Endüstri işletmeleri için mali başarısızlık tahmini-çok boyutlu model uygulaması

Abstract

SUMMARY This study aims to predict the financial failure by using multivariate predictive models. The literature survey covered in the second section of this study has shown that the multivariate predictive models have been used extensively and have been the most accurate in predicting financial failure. On the other hand, the studies examined in the literature survey indicate that the models have their own limitations. This study focuses on developing a model which tries to minimize the shortcomings revealed in the literature survey. For this purpose, data of the samples were obtained from `Capital Market Committee`. The sample firms whose financial tables are included in this study are as follo.ws: 25 failed firms and 35 non-failed firms for one year before financial failure ; 23 failed and 35 non-failed firms for two years before financial failure and 19 failed and 35 non-failed firms for three years before financial failure. Logit, Probit, Multiple Regression Models (MRM), Multiple Linear Discriminant Analysis (MtDA) and Multiple Quadratic Discriminant Analysis (MQDA) were used in the analyses including 23 financial ratios. The following findings were obtained at the end of the analyses. The explanatory power of financial ratios in predicting -202-the financial failure one, two and three years before the financial failure are statistically important. The selection of prior probabilities in MLDA and MQDA has an important impact on the accuracy of the model and the prior probabilities provide a flexibility in developing a model satisfying the objectives of the financial analysts. It has been showed that the validity tests ignored in the previous studies are actually important and should be carried out in such a case, No great difference between the accuracies of MLDA and MQDA has been observed. It is observed that the accuracies of MLDA and MQDA are not better than that of MRM. Almost all the analyses comparing the accuracies of the models have shown that logit and Probit have the best results in every comparison. The accuracies of these two models in predicting the financal failure one, two and three years before financial failure are 90.1 %, 86.2 % and 87 %. Besides, these findings are supported by the validity and control tests used for these models. The findings related to financial ratios can be summarized as follows: Liquidity, activity and financial ratios have a multi- collinearity relationship with the profitability ratios, and -203-if the necessary number of ratios from these three groups are included in the model then there is almost no need for the profitability ratios. The following ratios were included in the models predicting the financial failure. The model for one year before financial failure :Totel Debts/Total Assets, Current L iabi li ties /To tal ftsssets, Gross Profit/Sales and Earnings Before Taxes /Stockholders ' Equity. The model for two years before financial failure:Cash and Marketable Securities/Current Liabilities, Total Debts/Total Assets, Sales/Total Assets and Gross Prof it/Sales. The model for three years before financial failure : Total Debts /Stockholders ' Equity.Earnings Before Interest and Taxes / Interest Expense, Sales/Stockholders ' Equity, Operating Profits/ Sales. For one year before financial failure, total debts/ total assets, earnings before taxes/ stockholders ' equity ; for two years before financial failure, total debts/total assets, sales/total assets; for three years before financial failure, total debts/stockholders ' equity, earnings before interest and taxes /interest expenses were found more important than other ratios in the models. The univariate models including only one of these ratios distinguished failed firms from non-failed firms with an accuracy of approximately 70 percent. These high accuracy scores -20 4-encourage the use of univariate models in predicting the financial failure. Other important findings of this study are the determinan- tion of the consistency of financial ratios in predicting financial failure and the determination of the leverage (financial ) ratios as the most important ratios in predicting the financial failure. As a result, in this study, the importance of financial ratios in predicting the financial failure has been shown. The benefit of these financial ratios can be maximized if they are included in the use of multivariate predictive models.. This study whose main features and findings were summarized above was carried out according to the following plan. The first section of this study is devoted to the definition of financial failure and the causes of financial failure. The importance of predicting financial failure are discussed in detail. The second section of the study covers the literature survey. In this section, a new classification based on different criteria has been developed. The third section of the study, the largest part, covers the methodology and the research. At the end of the study, the main results, deductions and suggestions are recovered in details. -205-