynamics for oil-stock world interactions

A previous research ignores the distinction between short term and long term, and by decomposing financial variables (world general and stock market indexes) and the macroeconomic variable (oil prices) at various time scales, we study the relationship among series on a daily scale by scale basis. Continuous time wavelets help to circumvent the problems associated to basic linear regressions and given that stock-oil relationships are usually described as complicated we extend previous findings by providing more generalized and convincing results, in analyzing contagion and interdependence issues as well as lead and lag effects for both world general and sector stock levels between December 1992 and October 2012. The relationship between oil prices and sector stock returns is ambiguous, because results seem to show that there are both phase and anti-phase relationships, where mostly it is oil that is the lagging variable, independently of the sector under analysis. There is higher coherence among series for higher scales thus supporting the interdependence hypothesis, showing that long run market dynamics are more uncertain. Empirical results indicate a bidirectional relationship between both series for large time horizons, which can be associated to fundamentalist traders, especially fund managers and institutional investors, and which depend on the historical period under analysis.