Why Does Aggregated Earning Surprise Convey Future Inflation Information?

The earning information of listed companies conveys future inflation news at the aggregated level,which has been discussed and empirically confirmed.But how this earnings information can be used to forecast the future inflation are still under discussion.In the paper,we attempt to analyze the mechanism of “bank credit” and provide empirical evidence from both the macro and the micro level.Referring from most of the literature,we use aggregated earning surprise to measure the companies earning information.Earning surprise refers to the deviation of the management or analysts forecasted earnings from the company's realized earnings.And aggregated earning surprise means total weighted average of earning surprise for all the companies.Firstly,we analyze the theoretical basis that aggregated earning surprise conveys future inflation news through the credit mechanism.The theory are as follows:(1)Enterprise earnings can affect the financing behavior of the company(the choice of internal and external financing),which affects credit demand;(2)Enterprise earnings can affect the company's investment decisions and investment needs,which in turn affects credit demand;(3)Enterprise earnings can affect the willingness of banks to lend to the company,which in turn affects corporate credit demand.Secondly,we establish VAR model which contains three variables,including aggregated earning surprise(AES),inflation(CPI)and control variable(GDP).Granger causality tests conducted on VAR models reject the null hypothesis that “Aggregated earning surprise is not a Granger cause of CPI inflation”,which implies that aggregated earning surprise conveys future CPI inflation information.Finally,we adopt the test of mediating effect,and examine the credit mechanism from the macro level and the micro level.The empirical process in the macro level is as follows:(1)Establish VAR model including aggregated earning surprise(AES),mechanism variable(credit demand,LOAN),and control variable(GDP).Granger causality tests conducted on VAR model reject the null hypothesis that “Aggregated earning surprise is not a Granger cause of credit demand”,that is,aggregated earning surprise has conduction to the intermediary mechanism variable(credit demand,LOAN);(2)Establish VAR model including aggregated earning surprise(AES),mechanism variable(credit demand,LOAN),inflation(CPI).Granger causality tests are conducted on VAR model.The results show that “Credit demand is not a Granger cause of inflation” is rejected.That is,the mediator variable has a conduction to inflation.It confirms the establishment of the “credit mechanism” from the macro level.At the micro level,we further test the impact of earning surprise on individual credit behavior of the company.We select non-financial Shanghai and Shenzhen A-share listed companies as a sample,and take the net increase of long-term borrowings as the agent of the company's credit demand,and adopt a fixed effect model.The panel regression shows that the company's credit demand is significantly negatively correlated with earning surprise,that is,the unexpected earnings will reduce the company's future bank loans.It confirms that earning surprise will affect the company's credit behavior.The empirical results show that aggregated earning surprise significantly conveys inflation information through the credit demand mechanism.The main innovation and contribution of the thesis is that it explains the credit mechanism of aggregated earning surprise conveys future inflation information.And it provides direct evidence from the macro and micro levels,it has marginal contribution to the existing research literature.