A Study on the Prediction of the Level of Tax Avoidance Using Financial Characteristics

Previous studies on tax avoidance delve into the financial characteristics of firms that conducting tax avoidance, yet studies predicting the level of tax avoidance with financial characteristics are scarce. Financial ratios can be used as a source of information for various decisions including valuation of firm, credit assessment, and screening for lending decision, and which can be employed as a useful indicator of firm insolvency.
The methodology is summarized as follows: first, firms are categorized into high tax avoidance firms and low tax avoidance firms. We predict tax avoidance by dividing the total sample into two groups; the first group is used to estimate the discriminant function, and this function is used in the second group to estimate corporate tax avoidance, which is then compared to actual tax avoidance. In other words, we find the discriminant coefficients and significant financial ratios that influence corporate tax avoidance from the first group, then we use these values to estimate corporate tax avoidance in firms of the other groups. This estimated tax avoidance and actual tax avoidance are compared to measure the accuracy of the prediction of discriminant analysis result.
The results demonstrate that among financial characteristics, Inventory Assets to Current Assets (IACA), the Cash Structure ratio (CSR), Net Profit to Total Debt (NPTD), and the Labor Equiptment Ratio (LER) are found significant from our multi-step estimation method. Estimation using the selected financial characteristics can accurately forecast tax avoidance about 78% in the first group. When we implement the same variables in the second group, the predictions yield a substantially accurate prediction rate of 89%.
This study contributes to the literature by proposing an additional index to predict tax avoidance, the insolvency of a firm. We examine the difference in financial characteristic ratios between firms with levels of corporate