Improving Predictive Ability of Financial Ratios: The Application of Log Transformation

Abstract

This study aimed to investigate whether log transformation can help improve predictive ability of financial ratios in Thailand context. Four DuPont-based financial ratios of 104 (using 3 years data), 93 (using 5 years data), and 81 (using 8 years data) Thai large capitalization listed companies are used in this study. Applying yearly data, the four ratios included net profit margin, total asset turnover, equity multiplier, and return on equity. The study period is between 2012 and 2019, totaling 8 years. Statistics used included mean, standard deviation, skewness, kurtosis, logarithm, and multiple regression. The findings suggest that log transformation can significantly improve predictive ability of financial ratios. Therefore, financial researchers and/or financial analysts should initially examine whether financial ratios are normally distributed before testing hypothesis. Also, the researchers and/or the analysts should not overlook financial ratios as a comprehensive financial tool to help maximize stockholders’ wealth.