Abstract: With the continuous improvement of the status of the entity economy industry in the national economy, financial risk has showed the characteristics of cross-industry contagion obviously. In April 2017, the Financial Stability Work Conference held by the People’s Bank of China proposed that “the prevention of cross-industry and cross-market financial risks should be taken as a key area to maintain financial stability”. As a part of the overall risk, an idiosyncratic risk constantly diffuses in the system under the push of external impact, and finally converges into the systemic risk and may also become a source of the systemic risk. Therefore, the idiosyncratic risk should attract the attention of scholars. In this context, it is necessary to further analyze and understand the interaction between industries in order to prevent large-scale the contagion and cyclic amplification of idiosyncratic risks, which has an important guiding significance for preventing the cross-industry risk contagion.
In order to explore the taking and contagion of the industry idiosyncratic risk, this paper selects the industry index data of the stock market from 2005 to May, 2021 from CSMAR database as the study sample. After removing market factors represented by HS300 index from the mean equation of GARCH model, effectively extracting the idiosyncratic part of industry return rate and taking the idiosyncratic volatility of industry as the measurement of the industry idiosyncratic risk, this paper combines with DCC-MVGARCH model to construct the industry idiosyncratic volatility association network. According to the topological nature of the network, the overall analysis of the risk taking and risk contagion of various industries in the national economy is carried out. Further, according to the 2008 financial crisis and the 2015 “stock market crash” in China, the research samples are divided into five stages, and then the stage analysis of the risk taking and contagion is carried out.
The results are as follows: (1)The impact of financial events intensifies the heterogeneity risk of industries and enhances the correlation between industries. The inter-industry risk transmission effect and level of risk taking from being strong to weak and from high to low are followed by the period of Chinese stock market turbulence, the period of financial crisis, the late period of financial crisis, the stable period and the early period of financial crisis. Although the correlation between industries gradually weakens after the two sub-extreme events, industry risks still cannot be ignored, and especially when internal shocks occur, industry risks are still highly contagious. (2)The structure and central node of the industry network in different stages are time-varying, and the industry risk level caused by endogenous shocks is higher than that caused by exogenous shocks. During the period of economic turmoil or crisis, light manufacturing, textile and machinery equipment industry have a strong control over the network, for the risk level and risk transmission ability are stronger, while the automobile, food and beverage, transportation, national defense and military industry, communications, banking and non-banking financial industries are all located at the edge of the network, with a low risk bearing and contagion level, and to a large extent are passive recipients of risks. (3)The core nodes of the industry association network are mostly in the middle of the industrial chain or have higher industrial risks, while the edge nodes of the network are mostly at the end of the industrial chain. (4)As the main body of the financial system, banking and non-banking financial industry have negative correlation between the yield rate generated by their own idiosyncrasies and the corresponding yield rate of other industries. The industry position has certain particularity, and the idiosyncrasies of the industry are not significant.
According to the research conclusions of this paper, we can get some inspirations: On the one hand, in the face of economic turbulence and crisis, the industry characteristic risk still has a strong destructive effect, so regulatory authorities should not ignore the industry risk caused by the industry characteristic level while attaching importance to the systemic risk. On the other hand, under the external impact, the mutual connection between industries will be enhanced, and the risk spillover effect will increase. Therefore, it is of great practical significance to classify the systemically important industries, cut off the transmission channels, and establish a good firewall to curb the spread of risks and prevent financial risks.

Key words: DCC-MVGARCH; industry association network; risk contagion; risk taking

摘要： 本文从行业特质波动的视角出发,结合DCC-MVGARCH模型构建行业特质波动关联网络,通过经济体系内部行业间网络的拓扑性质分析了国民经济各行业的风险承担和风险传染问题。研究发现:金融事件的冲击加剧了行业特质性风险,增强了行业之间的关联性;行业关联网络的结构和中心节点具有时变性,而内生冲击引致的行业风险水平高于外生冲击产生的行业风险水平;在经济动荡或者危机期间,轻工制造、纺织服装和机械设备行业对网络具有较强的控制力,风险承担水平和风险传染能力更强,而汽车、食品饮料、交通运输、国防军工、通信、银行和非银金融行业均位于网络边缘,具有较低的风险承担水平和传染水平,在很大程度上属于风险的被动接受者;行业关联网络的核心节点多处于产业链中端或者具有较高的行业风险,而网络的边缘节点多处于产业链的末端。

关键词: DCC-MVGARCH, 行业关联网络, 风险传染, 风险承担