Multivariate Jump Diffusion Model with Markovian Contagion

Asset prices exhibit significant deviation from log-normality, with time-varying stochastics and ample evidence of jumps transmitting from one asset or market to another. We propose a multivariate jump diffusion model with Markovian contagion to capture these asset price dynamics, where the contagion channel periodically switches from an active to an inactive state. Using a dynamic conditional correlation network approach to estimate the Markovian contagion model, we apply the model to an inter-national equity and currency portfolio allocation. The fat tail characteristics captured help evaluate the extent of model risk, intra-asset class, inter-asset and inter-region contagion.