This paper study the relationship between oil and stock markets in G7 countries, by distinguishing between interactions based on fundamentals (long-term interdependence: high memory impact) and contagion (short-term interaction: transitory contamination). To do this, we propose in the first time two complementary frequency approaches based: the evolutionary co-spectral analysis and the wavelet approach allowing a time-varying measure of the dynamic correlation between the oil and stock markets over time and across time horizons. We find that interdependence between oil price and the stock market is more pronounced in the short and medium terms than in the long term. In addition, we prove that stock markets are more sensitive to oil shocks originating from demand shocks. These findings provide important policy implications for both policymakers, in terms of taking relevant actions regarding oil shocks originating from the demand side, and investors, in terms of a policy of diversification that depends on horizons.