법인세비용 계산의 복잡성과 법인세비용 계산시의 시간제약, 미래이익을 추정하는데 있어서의 자의성은 경영자와 감사인 및 주주 사이에 정보불균형을 가져오며, 따라서 경영자가 목표이익을 달성하기 위하여 법인세비용을 마지막 기회로 이용하는 것이 가능하다(Dhaliwal et al. 2004). 본 연구에서는 보고이익을 공시하기 전에 이익조정 전 이익(당기의 연간 세전이익을 이용한 방법과 3분기까지의 주당순이익을 이용하는 방법으로 측정)이 목표이익(3분기말의 재무분석가의 이익예측치와 전기 주당순이익으로 측정)에 미달된 기업들이 4분기에서 법인세비용을 이용하여 이익조정을 하는가를 알아보았다. 재무분석가의 이익예측치 자료가 이용가능한 2000년부터 2006년까지의 기간을 분석기간으로 하여 금융업을 제외하고 12월 결산법인을 대상으로 최종표본 751개 기업-연도 자료를 이용하여 분석하였다. 논문접수: 2008. 7 게재확정: 2009. 2 *2008년도 한국회계학회 하계국제학술대회에서 유익한 토론을 해주신 김상헌(단국대), 고종권(한양대) 교수님께 감사드리며, 논문심사과정에서 유익한 조언을 해 주신 두 분의 심사위원께도 감사드립니다. 본 연구는 숭실대학교 교내연구비 지원으로 이루어졌음. **충북대학교 경영학부 조교수, Assistant Professor, School of Business, Chungbuk National University (E-mail: parkjil@chungbuk.ac.kr) *** 숭실대학교 경영학부 부교수, Associate Professor, School of Business, Soongsil University (E-mail: kajeon@ssu.ac.kr) 본 연구의 실증분석결과는 다음과 같다. 첫째, Dhaliwal et al.(2004)의 방법과 동일한 방법을 이용하는 경우에는 3분기말에 발표된 목표이익에 미달한 기업들과 4분기의 법인세비용을 이용한 이익조정 간에는 관계가 없는 것으로 나타났다. 둘째, 본 연구에서 제안한 방법을 이용하여 분석한 경우에는 목표이익에 미달한 기업들이 3분기까지의 유효법인세율(effective tax rate)보다 4분기의 유효법인세율을 낮추는 것으로 나타났다. 이는 경영자가 법인세비용 을 낮추어 이익을 상향조정하는 것을 의미한다. 즉 Dhaliwal et al.(2004)의 방법을 개선하여 본 연구에서 제안된 방법을 이용하여 분석하는 경우에서 Dhaliwal et al.(2004)의 연구에서처럼 3분기에서 목표이익에 미달한 기업들이 4분기의 법인세비용을 이용하여 이익조정을 한다는 결과를 발견하였다. 본 연구에서는 Dhaliwal et al.(2004)에서 이용된 방법보다 개선된 방법으로 측정함으로써 경영자가 목표이익에 미달한 경우 법인세비용을 이용하여 이익조정을 한다는 실증분석결과를 보여주었다는 점에서 의의가 있다. 경영자들이 법인세비용을 이용하여 이익조정을 한다는 본 연구의 결과는 투자자나 채권자, 감독당국 등이 기업(특히 실제 이익을 공시하기 전에 목표이익에 미달한 기업)의 재무제표를 분석할 때 법인세비용에 특히 주의를 기울여야 함을 제안하고 있다는 점에서 의의가 있다.
