关键词: 金融子市场, 自我防御, 风险传染, 非线性MSBIARCH模型

Abstract: Identifying the risk contagion path in financial sub-markets under the impact of extreme events has become an important part of financial risk prevention and control in the new era. Based on non-linear MSBIARCH model, this paper conducts a study to discriminate the volatility and contagion relationships and the volatility clustering trends among stocks, real estate, bulk commodities, exchange rates and bond markets. The findings show that, firstly, except for the one-way volatility contagion effects from the exchange rate market to the real estate market, from the bond market to the stocks, real estate, exchange rates markets and from the bulk commodity market to the bond market, there exists a two-way volatility contagion relationship between the other markets; secondly, the extreme risk events trigger the volatility contagion through market sentiment and expectations, the public emergencies generate volatility contagion based on market sentiment, and the policy announcements cause volatility contagion through fundamental factors and market expectations; finally, most of the volatility cluster trends in financial sub-markets originate from the volatility contagion in other sub-markets. The exchange rate market has weaker risk contagion ability, while the bulk commodity market and the bond market have stronger self-defense ability, which mainly carry out risk outputs. Therefore, it is necessary to further establish a full-process financial risk prevention and control system, improve the financial sub-market trading mechanism, strengthen communication and cooperation between markets, and reasonably guide public expectations, so as to enhance the anti-risk ability of financial markets and maintain the steady development of financial markets.

Key words: financial sub-markets, self-defense, risk contagion, nonlinear MSBIARCH model