In recent years,with the weakening of macroeconomic growth momentum and the negative impact of the Corona Virus Disease 2019(COVID-19),the company’s performance has faced greater downward pressure.Earnings downside risks have further affected the liquidity of bond issuers and bond redemption capabilities,intensified credit risk in the bond market,resulting in frequent bond defaults and severe bond market turmoil.Therefore,how bond rating agencies perform their duties as a "goalkeeper" in the bond market has attracted much attention.However,whether the bond credit report issued by bond rating agencies can fairly convey the bonds’ and bond issuers’ credit risk signals to external investors?Or whether the bond credit report is obscured or concealed due to the self-interested motives of bond rating agencies?There is no clear conclusion so far.The relevant evidence in the existing literature is mainly limited to the single dimension of credit rating.The credit text tone is an important component of bond rating reports,but there are relatively few studies on its signaling effect.In this context,this paper uses the bond credit report of A-share listed companies from 2009 to 2019 as a research sample to examine the correlation between bond issuers’earnings downside risks and bond credit risks,and examine the role of bond credit ratings and credit text tone in earnings downside risks transmission.The research results show that:(1)The earnings downside risk of bond issuers is significantly positively correlated with bond credit risk,so the earnings downside risk provides a new dimension to identify bond credit risk.After further decomposing the earnings downside risk,the operating cash flow downside in earnings downside risk and bond credit risk are significantly positively correlated,while the accrual downside risk is not correlated;(2)The bond credit rating is significantly negatively correlated with the earnings downside risk,proving that credit ratings can transmit earnings downside risk.The credit text tone is significantly positively correlated with earnings downside risk,proving that the bond rating agencies have the tendency to obscure the bond issuers’ unfavorable performance through credit text tone;In the group with lower bond credit ratings,the correlation between the text tone and the earnings downside risk is more significant.Therefore,in the worse bonds,the bond rating agencies will use rating texts with less regulatory constraints to obscure the unfavorable performance of bond issuers(3)After further grouping in terms of bond issuers’ financing constraints,it shows that the earnings downside risk transmission effect on bond credit ratings is stronger in the higher financing constraints group.The above research results are conducive to evaluating the changes in the company’s future operating performance and preventing credit risks in a timely manner.The research significance of this article lies in:(1)Innovate text analysis samples.This paper manually collected the credit rating reports of A-share listed company bonds from 2009 to 2019,and used machine-learning methods to extract each credit text tone.These manually collected credit rating report are unique in text analysis;(2)Enrich the research of machine learning methods in text analysis;(3)Deepen the research on earnings downside risks.This paper takes the earnings downside risks as the entry point for credit risk research,verifies the correlation between earnings downside risks and credit risk,provides a new dimension to identify bond credit risks,and further split it into the operating cash flow downside risk and the accrual downside risk for credit risk research.Besides,this paper innovatively examines the earnings downside risk transmission on bond credit ratings and credit text. |