國內文獻通常KMV模型來預測公司財務危機。KMV模型以資產小於流動負債加二分之一長期負債做為違約點(DPT, default point)。然而本文蒐集了台灣在2001年到2008年共67家違約公司,發現這些公司在發生違約當季的資產並未達到違約點,但卻有過半公司的流動資產小於流動負債。因此本文修正KMV模型,以流動資產小於流動負債時,公司就會發生財務危機。在67家違約公司與132家正常公司配對資料下,檢視發生違約時點前四季的違約距離的變化。根據群內分析法(the intra-cohort analysis)與Logistic廻歸分析,發現本文所發展的修正模型對財務危機預測能力明顯優於KMV。且隨著違約時間點的接近,修正模型的預測效果更佳。
The domestic literatures often used the KMV model to forecast the financial distress. The default occurs when the asset value below the short term debt plus half the long term debt considered by KMV model. However, many literatures documented this methods exhibit poor forecast ability. Moreover, we collect 67 distressed firms in Taiwan from 2001 to 2008, and find the asset values of all distressed firms are still higher than default points defined by KMV. On the contrary, about half of the distressed firms suffer current assets are lower than current liabilities. Using the data of 67 distressed firms to collocate with 132 normal firms, we investigate the variations in the distance to default from four seasons prior to default. According to the intra-cohort analysis and logistic regression, we found that forecast ability of the distance to default defined by current assets and current liabilities is significantly better than the distance to default defined by KMV. Furthermore, when the default time comes closer, the modified KMV forecast ability will become more accurate.