Corporate failure results in enormous losses to most stockholders, especially consumers, suppliers, workers, financial institutions and many investors, and impacts significantly on the national economy. Corporate failure is manifested by a combination...
Corporate failure results in enormous losses to most stockholders, especially consumers, suppliers, workers, financial institutions and many investors, and impacts significantly on the national economy. Corporate failure is manifested by a combination of factors, rather than being influenced by one factor. Futhermore, since industrial structural constraints as macroeconomic variables affect the corporate's business strategy, financial structure, and profitability, etc. resulting in eventual effect on the probability of corporate failure. Therefore, it is necessary to include industry-related variables in business failure prediction model, in addition to the financial ratios. In order to meet these needs, this study was conducted on unlisted firms, especially SMEs, which included industry - related variables besides financial ratios as explanatory variables of business failure prediction model, and lacked previous studies. As for the methodology for the prediction of corporate failure, various techniques have been used such as discriminant analysis, logistic regression analysis (logistic regression: logit), probit, survival analysis, Artificial Neural Network (ANN), Case-Based Reasoning (CBR), Decision tree, etc. As for the explanatory variables of the forecasting model, accounting based measures, such as EBT and financial ratios are mainly used, while firm age, firm size, industry, region, credit rating, appropriateness of accounting information system, and macroeconomic indicators are used as non-accounting variables. In this study, we examined whether the possibility of failure is affected by profitability, liquidity, interest payment ability, financial stability, growth potential, activity, industry concentration and industry dependency. In addition, whether industry concentration and industry dependency moderate the effect of financial ratios on corporate failure is also examined. The dependent variable is a dummy variable that indicates whether the firm failed. According to the asset quality classification standard of the banking supervision regulation, ‘1’ is assigned to firms below ‘Fix’, while ‘0’ is assigned to others. And the independent variable is the financial ratio variable. Among the industry-related variables that are independent and moderating variables, the industrial concentration indicates the competition level of the industry. In this study, it is measured by the HHI data published by the Korea Fair Trade Commission. The industrial dependency indicates the extent to which producers in one market become buyers or sellers directly or indirectly in other markets. In this study, we measured customer dependency and supplier dependency separately according to Shih (2007). As for the control variables, we selected the firm size and age, and the firm size was measured by the natural log of total assets. The validation data used in this study are financial and industry-related data of unlisted SMEs excluding financial industry, and extracted from the internal database of "A" bank which is one of the domestic commercial banks. In this study, 100 normal companies and 100 insolvent companies were selected. The Fair Trade Commission has calculated the Hirschman-Herfindahl Index for each industry and has been compiling since 2006. The sample period is from 2010 to 2013 because it is now publicly disclosed to the index in 2013. The results of empirical analysis are as follows: The financial ratios are divided into six areas such as liquidity, profitability, financial stability, interest payment ability, activity, and growth potential. One of the highest t values per area, such as cash ratio, retained earnings ratio, SDBV, financial expenses to net sales ratio, turnover ratio of accounts receivable, and growth rate of equity capital, are selected. In the estimation of the regression model in the financial stability area, we selected SDBV instead of ratio of net worth to total capital, debt-equity ratio and debt ratio, which caused multicollinearity. According to a result of analysis using logistic regression on the data of the year of occurrence of the insolvency, the age showed the negative effect on the occurrence of the failure, while the size effect indicated by the amount of assets showed the negative direction. Among the independent variables, the financial expenses to net sales ratio has a positive (+) effect on the failure at the 1% of statistically significant level, and the SDBV has the positive (+) effect on the failure at the 10% level. According to a result of estimating the insolvent companies two years ago, the financial expenses to net sales ratio has a positive affect at the 1% level. As for a result of estimation three years ago, only the turnover of accounts receivable turns out to be statistically significant at 1% level. The industry concentration variable showed a positive (+) value at the 5% level, and both the customer dependency variable and the supplier dependency variable showed no statistical significance. The moderating effect was verified by creating a dummy variable that is 1 if the industry concentration, customer dependence, and supplier dependency are above the median, and 0, if they are below the median. The industry concentration dummy variable had a significant moderating effect only on the relationship between the financial expenses to net sales ratio and the failure probability.
it is shown that the customer dependency dummy variable has a significant moderating effect only on the relationship between the financial expenses to net sales ratio and the failure probability. On the other hand, the supplier dependency dummy variable has a significant moderating effect only on the relationship between the financial expenses to net sales ratio and the failure probability. The contribution of this study is as follows: First, SMEs have important implications for managers and policy making authority that financial ratios deeply related to the occurrence of insolvency are significantly different from those of large corporations. Second, since the financial ratios with high predictability change depending on the number of the years before the failure gets bigger, the possibility of failure can be predicted from a long-term perspective, based on changes in financial ratios. Third, we extended the time and spatial range of forecasting by predicting corporate failure by integrating micro-financial ratios and macro-industry-related variables.
However, as this study has the following limitations, it needs further study to complement this. First, as this study has limitedly collected data at one specific bank, it is limited to generalize the analysis results. Further research is required to collect data simultaneously from different banks' databases and to increase the likelihood of generalization of results. Second, because the sample of this study is limited to the manufacturing industry, it is difficult to apply it to the financial industry. Therefore, it is necessary to develop a model that extends to the financial industry, and to conduct a research that can analyze the difference between the financial industry and the manufacturing industry. Third, it is necessary to make a comparison with other methods in addition to logistic regression analysis.