過去有關國內外財務危機預警之實證文獻，甚少從銀行徵信之角度分析。本文主要目的在於建構預測財務危機的完善模式，有效地偵測企業發生危機的徵兆，降低公司發生破產、倒閉等行為，可有效降低金融機構呆帳的問題。本研究引用銀行企業徵信報告中所採用的財務比率，再加入公司治理變數等非財務因素，檢視企業發生危機前三年，其整合性變數是否能提早發現危機徵兆及預測危機的能力。本研究使用Logit模型進行樣本測驗，研究結果發現：危機公司越接近危機發生時其支付利息的費用越高，而依賴借款的程度越明顯，且營運現金淨流量較差，債權保障性低，財務結構穩健性欠佳。而正常公司在營業收入及毛利上的控制相對優於危機公司。此外，再以財務比率驗證企業危機前一年至前三年之預測正確率，其分類正確率分別為70.80%、70.80%、68.80%。再者，加入非財務比率進行整合樣本驗證後，其分類正確率於危機前一年至前三年各為96.90%、90.60%、83.30%。因此，若建構企業財務危機預警模式時加入非財務比率後可顯著的提升模式整體預測能力，有助於及早預測出企業發生財務危機之徵兆，且越接近危機發生時機，其預測正確率更明顯提升。 Previous studies on the indication of corporate finance crisis seldom analyzed from the viewpoint of bank credit information. The main purpose of this study is to establish a prediction model of corporate financial crisis in order to reduce the bad debt of the bank. The Logit model was used to analyze the financial ratio variables adopted from corporate financial statement as well as other non-financial factors dealing with corporate governance. The result shows that a corporate is approaching finacial crisis when it was facing higher interest expenses and lower net operating cash inflow, higher dependence on loan, lower indemnification of creditor’s right and inferior steadiness of financial system. The predictive accuracy of the corporate financial crisis based on the financial ratio data of previous 3 years before crisis is 70.80%, 70.80% and 68.80%, respectively. Furthermore, when the non-financial factors of previous 3 years before crisis were integrated into the analyses, the accuracy of this prediction arose to 96.90%, 90.60% and 83.30%, respectively. Therefore, besides the financial ratio data, the corporate governance of a corporate is also important in detemining a bank loan in order to protect the finacial assets of bank.
