Effect of Leverage on Real Earnings Management: Evidence from Korea

Round 1

Reviewer 1 Report

I would like to thank the authors for their effort to incorporate all comments raised. I believe the paper suffices for publication in its current form.

Author Response

Thank you very much for your previous comments to improve our paper.

Reviewer 2 Report

Thank you for having the opportunity to read your revision paper, which has a lot of improvement with the rigorous analysis. Here are my three concerns:

1/ On Table 6, the sample size of the suspicious firm is 682, but the sample size of the nonsuspicious firm is 5,525. For identifying the causality, I suggest applying coarsened exact matching or propensity score matching. 

2/ Thanks for adding 2SLS model to address the endogeneity issue. It seems that Size_Dummy is the instrument variable. Could you report the results of the first stage model and show it is a strong instrument variable? A weak correlation may provide misleading inferences about parameter estimates and standard errors.

3/ In the implementation part, how is the major finding on the relationship between leverage and EM related to the sustainability issues?

I hope that my comments help the authors improving the paper. I wish the authors good luck in improving the paper into a publishable one!

Author Response

Reviewer#2, Concern #1: On Table 6, the sample size of the suspicious firm is 682, but the sample size of the nonsuspicious firm is 5,525. For identifying the causality, I suggest applying coarsened exact matching or propensity score matching. 

=> The main concern of our paper is to test the relationship between leverage and REM, rather than the comparison between suspicious and non-suspicious firms. We have empirically examined whether leverage motivates managers to engage in REM activities. We aim to detect the managers' preferred quarter -seasonality- to conduct REM, given the firm's level.

=> CEM and PSM are two matching techniques for quasi-experimental study design (Wells et al., 2013) and the scope of our paper entails finance and accounting using financial statements of non-financial firms. A quasi-experimental design evaluates comparability between the treatment and comparison group based on objective metrics while the methodology of our study does not compare observations into distinct strata.

Reviewer#2, Concern #2: Thanks for adding 2SLS model to address the endogeneity issue. It seems that Size_Dummy is the instrument variable. Could you report the results of the first stage model and show it is a strong instrument variable? A weak correlation may provide misleading inferences about parameter estimates and standard errors.

=> First-stage and second-stage regression results are provided in section 4.3.3 two-stage least square regression analysis  (see pp. 16-17)

=> Size_Dummy is considered a strong instrumental variable since the covariance results are close to zero in the first-stage regression (see Table 8 p. 16)

Reviewer#2, Concern #3: In the implementation part, how is the major finding on the relationship between leverage and EM related to the sustainability issues?

=> Most important findings and implications for the relationship between leverage and REM are provided in sections 1 introduction (see pp. 2-3) and 5 conclusion (see pp. 17-18)