Multi Time Scales Inter-Industry Risk Contagion Networks Based on EMD and Granger Causality

Since the 2008 global financial crisis, identifying and measuring risk in financial system has been a crucial research direction. In our paper, we construct inter-industry causality networks of volatility risk on different time scales with the help of the empirical mode decomposition method from signal processing via Granger causality test. We find that the most influential industries on different time scales are similar despite that the shortest-term components are disturbed. As the time scale increases, the pattern tends to be more clarified that the comprehensive, financial, and emerging industries such as computer are the most influential in risk contagion.