非集團業自有資金、代理問題與信用風險之實證研究

本文利用Logistic 迴歸統計科學方法，建構非集團企業信用風險模型(Credit Risk Models)，考慮自有資金及代理問題(Agency Problems)，試圖找出預測非集團企業發生財務危機之因素，期望在企業發生信用風險之前，提前預測信用風險發生的機率，以達事前預警與風險管理之效。本文蒐集2000年至2012年非集團上市上櫃公司為研究樣本，採Beaver(1966)之 1：1配對原則，選取和危機公司同產業、同屬性、與固定資產規模相近之財務正常公司作為配對樣本，合計302家。利用28項財務比率變數及4項公司治理變數，進行K-S tests、M-U tests、Pearson相關檢定與Logistic迴歸分析，消除共線性並篩選出影響公司績效之顯著變數，以作為預測企業財務預警之用。實證結果發現，當自有資金大於50%，影響非集團業績效之主要指標有償債能力、獲利能力、現金流量、公司治理指標。在財務變數方面，提高利息保障倍數、稅前淨利率與總資產報酬率，降低營業費用率、內部保留比率與借款依存度，可增強企業償債與獲利能力。在公司治理方面，避免董事長兼任總經理，強化公司治理能力，可降低發生信用風險的機率。實證價值與管理意涵為，非集團業聘任專業經理人專業經營，降低信用風險發生的可能性，值得投資決策參考。

關鍵字

自有資金；代理問題；信用風險管理； Logistic迴歸

並列摘要

This paper uses logistic regression statistical method to construct the credit risk models for non-group companies. By taking into consideration of own funds and agency problems, this paper attempts to identify the factors that can predict the financial distress of non-group companies, expecting to forecast the probability of credit risk occurrence in advance for warning and risk management. With the data of non-group listed companies and OTC-traded companies during the period from 2000 to 2012 as the research samples, this study adopts the 1:1 matching principle proposed by Beaver (1966) to select financially normal companies of the same industry, attribute and similar fixed asset scale. The companies in financial distress are treated as the matching samples. There are a total of 302 companies. By using 28 financial ratio variables and 4 corporate governance variables, K-S tests, M-U tests, Pearson correlation test and logistic regression analysis are performed to remove col-linearity and select significant variables that can affect company performance, in order to predict financial crisis of the companies. The empirical results suggest that when the own funds ratio is higher than 50%, the major factors affecting the performance of non-group companies include solvency, profitability, cash flow, and corporate governance indicators. In financial variables,enhancing the times interest earned, return on sales, and return on assets, lower operating expense ratio, retention ratio anddebt/equity ratiocan enhance corporate solvency and profitability. Regarding corporate governance, avoiding management and ownership are combined,and strengthening corporate governance can reduce the probability of occurrence of credit risk. The empirical value and managerial implication ofthis study are that hiring professional managers can reduce the likelihood of credit risk innon-group companies. The findings can serve as references for investment decision making.