傳統產業財務槓桿、代理問題與信用風險模型之實證研究

本研究主要以台灣上市櫃之傳統產業為研究對象，利用統計方法Logistic 迴歸分析，針對企業資本結構之負債比率＞50%，建構企業信用風險模型(Credit Risk Models)。樣本資料選取期間為2000年至2012年，共14項產業類別93組樣本資料，採Beaver(1966)之1比1配對原則，選取固定資產規模相近之同類型產業，合計共186組樣本資料進行分析。實證分析上，考慮資本結構比例，首次嘗試針對傳產業高財務槓桿公司，利用28項財務比率變數及4項公司治理變數，進行K-S test、M-U test、Pearson test與建立Logistic 迴歸模型，找出影響公司發生信用風險之顯著變數，以達事前預警與風險管理之效。實證結果發現，在有效的資本市場中，影響公司績效的主要變數除了現金流量指標之外，公司治理指標亦是重要因素。研究建議，提高營業活動現金流量以增加企業流動性，能支應到期債務利息壓力，增加每股現金流量比率，可改善企業的財務結構，增加投資人信心，有利於企業籌資；增強公司治理能力，考慮公司專業經理人(CEO)的經營權與股東的所有權(Ownership)分開，提高專業經營能力與公司績效。因此，本文實證價值與管理涵意為，傳統業公司應重視公司管理階層，採分權式的雙重領導結構，提高專業經營，增強公司治理能力，有效降低信用風險發生的可能性，進而提高公司營運績效。

關鍵字

財務槓桿；代理問題；信用風險；公司治理

並列摘要

This study took listed the traditional industries of Taiwan as subjects. By logistic regression statistics, Against the Debts ratio for the corporate capital structure> 50%, this study constructed a Credit Risk Models. The sample data were collected during year 2000 to 2012, including 93 sample data of 14 industry categories. By Beaver (1966) 1 to 1 matching principle, selecting the similar scale of fixed assets and the same type of industry, for a total of 186 sample data analysis. According to the analysis, discussed the highly leveraged companies for traditional industries under the proportion of the capital structure, this study conducted K-S tests, M-U tests, Pearson tests and logistic regression models by 28 financial ratio variables and 4 corporate governance variables to extract the significant variables of corporate performance as the criteria for prediction and warning of corporate finance. According to the fingings, the main variables of corporate performance are corporate governance except cash fiow under efficient capital markets. This study thus suggests that raising cash flow of operating activities in order to increase liguidity can slove the debt interest and increase cash flow rate per share to improve financial structure and increase investor confidence to finance; In order to enhance the corporate governance power, considering companies with seperated management and ownership could adversely affect professional capability and corporate performance. Lastly, a seperated management and ownership arrangement would reduce the probability of credit risk and increase the forecasting capacity.