Abstract:

This paper uses the LASSO-CoVaR model to measure the correlation of the tail risk of real industries in China. Moreover, the influencing factors of the risk spillover in real industries are investigated by using the industry characteristics data. The main findings are as follows: First, the tail risk spillover among real industries in China presents the features of “Near Crisis”, and there are significant differences in the fluctuation of the risk condition of different industries. Second, the electronic and information industry has become the prime risk spillover contributor in the tail risk connectedness network of real industries. Computer, nonferrous metals and national defense industries are the major risk absorber. Third, the risk correlation among real industries with relevant business shows steady homogeneity. Finally, the specific characteristics of every industry, such as the risk, the scale and the leverage ratio, can significantly affect the intensity of the risk spillover.

Key words: Real Industry, Tail Risk Interconnectedness, Risk Spillover, LASSO-CoVaR

摘要：

利用LASSO-CoVaR模型对我国实体行业风险溢出机制与特征进行分析,并利用行业特征数据考察实体行业风险溢出的影响因素。研究发现:我国实体行业尾部风险溢出总体水平呈现“近危机”特征,且不同行业的风险波动具有显著差异;以电子、计算机为代表的电子信息产业已经成为主要的风险溢出者,计算机、有色金属与国防军工为主要的风险吸收者;业务相关性较高的行业间尾部风险关联呈现出稳定的同质特征;行业自身风险和杠杆率等特征对行业尾部风险关联具有显著影响。

关键词: 实体行业, 尾部风险关联, 风险溢出, LASSO-CoVaR