Abstract: This paper extracts common components in quantiles of stock market returns using a high-dimensional factor model to evaluate the tail systematic risk. It investigates the macro sources of the tail systematic risk by using regularized regressions and explores the effects of firm characteristics on the heterogeneity of the sensitivity of the tail systematic risk to macro risk factors via cross sectional regressions. The results show that the upside and downside tail systematic risks are asymmetric. The upside risk is more related to the increase in market uncertainty, has a higher volatility and a more complex relation with macro risk factors compared to downside risk. The asymmetry becomes more significant when the quantiles move to tails. Firm characteristics do have an explanatory power on the heterogeneity of the sensitivity of the tail systematic risk to macro risk factors, with the firm size, the debt asset ratio, and the self-sustainable growth rate being the most influential characteristics.

Key words: quantile factor analysis, tail risk, systematic risk, stock market return

摘要： 利用高维分位数因子模型提取个股分位数变化的共同成分以衡量尾部系统风险，采用正则化回归分析尾部系统风险的宏观来源，通过截面回归探讨企业特质对尾部系统风险受宏观风险因子影响差异的作用。研究表明，中国股市尾部系统风险具有较强的高低位不对称性，高位尾部风险更多地与市场不确定性增大相联系，具有更强波动性且受宏观风险因素影响复杂度更高。尾部系统风险的不对称性随极端条件的加剧增强。企业特征能够对尾部系统风险受宏观风险因素影响的差异进行解释，企业规模、资产负债率及可持续增长率是影响这一差异性的重要因素。

关键词: 分位数因子分析, 尾部风险, 系统风险, 股市收益率