时频视角下国际股市间高阶矩风险溢出效应研究

Research on Higher-Moment Risk Spillover Effects among International Stock Markets from the Time-Frequency Perspective

本文将高阶矩波动(GARCHSK)模型和Frequency Connectedness方法相结合,分别从时域和频域视角量化国际股市间高阶矩风险溢出效应,并运用滚动时间窗法刻画国际股市间的动态时变风险溢出效应。实证结果表明,除均值和波动溢出效应外,国际股市间的高阶矩风险溢出效应同样十分显著;美国、英国、德国、法国和加拿大股市主要扮演风险净溢出者的角色,而澳大利亚、日本、新加坡、中国内地和中国香港股市主要扮演风险净接受者的角色;国际股市间的均值和峰度溢出效应主要在短期(1~5天)传播,波动和偏度溢出效应主要在长期(22天以上)传播;重大国际危机事件往往会增强国际股市间的风险溢出效应,且危机期间的风险溢出效应强度显著高于危机前期和危机后期;中国香港股市不仅是中国内地股市的重要风险传染源之一,也是中国内地股市与国际股市间的主要风险传染中介。

The measurement of risk spillover effects among international stock markets has always been a hot topic in the academic circles and a frontier research topic in the field of financial engineering. However, no literature to date has investigated the higher-moment risk spillovers among international stock markets. Given the drawbacks of existing literature,this paper aims to quantify the higher-moment risk spillovers among 10 major international stock markets from time-domain and frequency-domain perspectives simultaneously.Given the advantages of the Frequency Connectedness and GARCHSK model, this paper combines them together to analyze the higher-moment risk spillovers among international stock markets. First, the GARCHSK higher-moment volatility model is estimated on each return series to obtain the conditional volatility, skewness, and kurtosis series. Second, the DY spillover index and BK frequency connectedness models are applied to obtain the time-domain and frequency-domain spillover effects. Third,the directional net-pairwise connectedness networks are established based on the net-pairwise spillover index extracted from the connectedness tables. Fourth, the 200-day rolling window method is utilized to obtain the dynamic time-varying risk spillovers. Finally, a robustness check is conducted by choosing different rolling window sizes.The empirical results obtained in this paper demonstrate that besides the mean and volatility spillovers, the skewness and kurtosis risk spillovers among global stock markets are also significant. The U.S., the UK, German, French, and Canadian stock markets mainly act as the risk spillover net-transmitters. Correspondingly, Australian, Japanese, Singapore, China's Mainland and Hong Kong SAR stock markets mainly act as the risk spillover net-recipients. The mean and kurtosis spillovers mainly spread in the short-term(1~5 days),while the volatility and skewness spillovers mainly transmit in the long-term(22~inf days). The international major crisis events tend to enhance the risk connectedness among global stock markets, and the connectedness during the financial crisis is significantly stronger than pre-and post-financial crisis periods. Moreover, the Hong Kong SAR stock market is not only the risk connectedness transmitter to China's Mainland stock market but also one of the crucial risk spillover intermediaries between China's Mainland stock market and other stock markets.To highlight the significance of higher-moment risk spillovers and improve the effectiveness of financial risk management,this paper provides several implications for international investors and regulators. First, the market participants should keep an eye on the spillover effects transmitted from the U.S., the UK, German, French, and Canadian stock markets to their own stock markets so that the risk-averse measures can be taken in time for the outbreak of extreme risk events in the international stock market. Second, the Chinese government should further improve relevant laws and regulations, and enhance the international influence of the mainland stock market, and strengthen the ability to withstand the risk spillovers from other stock markets. Third, investors and regulators should guard against the risk contagions among international stock markets,especially during the financial crisis, pay more attention to the short-term mean and kurtosis spillovers, as well as the longterm volatility and skewness risk spillovers. Last but not least, regulators should formulate contingency plans for major international crisis events, establish an information-sharing mechanism, and construct a cooperative risk supervision system,thus achieving efficient and collaborative supervision of the risk spillover effects among international stock markets.