Asymmetric volatility transmission in Japanese stock market in the presence of structural breaks

This research shows that accounting for structural breaks within the asymmetric GJR-GARCH model reduces volatility persistence in stock and oil returns. More importantly, we ascertain that good news has less impact on current volatility if sudden changes are accounted for, while bad news has more impact in all cases with an exception of Nikkei225. Our findings are in line with previous studies and highlight that it is plausible to include structural breaks in asymmetric GARCH models to model volatility dynamics more precisely. The results of the univariate analysis provide evidence of bi-directional volatility spillover effects between the Japanese stock market and the rest of the markets, independently of the model choice. Thus, neglecting structural breaks in variance may lead to different results but not to the extent of distortion of the direction of volatility transmission effects. The latter finding does not hold true under the multivariate context highlighting the power difference between the CCF approach and the BEKK-GARCH models. Finally, we ascertain that data frequency may also affect the direction of the volatility spillovers among the markets.