서강대학교

초록/요약

There has been considerably numerous study results related to bankruptcy prediction so far. However, looking into the previous study papers, most are studies that do not have a classification according to industrial sectors, or although classified, includes all small firms with low financial reliability that are not liable to external auditing, or studies that are limited to manufacturing industry or construction industry. As these existing studies establish logistic regression model without industrial classification, based on small firms with low financial reliability and no external auditing as the sample group, it is assumed that there is a possible biased sampling problem. The major purpose of making bankruptcy predictions is to minimize the loss by defaults generated due to misjudgement of bankrupt firms as normal operating firms. In other words, the biased sampling problem may increase this kind of Type I error. Thus, placing importance on the independent variable, financial ratio reliability, this study selected only a total of 510 firms that are subjected to external auditing as its sample group, and performed individual logistic analysis for each sectors of manufacturing industry, construction industry and wholesale and retail industry in order to allow comparative analysis among same category of business. It is assumed that through this selection of sample group and classification by industry, the Type I error rate decreased and the overall predictive power increased compared to previous studies. As a result of the logistic analysis in this study, the statistically significant financial variable that affects insolvency of each industry was defined to be a total of 5 financial ratios, including ROA, capital adequacy ratio, total depts to total assets ratio, dept-service coverage ratio, total cash flow to dept ratio, for manufacturing industry, a total of 2 financial ratios, capital adequacy and total cash flow to dept ratio, for construction industry and a total of 2 financial ratios, net income to net sales ratio and dept to sales ratio, for wholesale and retail industry. Through comparative analysis according to industry, it is revealed that the financial variable that affects insolvency differs for each industry, and it explains which feature of the industry makes the corresponding financial variable different. Finally, to provide a significant financial variable that affects the firm insolvency, this study performed an individual logistic analysis for all three industries of manufacturing, wholesale and retail and construction, with only the firms that require external auditing. On this basis, the study raises a need to establish bankruptcy model per each industry in making bankruptcy prediction for small and medium-sized enterprises and presents a more efficient and significant bankruptcy prediction model such as application of financial ratio value for each industry.