摘 要:本文在分別借助因子分析和非參數檢驗方法對公司財務指標和治理因素進行統計處理的基礎上,構造并實證檢驗了用于預測我國上市公司財務困境的兩大模型,即僅是包含財務信息與融合財務信息和公司治理因素的Logistic回歸預測模型。實證結果表明,作為通過了非參數檢驗的公司治理因素之關鍵特征變量——第一大股東派出董事比例,不但與公司發生財務困境的概率之間呈現顯著的負相關關系,而且能夠顯著提高包含該變量的預測模型的回判和預測準確率。因此,公司治理信息應當是利于預測我國上市公司是否可能發生財務困境的重要因素。
關鍵詞:公司治理;財務困境;因子分析;Logistic回歸;第一大股東派出董事比例
中圖分類號:F272.1 文獻標識碼:A 文章編號:10035192(2007)02006305
Corporate Governance and Financial Distress Prediction
HUANG Shandong, YANG Shue
(School of Management, Xi’an Jiaotong University, Xi’an710049, China)
Abstract:Statistically disposing corporate financial index and governance ingredient in virtue of factor analysis and nonparameter test respectively, the paper construct and empirically test two logistic regression models for predicting Chinese listed companies financial distress, i.e. model which just include financial information and model which syncretize financial information and corporate governance ingredient. Empirical results suggest that key characteristic variable of corporate governance ingredient,the percentage of director occupied by the first large shareholder, which pass nonparameter test,not only are prominent negatively related to firm financial distress probability, but also can remarkably improve backwards differentiate exact rate and predict exact rate of the model which include the variable. Thereby corporate governance information should be redound to predict if Chinese listed companies occur financial distress.
Key words:corporate governance; financial distress; factor analysis; Logistic regression; the percentage of director occupied by the first shareholder
1 問題的提出
企業發生財務困境將給企業管理者、股東、員工及其他人員帶來巨大的經濟損失,對整個國家也帶來巨大的社會和經濟成本。一個財務困境預測模型能起到預先警告利益相關者的作用,從而使管理者和投資者及時采取應對措施。因此,對財務困境進行準確的預測是一個重要的財務問題。
自Fitzpartrick、Beaver、Altman等經典文獻問世以來,大量的研究人員對財務困境預測進行了研究。從多元判別分析等線性預測模型,到以神經網絡模型為代表的各種非參數預測模型,相關的研究成果層出不窮。然而,要開發完全滿足需要的模型對這一領域的研究者仍然是一個挑戰。……