企業危機預警分析之研究-以Logit模型進行財務指標實證分析

Research of Enterprise Crisis Warning Analysis – Diagnosis Analysis of Financial Indicators Based on Logit Model

曾相海(Shiang-Hai Tseng)

指導教授 ： 吳瑞山

國立臺北大學/商學院/企業管理學系/碩士(2008年)

摘要

隨著金融自由化的趨勢，企業進行金融操作手法日趨繁雜，再加上經營層人員品德、操守違反正義原則及專業知識之不足，未能掌握金融環境之驟變，導致諸多企業頻頻發生財務狀況。當企業發生財務危機或倒閉時，對社會安全與金融秩序造成嚴重的衝擊。影響範圍遍及投資股東、企業員工、供應商、放款之金融機構，全體社會必須共同承受損失。因此，建構穩定且有效的企業財務危機預警模型，防範於未然實有必要。
本研究以財務指標為主體，建構企業財務預警模型，主要的研究發現如下：
1.財務危機的定義，採用「實質違約」方式認定，亦即實質違約的定義包括延期支付本息、紓困、跳票、重整、破產、倒閉，影響到債權人權益的事項，可避免選取樣本時，將正常公司誤認為危機企業，也可避免將全額交割股認定為危機企業，但又沒有違約的事實。
2.本研究將企業財務危機類型依其原因概分為：營運績效不佳、高槓桿與掏空三種類型，其財務特性：營運績效不佳型企業具有負債高、流動性低以及獲利差的財務特性；高槓桿型財務危機企業則具有流動性低、財務費用負擔沉重以及營運現金流量呈淨流出的的財務特性；營收虛造與資產掏空型企業財務特性為財務與償債能力指標正常，但部分經營能力指標出現異常現象，財務槓桿度則偏高。
3.在1%之顯著水準下，拒絕危機企業與正常企業的財務比率無顯著性差異的假設有：負債比率、借款依存度、流動比率、速動比率、現金比率、應收帳款週轉率、資產報酬率、存貨與應收款對淨值、營運現金流量率。
4.本研究選取2003至2004年間發生財務危機計有29家，總觀測樣本合計132家，作為原始樣本；危機企業的樣本抽樣最好能採取大樣本規模。並以負債比率、速動比率、長短期投資、營運現金流量與資產報酬率建構Logit財務預警模型，經實證結果顯示，整體模型的區別率達到91.70%的準確度，其中正常企業的區別
率高達93.20%，財務危機的區別率為86.20%，具有參考價值。

關鍵字

Logit模型、財務危機、違約率

並列摘要

Under the trend of financial liberalization, more and more complex financial manipulation by enterprises, injustice of management due to immorality, and insufficiency of professional knowledge have all resulted in a lot of financial difficulties faced by enterprises because of their inability of accommodating to the enormous changes of financial environment. As the financial crisis or bankruptcy occurred, social security and financial order would be severely impacted. The impact may involve enterprises’ shareholders, employees, suppliers, and financial institutions as creditors, whole society would also have to abide by the consequential costs at the same time. Accordingly, it is substantial to establish a stable and efficient enterprise financial crisis warning model.
This research report bases on financial indicators to build the enterprise financial crisis warning model. Major research findings are as follows:
1. The definition of financial crisis is granted by the occurrence of actual default, instead of subjective classification that popularly used by most banks.
2. This research categorizes enterprise financial crisis include three types by the causes of crisis: bad operation performance, high financial leverage, and embezzlement. Respective three characteristics are adopted by the research. enterprises which inflate profits and embezzle are possessed of normal financial indicators and general ability to pay, but the indicators of management turn out to be abnormal and the financial leverage is higher than average.
3. Under the significant level of 1%, the hypothesis is rejected and shows the fact that the crisis and normal enterprises have no evident difference.
4. The real diagnosis shows that the differentiation ratio of the whole model goes up to 91.70% in accuracy as the differentiation ratio of normal enterprise is as high as 93.20% while the differentiation ratio of financial crisis enterprise is 86.20% that is definitely referable.

並列關鍵字

Logit model, Financial crisis, probability of default(PD)

參考文獻

10.陳建宏、陳麗芬、戴錦周，樣本偏誤對財務危機預警模型之研究，東吳經濟商學學報，頁29-47，第57期，2007年6月。

1.Altman, E. I., “Financial ratios: Discriminat analysis and the prediction of corporate bankruptcy.＂ Journal of Finance 23(4):589-609,1968.

2.Altman, E. I., Haldeman, R.G. and Narayanan, P., “ZATE analysis－A new model to identify bankruptcy risk of corporations.＂ Journal of Banking and Finance, 29-54,1977.

3.Beaver, William H., “Financial ratio as predictors of failure.＂ Empirical research in accounting: selected study, Supplement to Journal of Accounting research, 71-111,1966.

4.Blum, M., “Failing company discriminat analysis.＂ Journal of Accounting Research 12(1):1-25,1974.