본 연구에서는 기업의 최고경영자 교체시에 외부영입한 경우와 내부에서 승진한 경우에 이익조정 정도에 차이가 있는지를 분석하였다. 또한 이러한 영입경로(외부영입 또는 내부승진)에 따른 이익조정 정도의 차이가 경영자 교체사유에 의해 달라지는지를 살펴보았다. 본 연구에 사용된 표본은 2001년부터 2005년 사이 최고경영자의 교체가 이루어진 총 239개의 기업-연도 표본이다. 실증분석에서 이익조정의 대용변수로는 경영성과를 통제하여 계산한 성과대응 재량적 발생액을 사용하였다. 실증적 분석의 결과는 다음과 같다. 우선 재량적 발생액의 절대값을 이용한 분석결과 외부영입한 경영자가 경영자 교체연도에 이익조정을 더 많이 수행한다는 것이 발견되었다. 이익조정의 방향성을 고려한 추가분석에 따르면, 외부영입과 내부승진 경영자 사이에 이익을 늘리는 방향인 양(+)의 이익조정 수준에는 유의적인 차이가 존재하지 않았으나, 이익을 줄이는 방향인 음(-)의 이익조정 수준에는 차이가 있었다. 즉 외부영입한 경영자가 더욱 적극적으로 음(-)의 이익조정을 실시한다는 발견이다. 그러나 경영자 교체 직전연도나 교체 직후연도에는 양 집단 간에 유의적인 차이가 전혀 발생하지 않았다. 이는 외부에서 영입한 경영자가 취임할 경우, 상대적으로 미래 기간의 이익을 높이기 위하여 현재의 이익을 낮추는 이익조정을 실시하여 취임연도에 저조한 경영성과를 보고할 가능성이 높다는 것을 나타낸다. 이에 반해 내부승진한 경영자는 전임 경영자와의 긴밀한 유대관계 때문에 이익을 낮추는 이익조정을 적극적으로 실시한 결과 과장되어 나타난 부진한 경영성과를 전임 경영자의 잘못 때문이라고 책임을 전가시키기가 힘들다고 할 수 있다.
추가적으로 경영자의 교체이유를 살펴본 결과, 경영자가 비정상적으로 교체되었을 경우 그리고 전연도의 경영성과가 부진했을 경우 더욱 더 이러한 추세가 뚜렷하게 나타났다. 이는 전임 경영자가 부진한 업적 때문에 문책성 교체를 당했을 가능성이 높은 표본들의 경우 외부영입 된 신임경영자가 이익을 낮추는 이익조정을 교체연도에 적극적으로 실시한다는 것을 의미한다.
This study examines the association between the Chief Executive Officer (CEO, management) changes and the magnitude of earnings management of the firm. Especially, we examine if the types of management changes (either internal CEO promotion or external CEO recruitment) influence the magnitude of the earnings management. We define internally promoted new CEO as an internal management change and externally recruited new CEO as an external management change. In addition, we examine if the observed reason for the CEO changes is associated with the magnitude of the earnings management.
Although there have been voluminous prior studies which investigate the magnitude of earnings management at the time of CEO changes, only the study of Strong and Myers (1987) and Lee et al. (2007) investigated this issue. However, they fail to find any consistent and significant association between the two, possibly due to several methodological limitations. This study revisits this issue after correcting those methodological problems. In addition, this study extends the scope of the research to the reason for the CEO changes: whether the reasons for the CEO changes are associated with the magnitude of earnings management by new CEO.
On the one hand, it is possible that externally recruited CEO is more likely to take a big bath or manage earning downward at the year of appointment in order to lower the reported earnings. It is because he could blame inappropriate management of previous CEO for the poor performance of the year more easily. In contrast, internally promoted CEO may have difficulty in taking a big bath or managing earnings downward because he or she knows the former CEO well and has maintained close relationship with the former CEO. Thus, it may be difficult for the new CEO to blame the former CEO for the poor firm performance caused by the big bath or downward earnings management.
With 239 CEO changes data collected over 2001-2005 periods from Korean listed companies, we empirically examine these predictions. The samples used in our analyses are taken from the study of Lee et al. (2007) but we remove 11 observations from their samples judged as outliers. Our sample is composed of all non-financial industry firms listed in Korean Stock Exchange. We use performance-matched discretionary accruals developed by Kothari et al. (2005) in order to proxy for the level of earnings management. During the sample selection process, we delete samples that the absolute value of the discretionary accruals is greater than 0.5 as outliers. Prior study of Ball and Shivakumar (2008) reports that it is very important to remove the effect of outliers in the research that uses the measure of discretionary accruals. Ball and Shivakumar (2008) suggest that Teoh et al.’s (1998) study fails to control outliers and their results do not hold when Ball and Shivakumar (2008) remove a few outliers from the analyses.
The empirical findings are summarized as follows: First, using the full sample, we document that earnings management is more likely to occur at the time of management changes (both internal and external management changes). This finding is consistent with that in prior research of Lee et al. (2007) which use Korean data, and Murphy and Zimmerman (1993) and Pourciau (1993) which use the data of U.S. Second, using the full sample, we document a significant positive association between the level of earnings management and external management changes, compared with internal management change. Third, we separate the full sample into two groups of sub-samples based on the direction (upward (i. e., income-increasing) or downward (i.e., income decreasing)) of earnings management. When we use the sub-sample which adopts income-increasing discretionary accruals, we do not find any significant differences between the samples having external and internal management changes. However, when we use sub-sample which adopts income-decreasing discretionary accruals, we find that firms with external CEO changes are more likely to engage in income-decreasing earnings management, confirming the prediction that externally recruited CEO is more likely to deflate earnings in order to take a big bath or manage earnings downward. Further tests reveal that these findings are mostly due to the cases when the CEO changes occur in abnormal situation, for example, when poor financial performance of firms might cause the CEO change or when former CEO does not hold any position such as advisors or honorary president within the firm at all. In these cases, it is more likely that former CEO is replaced due to poor firm performance, not due to promotion or happy retirement.
Similar to our study in part, Lee et al. (2007) investigate if there is any significant difference in terms of the magnitude of earnings management between internally promoted and externally recruited CEOs. They fail to find any significant association between the two, which is different from this study. We investigate what causes the differences and find out that they fail to correct or remove a few outliers in the discretionary accruals measure. When we judge outliers by using various criteria, our results are robust that externally recruited CEOs tend to manage earnings more. Our results which are different from those in Lee et al. (2007) suggest the importance of the need to control for the influence of outliers in discretionary accrual estimation processes. Our results suggest the importance of outlier controls in empirical analyses using ratio variables generated from complicated estimation procedures.
This finding provides valuable insights to regulators, academics, firms, employees as well as many related parties into the behavior of new CEOs with respect to their motivation to manage earnings.
