Return to search

企業財務危機預警模型之建構-以類神經網路為工具

由於財務報表資訊易遭管理當局操縱,因此財務預警模型若僅考慮財務比率變數,即有其限制。本研究因此結合財務比率變數與公司治理變數,以期建構更良好的財務預警模型。此外,本研究使用倒傳遞網路為工具,以避免前述限制,並預期結果顯示綜合採用財務比率及公司治理二類變數,在預測期間短時,所建立的財務預警模型,其錯誤率的確較低。本研究同時發現,樣本公司中的危機公司大多屬於「急速失敗公司」。 / Early warning models used to predict financial distresses of corporations confront with limitation, when the model specification consider only financial ratios based on financial statements, because of the possibility of manipulated financial statements. This study intends to construct a early warning model with not only financial ratio variables, but also corporate governance variables. The corporate governance variables may affect the corporation with financial distresses dramatically. This study constructs a new early warning model, considering the two kinds of variables, both financial ratio and corporate governance, and improves the predictability of sample firms of the one-quarter period. The study shows that Back Propagation Neural Network model can learn from the data of failed corporations and a matched group of survivor firms and hence predict the financial distresses. The study also finds the sample failed corporations are more likely to be “acute failure” ones.

Keyword: BPN, Corporate Governance, Financial Distresses.

Identiferoai:union.ndltd.org:CHENGCHI/G0090353008
Creators楊謹瑜
Publisher國立政治大學
Source SetsNational Chengchi University Libraries
Language中文
Detected LanguageEnglish
Typetext
RightsCopyright © nccu library on behalf of the copyright holders

Page generated in 0.0016 seconds