![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
This study employs the panel threshold model to reexamine the non-monotonic relationship between CEO stock-based compensation and firm earnings across various firm-size conditions. The feasibility of the model is tested using data for US non-financial firms from 1993 to 2005. Our empirical results indicate that while a positive relationship between the CEO stock-based pay and earnings is presented for small-size firms, a negative impact of CEO stock-based compensation on earnings is shown when large-size firms are concerned. Further, the longstanding puzzle of whether the CEO stock-based pay could enhance earnings among earlier studies could be satisfactorily explained by our empirical results.
Acknowledgements
The authors acknowledge the two anonymous reviewers for their very helpful comments and gratefully acknowledge the funding from the National Science Council of Taiwan (NSC96-2416-H-006-023-MY3).
Notes
See Li and Miu Citation35 and Li Citation31 Citation33 for the related discussions.
The estimation procedure for the fixed effects model is well documented in the literature, and therefore, this study omits any discussion of this.
Following Hansen Citation23, the errors in the panel threshold model are assumed to be independently and identically distributed (i.i.d.). Under i.i.d. errors, Hansen Citation23 defined the panel threshold model estimator and proved its consistency and asymptotic normality.
The concept of threshold model has generated considerable applied interest. See Li Citation32 Citation34 for a summary.
As part of this study's focus was on the non-uniform relationships between y
it
and x
it
(i.e. the β parameters), we take the “group difference” between variables and redefine EquationEquation (3) to have EquationEquation (4)
, in which the intercept terms (i.e. μ
i
) are included. Further dealing with the dynamic intercept terms by allowing that they change over time is a valuable research direction.
The procedure for the bootstrap method is well documented in the literature. Therefore, this study omits any discussion of it and refers readers to Hansen Citation22 Citation23 for further details.
We use firms listed in US S&P 500 since their accounting and financial data are arguably of relatively reliable.
The study of Hansen Citation23 also suffers the balanced samples requirement and thus selects the relatively small subset of US firms. Including more firm samples by dealing with the unbalanced panels is a valuable direction for future research.
The list of all 64 companies selected for this study is given in the appendix.
The F statistics and their corresponding p-values are calculated using the bootstrap method proposed by Hansen Citation24.
Our findings suggest that the stock-based/total variable is significantly positive (negative) in regime I (III). We ever re-run our model by re-defining the threshold variable using the relative measure of firm size (e.g. individual firm's asset value is normalized by the median asset value of all firm samples in each specific year). Our results show that the positive CEO stock-based pay-to-earnings relationship is flattened out as the large-firm regime is examined, whereas this process does not reverse. The authors acknowledge the anonymous reviewer for this comment.
It should be noted that salary is the most important pay component of cash compensation.
Moreover, Gertler and Gilchrist Citation20 find that small firms are more constrained when the monetary policy is tightening.