Distressed Prediction Models in the Stock Market of Mainland China

• Main Authors: Steven Lo, 羅生賢
• Other Authors: none
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/90843733306972495246

碩士 === 東吳大學 === 企業管理學系 === 97 === This paper uses logit regression to construct a distressed prediction model in the stock market of Mainland China. There are 68 companies which first denounced as Special Treatment from 2006 to 2007 selected as distressed firms; and 794 companies which never denounced as Special Treatment until 2007 are selected as regular firms. The paper uses 18 financial ratios and 10 corporate governance variables，and seperately builds up the distressed prediction models of three years prior to distress and the two years prior to distress. The conclusions are as followings: 1. The variables in the distressed prediction model of three years prior to distress are debit ratio, total assets turnover ratio, cash flow per share. 2. The variables in the distressed prediction model of two years prior to distress are debit ratio, total assets turnover ratio, return on equity, non-operating expense and income/ net sales, cash flow per share, contingent liability ratio. 3. The distressed prediction model of two years prior has higher ability of prediction than the three years prior to distress. According to the conclusions, we can find that debit ratio, total assets turnover ratio, cash flow per share are both included in the distressed prediction models of three years prior to distress and the two years prior to distress, it shows the consistency and importance of these three variables. Furthermore, return on equity, non-operating expense and income/ net sales, contingent liability ratio are only included in the distressed prediction model of two years prior to distress, it indicates that these three variables would be the key indicators in predicting the Special Treatment company in Mainland China.