특이적 충격이 비정상 판관비 측정치에 미치는 영향

초록

Idiosyncratic shocks are firm-specific characteristics that differ from the industry average due to a firm's business model, competition, and strategy. When earnings management is measured by using regression residuals under the assumption of industry–year homogeneity, there is a risk that idiosyncratic shocks may be mistakenly interpreted as earnings management, thereby introducing measurement errors. This study aims to suggest the necessity of controlling idiosyncratic shocks by analyzing the Type 1 and Type 2 errors of the original model for abnormal SG&A(selling, general, and administrative) expenses and the alternative model that controls idiosyncratic shocks. A simulation technique was employed, wherein 100 observations were randomly selected without replacement and validated over 250 iterations. Type I error testing involved the random extraction of 100 observations from both the full sample and a partial sample, while Type II error testing was conducted by adjusting the SG&A expenses of 100 randomly selected observations by 5% and 10%. The analysis revealed that the alternative model, which accounts for idiosyncratic shocks, exhibited lower Type I and Type II errors compared to the original model, suggesting that incorporating idiosyncratic shocks is crucial for enhancing the accuracy of earnings management measurement models. Among the original models, the abnormal SG&A expense measurement model proposed by Huang et al. (2020) demonstrated the lowest error rates. This study offers lower-error abnormal SG&A expense measures, assisting practitioners in identifying SG&A adjustments, thereby contributing to improved accounting practices, corporate valuation, and ethical responsibility. Moreover, it offers an academic contribution to the development of models in the earnings management research field.

키워드