財務比率建立財務危機預警模型之實證研究－合併財務報表與母公司財務報表之比較

With the rise of enterprise financial crisis, it has been an important issue to build an effective warning mechanism. The purpose of the study is to build effective financial distress predicting models by combining the research of the parent corporations’ consolidated financial statements and stand alone financial statements, and to probe into the relative crises predicting power and the information content underlying in the above mentioned financial statements.
The financial distress predicting model of the study is established by applying logistic regression and evaluating the performances of models with the correctly predicting percentage and Accuracy Ratio (AR). The findings of the study are summarized as follows:
I. It is found that the predicting powers of the two different predicting models built by the consolidated financial statements and the stand alone financial statements of the parent corporation are similar. .
II. The empirical findings indicate that if in-sample is applied, the predicting model built from a single industry is superior to the model built from cross industries, and the difference is statistically significant. But out-of-sample can not lead to the same conclusion.
III. Because the correctly predicting percentage is involved in the selection of cut-off-point and is not objective, Accuracy Ratio is the main tool used to evaluate the performance of models.