题目：“经济学系列讲座”第43期 Conditional Heteroskedasticity with Risk Spillover Through Networks: An Exponential GARCH Approach
主讲人：Yang Yang 美国俄亥俄州立大学经济系
By introducing both intra-temporal and inter-temporal risk spillover through network, we propose a new multivariate conditional volatility model. For stationary case, the model can capture the dynamic of conditional heteroskedasticity structure when there are long-run stable links among multiple markets, and it is easy to be estimated consistently by QMLE approach. By Monte Carlo simulations, we show good finite sample performance when n/T→0. When applying the model to monthly stock return innovations of 11 eurozone countries from March 1999 to April 2021, by using geographical and institutional links to capture the network between the countries, the performance of our model dominates single variate GARCH(1,1), EGARCH(1,1) and multivariate GARCH with both constant correlation and dynamic conditional correlation settings by likelihood values and AIC criteria.
Yang Yang is a Ph.D. candidate in economics from the Ohio State University. His research focuses on spatial econometrics, finance, and labor economics.