A Study on Financial Crisis Warning Model of Enterprises in Taiwan—An Application of Polytomous Logistic Regression / 以多分類邏輯斯迴歸模型預警台灣企業財務危機之研究

碩士 / 東吳大學 / 財務工程與精算數學系 / 97 / This research divides the financial crisis degree into three classes of the normal operation, slight crisis and heavy degree of crises. Construct the explanatory variables as this research of every financial ratio data with five financial perspectives of profit ability, cash flow, debt paying ability, financial structure and operation ability. Use the polytomous logistic regression model and the cascaded logistic model to set up the financial crisis warning models according to different industry's classification respectively.

This result of study shows that the differences do exist between various industries by looking over the explanatory variables significant of twenty financial ratios data from food industry, textile industry, steel industry, building industry and electronics industry respectively. The macro correctly classified percentages by integrated all perspectives of each industry with better predictable ability than the warning model set up by five perspectives respectively.In terms of verifying the comparison samples, electronics industry gets a correct classification rate up to 81.8%, which means this model ,to enterprises of electronics industry, really with warning ability of predict whether the crisis will take place in company's financial situation.To food industry, textile industry, steel industry and building industry, there is a integrated correct classification rate by integrated all perspectives about 71.9% to 85.7% ,even without sufficient comparison samples as proving, which means the model has certain warning ability of crisis to every industry.

Identiferoai:union.ndltd.org:TW/097SCU05336005
Date January 2009
CreatorsChun-lin Yen, 顏君玲
ContributorsYoeng-Kuan Chang, 張永寬
Source SetsNational Digital Library of Theses and Dissertations in Taiwan
Languagezh-TW
Detected LanguageEnglish
Type學位論文 ; thesis
Format114

Page generated in 0.0016 seconds