세무상 이월결손금 공제규모 축소와 이익조정 = Tax Net Operating Loss Limitation and Earnings Management

국문 초록 (Abstract)

본 연구는 2014년부터 2020년까지 KOSPI 기업을 표본으로 하여 세무상 이월결손금과 이익조정의 관계를 연구하였다. 아울러 해당 시기 중 개정된 세무상 이월결손금의 공제한도 축소규정이 이월결손금과 이익조정과의 관계에 추가적인 영향을 미치는지 여부를 확인하고, 이러한 영향이 대기업과 중소기업에 차별적으로 발생하는지 검토하였다. 본 연구의 분석 결과는 다음과 같다. 첫째, 분석대상기간 동안 이월결손금은 발생액이익조정과는 양(+)의 유의하지 않은 관계를, 실질이익조정과는 양(+)의 유의한 관계를 보였다. 둘째, 이월결손금 공제한도 축소가 대기업과 중소기업의 이월결손금-이익조정의 관계에 차이를 발생시키지 않았다. 즉, 공제한도 축소가 이를 적용받은 대기업의 이익조정 행태에 평균적으로 유의한 영향을 미치지 못하였음을 확인하였다. 셋째, 강건성 분석을 통해 공제한도 축소가 이월결손금 규모가 크면서 (작으면서) 영업성과가 좋은 (낮은) 기업의 이익조정 행태에 유의한 영향을 미침을 확인하였다. 이는 공제한도 축소가 특정 상황의 기업에만 영향을 미치기 때문에 전체 표본을 이용한 분석에서 공제한도 축소의 유의한 영향을 관측할 수 없었음을 의미한다. 본 연구는 이월결손금 공제기한 연장의 영향을 통제한 후 이월결손금 공제한도 축소의 영향을 분석한 최초의 연구인 점에서 의미를 갖는다.

다국어 초록 (Multilingual Abstract)

We aim to examine the effects of the introduction of the tax net operating loss (NOL) limitation on the relationship between tax NOL and earnings management. Several studies conducted during the late 2000s and early 2010s documented a positive relationship, as the existence of tax NOL carryforward promotes earnings management to utilize NOL within the permitted period (Park and Song 2008; Kim and Lee 2009; Park and Yook 2011a; 2011b). However, these studies generally examined selected facets of this relationship; only a few, such as Song and Shim (2013) and Cho and Lim (2016), sought to expand research by identifying factors influencing the relationship between tax NOL and earnings management. Choi et al. (2019) cited the difficulty of collecting tax NOL data as the primary reason for the lack of relevant research. Consequently, studies of tax NOL have declined in recent years. Despite the decline in studies, there have been enactment or revision of several tax laws related to tax NOL. First, the tax NOL carryforward period has been extended. Any tax NOL generated on or after (before) the taxable year of 2009 (2008) is allowed to be offset within 10 (5) years. The carryforward period is further extended to 15 years for tax NOL generated on or after the taxable year of 2020. These extensions of the NOL carryforward period may weaken the ‘motivations’ for earnings management to utilize tax NOL. Second, the tax NOL limitation was introduced. Any tax NOL of firms other than small and medium-sized firms (hereinafter “large firms”) generated on or after the taxable year of 2016 is allowed to be offset by up to 80% of the total tax base. This limitation decreases gradually to 70% for 2018 and 60% for taxable years on or after 2019. The introduction of the NOL limitation represents a restriction on the NOL carryforward period extension. Because previous analyses of the relationship between tax NOL and earnings management used financial data ending on or before 2014, the results do not appropriately reflect the effects of tax law changes occurring after 2009. Our study uses financial data from 2014-2020, a period in which all tax NOL is subject to the carryforward period of 10 years. Thus, this study is the first to examine the relationship between tax NOL and earnings management after controlling for the effects of the extension of the NOL carryforward period, and the analysis yields three findings. First, we find a positive but insignificant association between tax NOL and discretionary accrual (DA), inconsistent with previous studies. However, we observe a positive and significant association between tax NOL and real activity earnings management (REM). These results confirm that the influence of the earnings management behavior of firms decreased after the extension of the NOL carryforward period. Choi et al. (2019) explain that the cost of DA is higher than that of REM because DA involves changes in accounting policy, which will be accompanied by potential legal or political costs. Our results support this view. Second, we find no significant impact of the tax NOL limitation on the relationship between tax NOL and earnings management. Specifically, the relationship between tax NOL and earnings management for large firms (firms limited in NOL offset) is not significantly different from that for small and medium-sized firms (firms not limited in NOL offset). We attribute this lack of significance to differences in the weight of the influence of the NOL limitation among large firms. That is, the NOL limitation only affects large firms under particular circumstances, and thus statistical significance is not observed on average. Third, based on the conjecture that firms with a larger (smaller) amount of tax NOL and higher (lower) operating performance will be affected by the introduction of the NOL limitation to a greater extent than those with a larger (smaller) amount of tax NOL and lower (higher) operating performance, we re-examine the effects of the introduction of the NOL limitation on firms that are expected to be particularly affected by the NOL limitation. The results show that the positive relationship between tax NOL and earnings management becomes stronger for these firms after the introduction of the NOL limitation. Therefore, we confirm that the introduction of the NOL limitation selectively influences firms under certain conditions. Finally, we further examine earnings management in the year prior to the introduction of the NOL limitation. No earnings management behavior is detected prior to the introduction. We also identify that the recognition of deferred assets in NOL does not influence the relationship between tax NOL and earnings management. Overall, our findings imply that tax law enactments or revisions related to tax NOL have a limited impact on decision-making by management. Therefore, we contribute empirical evidence and the political implication that the recent changes in tax laws related to tax NOL have achieved their policy goals of preventing significant tax planning by firms through earnings management.