公司發生危機可能是因為財務結構不良或經營效率不彰兩種因素，但過去的學者及研究人員在建立危機預警模式時，多半只考慮到公司的財務比率，而忽略了經營效率對於公司發生危機的影響。本論文之主要研究目的在於探討影響公司危機發生機率的變數研究，危機預測的重要性在於不僅可以讓企業自行診斷發生危機的可能性，更可以給予金融機構從事放款工作時審核與訂價的重要依據。本論文樣本選取1999年至2005年之間發生危機之25家上市公司，配對規模相近產業相同之正常公司25家。而本研究除了傳統的財務比率變數之外，又加入了經營效率變數進行探討。利用資料包絡法來求算效率變數，以及運用Logit分析方法建立較為準確之危機預警模型。實證結果顯示：1.在技術效率分析發現，離危機發生前第二年，危機公司可能就有經營惡化現象。2.正常公司的純技術效率是優於危機公司的，表示管理者決策較為適當。3.加入經營效率變數可以有效增加預測準確度。4.生產力跨期分析中發現距離危機發生時間點愈近，危機公司生產力愈不理想。 The reasons for company crisis may result from the financial factor or non- efficiency or both. In the exiting paper, when researchers establish crisis early warning system, they only consider companies financial ratio variables while neglect efficiency influence. The purpose of this study is to discuss the factors which can affect the occurrences of crisis of organizations. The importance of anticipating crisis can not only allow the corporation do self-diagnose, but also grant the institutes abilities to evaluate the interest rate when they are conducting the loaning process. Using a matched pairs design this study obtained a sample of 25 distress firms from 1999 to 2005 with the same industry and scales. This study applies DEA and Logit model the efficiency variables with in addition to the traditional variables. The empirical findings are summarized as follows：1. From the technical efficiency analysis, it showed that the crisis companies may be deteriorated two years before the crisis happened.2. That the pure technical efficiency of the normal companies is better than one of the crisis companies demonstrated the administrator’s policy making of formers is more proper. 3. Adding the efficiency variables into this model could increase the accuracy for the anticipation.4. From the analysis of productivity change, we found that the shorter the distance from the happening time of crisis, the less the productivity of the crisis companies.
