基于ST预测的财务困境测度与股票横截面收益

A Measure for Financial Distress based on ST Predictive Model and the Cross-section of Stock Returns

本文利用特别处理(ST)这一我国资本市场特有的现象和机器学习的方法构造了一个新的刻画上市公司财务困境的测度指标,并研究了财务困境与股票横截面收益的关系。首先利用机器学习的方法建立了上市公司ST的预测模型;再通过预测模型构造上市公司被特别处理的预测概率PrST,并以PrST刻画上市公司的财务困境;最后利用这一新的财务困境测度指标研究了PrST对公司股票收益的影响。研究发现:(1)事件分析结果显示,市场对特别处理公告做出显著的负面反应;(2)以预测模型得到的PrST构造资产组合,PrST高的资产组合收益率显著低于PrST低的资产组合,一个多空股票组合可以产生平均每月约-1.40%的超额收益;(3)构造PrST因子,Fama-MacBeth回归也得到显著的结果。综上,本文的实证研究表明:构造的财务困境指标PrST能有效解释股票横截面收益,PrST因子是一个被定价的风险因子。

Financial distress may lead to company bankruptcy in severe conditions.In fact,financial distress is a gradual process and it is predictable.Predicting financial distress is of great significance.On one hand,effective prediction of financial distress is beneficial to the protection of investors’rights and interests.On the other hand,financial distressed companies are facing greater business risks,and it’s important to study whether the financial distress risk is effectively priced by the stock market.In China’s stock market,to better remind investors of market risks and guide investors to invest rationally,the China Securities Regulatory Commission requires stock exchanges to implement special treatment(ST)for listed company stock transactions in abnormal conditions.This is a policy unique to China’s stock market.Special treatment(ST),a unique phenomenon in China’s stock market,and machine learning methods are used to construct a new measure that characterize the financial distress of listed companies.The relationship between financial distress and the cross-section of stock returns is also studied.The sample period is from January 1,2003 to December 31,2018.A machine learning model is built to forecast the ST of listed companies.Then a prediction probability index,PrST,is constructed which can characterize the financial distress of the public companies.Finally,the cross-sectional stock returns is studied with our new measure of financial distress.It is found that:(1)the event study shows that the market has made significant negative reactions to ST;(2)the portfolio with high PrST has significantly lower return than the portfolio with low PrST,and a long-short stock portfolio earns an average of 1.4%abnormal return per month,(3)Fama-MacBeth regression also yields significant results.In summary,this paper offers empirical evidence that our new measure of financial distress,PrST effectively explain the cross-section of stock returns.The PrST factor is a priced risk factor.This study sheds new light on the research of financial distress risk and asset pricing in China.