Predicting financial distress using financial ratios: Malaysian evidence / Enylina Nordin

This study attempts to construct and test distress prediction models for Malaysian Companies. This study also observes and evaluates the classification and prediction error rates for the models developed by utilizing a sample of 84 distressed firms in 2001 and 2002 and a matched (by industry and fir...

全面介紹

Saved in:
書目詳細資料
主要作者: Nordin, Enylina
格式: Thesis
語言:English
出版: 2003
主題:
在線閱讀:https://ir.uitm.edu.my/id/eprint/59963/1/59963.pdf
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
實物特徵
總結:This study attempts to construct and test distress prediction models for Malaysian Companies. This study also observes and evaluates the classification and prediction error rates for the models developed by utilizing a sample of 84 distressed firms in 2001 and 2002 and a matched (by industry and firm size) sample of 84 healthy firms. The models are constructed using pooled data and yearly data of 5 years prior to financial distress by employing logit maximum likelihood estimator as a statistical technique. Pooled data model utilize measures of current liabilities to total assets and total borrowings to total assets. The model demonstrates excellent and moderate accuracy of financial distress firms and healthy firms respectively. Meanwhile, the prediction accuracy for the five yearly models ranges from moderate to excellent for both distressed and healthy firms. The prediction accuracy remains almost at the same level when the models are applied to an independent holdout sample. In addition, the developed models seem to fit. In conclusion, the pooled data model can predict financial distress of a firm up to 5 years. However, the 5 yearly models are not useful in predicting the financial distress of the firm since the predictors are not consistent in each model. It is also believed that the developed model can be useful to different groups of users such as policy makers, financial institutions, creditors, managers, bankers, investors and shareholders for making investment decisions.