세무위험 측정치에 대한 문제점 고찰

[Purpose]Many prior literatures investigated tax risk measured by standard deviation of ETR (Hutchens and Rego(2015), Drake et al.(2015), Guenther et al.(2017)). We raise the question whether the measurement of ‘tax risk’ has the same meaning as the tax risk that prior literatures generally think about. Through various approaches, this study attempts to analyze the relevance or differences with economic tax risk. [Methodology]We adopt the research tool following Dyreng et al.(2008), Hanlon and Heitzman (2010) and Henry and Sansing (2014, 2018) and review following research subjects. [Findings]First, we analyzes two problems (truncation and winsorization) related to the data pointed out as the problem of this tax risk measure. As a result of the analysis, it was confirmed that the tax risk measure can use only a sample of 70% of the other measures that present tax avoidance using the ETR. In addition, in the case where the annual ETR was adjusted between 0 and 1 and the annual ETR was not adjusted, the average match ratio was 90.2%. Second, in order to identify the characteristics of variables, the pretax net income and the ETR of companies with high and low tax risks were analyzed, and then the predictability of the variables was confirmed. As a result of the analysis, it was found that companies with high tax risks had large changes in pretax net income or ETRs. Additionally, the predictability of tax risk was significant only in tax risk calculated from adjusted annual ETRs. Finally, the relationship between financial characteristics and tax risk was examined to identify the financial characteristics. The result shows that the companies that can calculate tax risks are superior companies with large total asset and low debt ratios. [Implications]Through this study, the characteristics and meaning of the tax risk measurement values ​​were clarified and the possibility of interpretation errors of the variables was noticed.