Bankruptcy Prediction : Analyses of Cash Flow Models

概要

論文の詳細を見る
The purpose of this study is twofold. One is to evaluate the effectiveness of a bankruptcy prediction model when the Statement of Financial Accounting Standards No.95,the Statement of Cash Flows, is applied to Japanese corporations. The other is to demonstrate management decisions in bankrupt corporations. A statement of cash flows is a hot issue in Japan, because it is scheduled to come into effect in FY1999 in our country. Meanwhile, a bankruptcy prediction is another hot issue, because we are experiencing the most bankruptcies we have ever seen. Utilizing cash flows and cash flow components, twenty-eight models are built via the use of logistic regression analyses in a variable selection method. The empirical results of these models have proved their efficiency as a bankruptcy prediction with over 90% accuracy. Logistic regression statistical technique in a variable selection method chooses variables suitable to data and constructs a model. The selected variables in models function to predict bankruptcy. The analyses of the selected variables demonstrate a maneuver in receivables, an aggressive investment spending and a heavy dependence on the stock market of bankrupt corporations. The study also reveals that management three years prior to bankruptcy is a key, although most previous researches predict it five years before.
関西学院大学の論文