Font Size: a A A

The Study On Failure Pre-warning Model Of Listed Companies Based On The Logistic Regression Method

Posted on:2007-03-16Degree:MasterType:Thesis
Country:ChinaCandidate:Y WangFull Text:PDF
GTID:2179360182486424Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
The pre-warning study on failure of companies has gained extensive attention both in the country and abroad. As an important composition of the research on economy pre-warning system, it contains not only high science value, but also great social application value. With the development of market economy mechanism and capital market, kinds of problems of listed companies gradually emerged, and more and more listed companies failed. Therefore, it's of great importance to forecast the failure of them for operators keeping away financial crisis, for investors and creditors protecting their own behalf, for Securities Regulatory Commission supervising the risk of listed companies and stock market.This dissertation puts failed listed companies as research objects. On the basis of large numbers of literature, combining with characters of security market of our country, we defined failure as special treatment (ST).Then we collected cases and data of failing listed companies, and researched on the financial pre-warning of failure from both the theoretical and empirical aspects.On theoretical analysis, this dissertation anatomizes cases of ST stocks, and analyzes reasons of failure from the following five aspects: Industry factor, stock frame, intendance mechanism, business diversification and finance management. Then considering reasons and cases, we summarize the characters of failure in listed companies as deterioration of liquidity and leverage, profitability, activity, growing ability, structure and cash flow. Such results are very helpful choosing variables.On empirical research, this dissertation chooses ST stocks in manufacture as research sample, and selects financial indexes. Based on such choice, we combine Paired-Sample Test, Factor Analysis and Binary Logistic methods together, and construct financial pre-warning model of failure in listed companies.Ultimately, we testify the model's validity, and conclude 82.5 percent predicting accuracy rate.
Keywords/Search Tags:Failure, Pre-warning, Factor Analysis, Logistic Regression
PDF Full Text Request
Related items