Volatility discovery in G-7 stock markets based on evidence from realized kernels

• We investigate the volatility discovery hypothesis in G-7 financial markets. • An FC-VAR model and spectral analysis are applied to realized kernels. • Our findings reveal that Canada and the UK are in equilibrium with the US. • Germany, France, Italy, and Japan often lead in volatility response. This study investigates empirically the volatility discovery hypothesis in G-7 financial markets by applying a Fractionally Cointegrated Vector Autoregressive (FC-VAR) model directly to realized kernels instead of returns. We further enhance our analysis using Wavelet Coherence to verify long and short-term co-movements, in both time and frequency domains. Our main findings reveal that the Canadian and UK markets have an equilibrium in volatility discovery with the US market. Conversely, empirical evidence support that the markets of Germany, France, Italy, and Japan lead the volatility discovery compared to US, suggesting potential diversification opportunities. Regarding the volatility discovery hypothesis, our analysis indicates that the most non-US markets (except for Canada and the UK) often lead in volatility response, likely due to exogenous shocks located mainly outside US (i.e., pandemia, Russo-Ukrainian war) during the sample period. This study contributes to the existing providing a broader understanding of financial market dynamics.