考慮產業修正因子的財務預警模型:Logistic模型的運用

近年來，隨著金融開放，人們參與金融市場的程度加深，財務危機公司的預警便成了當前最重要的議題，本論文旨在建立一個只需要輸入公開資訊的財務預警模型。傳統的財務危機預警模型，往往只採用財務比率變數，但過去研究發現亞洲公司往往有集團持股或家族控制之行為，公司治理充滿著人治色彩，且對於外在環境改變的衝擊並未有考量，故本研究加入公司治理與總體因素，試圖加強模型的預警能力。本研究選取的樣本為2007年1月到2009年12月被TEJ定義為財務危機之公司，並以同產業且資產規模相近的正常公司，以1:2的方式進行配對，並選擇過去文獻中最重要的18項理論變數來進行逐步迴歸，得出最有預警能力的變數共計14個來做為Logistic模型的使用變數。除此之外，並考慮不同公司分屬不同產業便具有不同的產業財務特性，透過以產業均數平減與以產業均數標準化的方式，來淨化不同公司的產業差異情形。研究發現以平均數平減的預警模型在整體的預測正確率表現最佳，且對危機公司的預警能力最為穩定。此外，並透過不同分割點的選擇來加強對危機公司的偵查能力，研究發現當分割點為0.3的一般模型對危機公司最有預測能力，但分割點的下修亦造成對正確公司判斷正確的能力下降。

關鍵字

財務危機； Logistic ；產業修正；分割點；逐步迴歸

並列摘要

Recently, with banking open, people get involved in financial market frequently and that brings the most important issue about the precaution of financial crisis companies at present. The thesis is built on financial precaution model via inputting public information only. Traditional financial crisis precaution model uses financial ratio variables. Studies in the past discover that consortiums hold stocks or control behavior by family in Asian companies. Company government is covered with human government and doesn’t consider the impact of external environment changes. This study adds company government and whole factors to try to enforce precaution ability of model. The samples of this study are financial crisis companies which are defined as financial crisis company by TEJ in January, 2007 to December, 2009. They match with the same industries and the similar scale companies in the method of one to two. They also choose the most important eighteen theoretical variables from the literatures in the past to do stepwise regression and get the most precautionary fourteen variables to be regarded as the variables in Logistic model. In addition, they also consider that different companies which belong to different industries have different industrial financial properties. They use this method of industrial variables minus mean and standardized mean to purify the industrial differences of the different companies. Studies discover that the precaution model of variables minus mean predicts most correctly in whole prediction and keeps precaution ability most stably in crisis companies. And moreover, they choose different cutting points to enforce discovered ability to crisis companies. Studies discover that average models whose cutting points are 0.3 have the best prediction ability to crisis companies, on the other hand, cutting points which are decreased also make discriminate ability to correct companies decrease.

並列關鍵字

financial crisis ； Logistic ； industry revision ； cutting point ； stepwise regression

參考文獻

4. 張大成、林郁翎、邱雅雯（2008），考慮盈餘管理之企業財務危機預警模型，亞太經濟管理評論，第11期，頁65-86。

5. 張大成 (2006)，產業差異與企業財務危機模型，臺灣金融財務季刊，第7期，頁1-28。

8. 張大成 (2002)，違約機率與信用評分模型，台灣金融財務期刊，第4期，頁19-37。

1. Altman, E.I. (1968), Financial Ratios, Discriminant Analysis and The Predicition of Corporate Bankruptcy, Journal of Finance, Vol. 23, pp.589-609.

2. Beaver, W.H. (1966), Financial Ratio as Predictors of Failure, Journal of Accounting Research, Vol. 4, pp.71-111.

國際替代計量

考慮產業修正因子的財務預警模型:Logistic模型的運用

未授權

主題瀏覽