2004 NTU International Conference of Finance on Financial Market Competition and Integration 台灣大學財金系 2004/12/20(一) 講演教授：淡江大學財務金融系所系主任兼所長 聶建中 博士 http://mail.tku.edu.tw/niehcc/ Nieh, Chien-Chung

債務比率的門檻效果對中國上市公司的價值影響 Panel Threshold Effect of Debt Ratio on Firm Value in Chinese Listed Companies 債務比率的門檻效果對中國上市公司的價值影響

Panel Threshold Effect of Debt Ratio on Firm Value in Chinese Listed Companies employs threshold regression model to investigate the panel threshold effect of financial leverage on operating performance (or firm value) of Chinese listed firms optimal capital structure ROA => operating performance debt-to-asset ratio => threshold variable

權益資本 (Equity Capital) 資 本 結 構 負債資本 (Debt Capital) 相對關係 此消彼長 財務槓桿的運用 (capital structure) 負債資本 (Debt Capital) 相對關係 此消彼長

負債資本 優點 缺點 適當的使用帶來稅盾效果 提高公司價值 過度使用會連帶產生破產成本、 代理成本、 負債成本 公司價值下降

理論上應存在 財務槓桿運用程度 最適負債比率 過低 過高 無法有效提昇 公司價值極大化 降低公司價值

Data and Variables 821 A-shares firms on the Shanghai stock exchange (SHSE) and Shenzhen stock exchange (SZSE) that went public before 1998 (343 firms are from SHSE and 478 from SZSE) sample period running from 1998 to 2002 ‘balanced panel data’ of debt-to-asset ratio and ROA, each with the total observations of 4105 (821*5) The sample excludes the listed financial firms.

Data and Variables industry pattern of capital structure of industry firms the ST and PT firms were not included in the sample 465 A-shares industry firms on the SHSE and SZSE sample period 1998 to 2002 again ‘balanced panel data’ with 2325 observations ST and PT are abbreviations for Special Treatment and Particular Treatment, respectively. When the firm is not profitable for two consecutive years, it will be classified as ST stock. If the ST stock is still unprofitable, it will be classified as PT stock. (Now this provision is abolished.)

Data and Variables Independent Variable (Threshold variable) Long-term Debt and Short-term Debt in China Book Value and Market Value in China Dependent Variable ** ROA (the ratio of net profits to total assets) ***Tobin q cannot well assess Chinese listed firm value because of the difficulty of evaluating the market value of non-circulating shares, the market value of debt and replacement cost

Tables Results Table 1 Descriptive Statistics of Variables for 821 A-shares firms Table 2 Descriptive Statistics of Variables for 465 A-shares industry firms

Methodology (Panel Threshold) (Basic Model) Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法

Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法 負債門檻值 公司價值 公司價值 小於 大於 正相關 負相關 受到槓桿關聯成本抵銷 稅盾效益有限 公司價值極大化

Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法 本文主要在討探負債比率對於公司價值的影響 解釋變數 被解釋變數

研 究 方 法 (五) Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法 單一門檻模型

研 究 方 法 (六) Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法 門檻效果檢定 研 究 方 法 (六) Hansen (1999) -- 縱橫門檻(Panel Threshold)迴歸方法 門檻效果檢定 在虛無假設成立下時，此時係數 ，迴歸方程式退化成一般單條迴歸式，故門檻效果並不存在；但若對立假設成立時，此時係數 ，即此係數與在兩區間會有不同之解釋現象，代表負債比率對於衡量公司價值及經營績效值會存在一上下非對稱之門檻效果。

Model Where

821家上市公司實證結果的估計模型： 465家上市產業公司實証結果： 此時我們把負債占資產的比率d當作自變數… 且今天以ROA來代替公司價值並把公司價值v當作應變數

Results triple threshold effect the coefficients for four regimes are all positive and significant  the operating performance is positively related to the debt level (However) firms with ‘very low debt’ has relatively higher coefficient  the positive effect decreases when debt-to-asset ratio increases (Moreover) the industry pattern of capital structure  the industry pattern does not match the Chinese listed firms