公司會計資訊、代理問題與信用風險模型之實證研究─家族與非家族企業比較

An Empirical Study on Accounting Information, Agency Problem and Credit Risk Models – Comparison of Family and Non-Family Firms

楊怡芬(Yi-Fen Yang)

指導教授 ： 余惠芳

共同指導教授 ： 金榮勇(Rong-Yong Jin)

健行科技大學/商學院/國際企業管理研究所/碩士(2012年)

摘要

2011年全球信評公司S&P 調降美國公債的信用評等與歐債危機引發全球股市大跌，國內經濟受到全球金融事件衝擊，如何建立信用風險模型(Credit Risk Models)，降低企業財務風險(Financial Risk)發生之機率，提供投資人、企業經營者、金融機構授信審核、學術研究與政府單位決策參考，成為相當重要的課題。目前台灣上市公司有76%為家族控制公司，家族企業多由家族成員負責公司經營並掌控管理階層。因此，本研究考慮降低代理問題(Agency Problems)，在經營人與所有人同一人的企業中，區分家族與非家族企業，嘗試找出影響公司績效之顯著變數，進一步比較分析其差異。期能在企業尚未發生信用風險前，預測可能發生的機率，達到事前預警與風險管理之效。
實證分析上，利用統計科學方法建構Logistic信用風險模型，依財務報表五大構面選取28項財務比率變數，及3項公司治理變數，計31項解釋變數篩選出影響公司績效之顯著變數。樣本期間為2000年到2011年，發生信用風險企業選出63家家族企業與37家非家族企業，採 Beaver(1966)的1：1配對原則，選取同一產業與固定資產規模相近的正常公司，共計200家樣本資料。實證結果發現，影響家族業公司績效的主要指標為財務結構、償債能力、經營能力與獲利能力指標；適度降低負債比率、提高利息保障倍數與財務槓桿度，可增強家族企業財務結構、償債與獲利能力。影響非家族業公司績效之主要指標為財務結構、償債能力、經營能力、獲利能力與現金流量指標；提高現金流當允當比率、股東權益報酬率(ROE)、總資產報酬率(ROA)，可提高非家族業經營與獲利能力，降低信用風險發生的可能性。因此，本文之實證價值與管理意涵，為公司會計資訊建立之信用風險模型，可有效運用於財務預測。

關鍵字

會計資訊 ； 代理問題 ； 信用風險模型,財務風險管理

並列摘要

The global credit rating company S&P lowered the credit ratings of U.S. Treasuries and European debt crisis triggered a global stock market crash and the domestic economy has been the impact of the global financial events in 2011. How to build credit risk models?Rreduce financial risk the probability of occurrence of the investors, business operators, financial institutions, credit review, academic research and government agencies decision-making reference, become a very important issue. Taiwan listed companies 76% of family-controlled company, family enterprises by family members responsible for the company and control of management. Therefore, this study consider the agency problem, the distinction between family and non-family business, try to identify the significant variablesaffectingcorporate performance, further comparative analysis of the difference. Of enterprises have yet to credit risk, to predict the probability that may occur to achieve the effect of prior warning and risk management.
Empirical analysis, using statistical methods to construct the logistic credit risk model, selected according to the financial statements of the 5 dimensions of 28 financial ratio variables, and 3 corporate governance variables,namely31explanatoryvariables tofilter out the significant variables affecting corporate performance. Sample period is from 2000 to 2011, a credit risk elect 63 family business with 37 non-family business, with Beaver (1966) 1:1 principle, select the same industry with similar scale of fixed assets normal company, for a total of 200 sample data. The empirical results show that the family industry, company performance indicators for the financial structure andsolvencyindicators; moderately reduce the debt ratio, interest coverage, and enhance the financial structure and solvency of the family business. Non-family industry, company performance indicators for the management capacity and profitability indicators; to improve the cash flow adequacy ratio, return on equity (ROE), return on total assets(ROA), can improve operations and profitability of non-family industry ability to reduce the likelihood of credit risk. Therefore, the empirical value and managerial implications, For the company's accounting information to build a credit risk model can be effectively applied to financial forecasting.

並列關鍵字

Accounting Information ； Agency Problems ； Credit Risk Model ； Financial Risk Management

參考文獻

(1) 李春安，徐傳瑛，「董事會獨立性與企業信用風險關係之研究」，管理與系統，第十七卷第三期，467~499頁，民國九十九年

(2) 余惠芳，王永昌，「財務預警與公司治理-台灣傳統產業之實證研究」，應用經濟叢刊，九十期，209~241頁，民國一百年十二月。

(3) 余惠芳，張士軍，黃于軒，「企業財務資訊、代理問題與信用風險模型之實證研究」，華人經濟研究，第2期，1~18頁，民國一百年。

(4) 余惠芳，「股權結構、股東權益與公司價值之實證研究」，全球管理與經濟，第五卷第一期，1~21頁，民國九十八年。

(5) 周行一，陳錦村，陳坤宏，「家族持股、聯屬持股與公司價值之研究」，中國財務學刊，第四卷第一期，115~139頁，民國八十五年。