Skip to main content

Table 5 Regressions using an alternative measure of earnings management

From: Tone of language, financial disclosure, and earnings management: a textual analysis of form 20-F

Variables

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Dependent variable: AVDA

Positive

 

9.097***

       
  

(2.86)

       

Negative

  

1.603

      
   

(1.14)

      

Uncertainty

   

4.392**

     
    

(2.44)

     

Litigious

    

0.404

    
     

(0.46)

    

Strongml

     

6.183

   
      

(1.16)

   

Moderml

      

14.935**

  
       

(1.98)

  

Weakml

       

6.320**

 
        

(2.25)

 

Irrverb

        

3.560

         

(0.97)

Controls

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Firm

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Industry

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Year

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

N

449

449

449

449

449

449

449

449

449

Adj R2

0.337

0.349

0.337

0.344

0.335

0.337

0.341

0.343

0.336

F- statistics

8.49

8.62

8.26

8.51

8.21

8.27

8.40

8.46

8.25

  1. This table reports the OLS regression results using an alternative measure of earnings management (AVDA). The dependent variable is the estimates of discretionary revenues developed by Kothari et al. (2005). The sample includes 75 U.S.-listed Chinese companies during 2002–2014.The total number of observations is 449. The sample is obtained from the S&P Capital IQ and supplemented from the Wind Database. Positive, negative, uncertain, litigious, strong modal, moderate modal, weak modal, and irregular verbs are the eight categories of wordlist defined from the 2014 Master Dictionary of McDonald, and Loughran and McDonald (2011) word list. For the sake of brevity, we use Controls to represent all the control variables. All regressions include firm, industry, and year fixed effects. The t-statistics are in parentheses
  2. ***, **, and* denote significance at the 1%, 5%, and 10% level, respectively