When the complexity of estimating tax expense and the timing of the tax accrual prior to earnings management is considered, Dhaliwal, Gleason and Mills (2004) (hereafter DGM) suggest that tax expense is a powerful and underexplored context for studying earnings management. DGM (2004) establish that, relative to firms that beat the forecast, firms which would miss analysts’ consensus forecast of annual earnings per share (EPS) are more likely to decrease tax expense to meet or beat the forecast. Therefore, DGM (2004) examine whether firms manage earnings through decreases in effective tax rates (ETRs) between the third and fourth quarters. They assert that changes in ETRs are consistent with earnings management because tax expense is one of the last accounts closed in determining reported earnings. DGM (2004) insist that although many pre- tax accruals must be posted in the year-end general ledger, managers estimate and negotiate tax expense with their auditors immediately prior to earnings announcement. So, We examine whether firms manage tax expense to reach earnings targets by applying DGM’s methodology and our improved methodology to Korean firms. The sample consists of non-banking firms (751 firm-years) with a December fiscal year-end listed on the Korean Stock Market over 2000-2006. The data on analysts’s forecasts is collected from the Fn-DataGuidePro Database. Therefore, the sample is limited to firms to be able to collect analysts’s forecasts data from Fn-DataGuidePro Database. We collect other firm-specific financial data from KIS-VALUE Ⅱ database. We support the DGM (2004) arguments that tax expense represents an opportunity for firms to earnings management using changes in ETRs. We conduct the different three measure in regard to targets earnings of empirical tests as DGM (2004) to investigate the sensitivity of DGM’s results to research design choices. We use the consensus earnings forecasts and earnings of the previous year as our proxy for targets earnings. Also, we define ETRQ4 - ETRQ1∼Q3 as the change in annual ETR from the accumulated three quarters(from first to third quarter) to fourth quarter and use this as our measure tax expense management. In our hypothesis, we expect ETRQ4 - ETRQ1∼Q3 to be negatively related to how much the firm would have missed the consensus forecast(or targets earnings). We extend DGM (2004) by exploring a plausible additional explanation for third-to-fourth-quarter decreases in ETRs that they acknowledge as a limitation of their study. DGM (2004) use analysts’ consensus forecast as targets earnings and current annual pre-tax earnings as unmanaged earnings. But we use analysts’ consensus forecast and previous annual earnings as targets earnings and current annual pre-tax earnings and first∼third-quarter earnings as unmanaged earnings. Managers have strong incentives to meet or beat three benchmarks(to avoid losses, earnings decreases, and missing analyst expectation). DGM (2004) focus on a benchmark(to avoid missing analyst expectation), but we consider two benchmarks (to avoid missing analyst expectation and earnings decreases). We don’t consider the benchmark to avoid losses because loss firms have negative ETRs and negative ETRs don’t have economic meaning. Specifically, we investigate the extent to which greater first∼third-to-fourth-quarter ETR reductions are associated to firms that would miss their consensus earnings forecasts(or targets earnings) absent ETR changes relative to firms that would otherwise reach or exceed these income goals. We posit that firms may use from first to third-to-fourth-quarter ETR changes to reduce reported income when earnings are expected to meet or exceed analysts’ forecasts. In our empirical tests, in contrast to DGM's results when we use DGM (2004) methodology, we find no statistically significant association between analysts' consensus forecasts (a proxy for target earnings) and changes in ETRs (last chance earnings management). However, we recalculate two measures as a proxy for missed targets earnings and find the negative association between the first∼third- to-fourth-quarter ETR changes and missing income goals absent ETR changes. Our study implies that DGM's methodology is not proper to Korean Stock Market and DGM’s results are sensitive to research design choices. Overall, our results suggest that, for firms that would miss consensus earnings forecasts in the absence of ETR management, the first∼third-to-fourth-quarter ETR changes are associated with greater reductions in ETRs between the first∼third and fourth quarters. We also find consistent support for the DGM (2004) argument that firms use ETR as a last-chance earnings management technique, having controlled for firms’ manipulation of other accounts. These results are robust to the inclusion of other common earnings management variables, including total accruals and deferred taxes. Our research contributes to the literature in several ways. First, we provide information to investors and policy makers about the extent to which reported tax positions are consistent with tax planning and earnings management by investigating whether income tax expense is commonly used to meet or beat earnings targets. Second, consistent with Healy and Wahlen (1999), we contribute to earnings management research by focusing on specific accruals, which improves the likelihood of discerning between earnings management and effective tax planning as the reason for decreases in ETRs. Like DGM’s assertion by studying the tax expense in total, rather than narrow components of deferred tax expense, we think that our results provide general evidence that reported taxes are used to manage earnings. Finally by improving DGM's measure we find that the first∼third-to-fourth-quarter ETR changes are negatively related to income goals absent ETR changes.
