從眾多籌資管道當中,企業選擇海外存託憑證作為其中長期資金的來源,掛牌公司具有那些特點影響其發行意願?海外掛牌後是否能增加公司績效?為瞭解海外存託憑證的發行因素以及對公司績效的影響,本文以臺灣上市上櫃電子資訊公司作為研究對象,研究期間自1991年至2004年,發行海外存託憑證有40家公司,無發行者有557家公司,為考量研究對象數目之限制,故採取含橫斷面與時間序列之混合型資料(Panel data)進行分析,扣除沒有資料及資料不全之樣本數後,共有211個年觀察值(year-firm observations)發行存託憑證,2,630個無發行,全體有效年觀察值為2,841個。 首先,我們發現海外交易所掛牌公司(SEC)、OTC與 144a公司及無海外掛牌公司三群在財務特性上有顯著不同。 本文並採用Logistic迴歸模型,探討那些財務特性會影響公司海外掛牌與否,並發現,公司績效低、公司規模大、低財務槓桿比率、低外銷比率、低固定資產比率、高資本支出比、股票流動性較高、成立時間較不久、上市公司,對存託憑證的發行有顯著的正向影響;並用Logistic迴歸模型預測存託憑證的發行機率。 再採用兩階段最小平方法聯立方程式模型(2SLS Model)探討存託憑證的發行與否及其發行機率對公司績效的影響,結果發現:存託憑證的發行與否與公司績效呈顯著負向影響,即高公司績效的公司傾向不發行存託憑證,低公司績效的公司傾向發行存託憑證,有發行存託憑證的公司其與公司績效呈顯著負向關係。
The enterprise chooses to issue Depositary Receipt as long-term capital source from multitudinous finance tools. What kind of characteristics has the enterprise that affects enterprise to issue DR? Can increase company performance after cross-listed? In order to understand the factors of issuing DR and the performance difference after cross-listed, this paper attempts to investigate correlative finance data. From 1991 to 2004 in Taiwanese electronic Industry, there are 40 companies issued DR and 557 companies not issued DR. Considering the research number, we use Panel data to analyze this paper including cross section and time section. After deducting the ineffective data, there are 211 year-firm observations issued DR, 2,630 year-firm observations not issued DR, entire 2,841 year-firm observations. First part, we find the three crowds are significantly different among SEC-listed companies, OTC and 144a companies, and non-cross-listed companies. The second part, we use logistic regression to analyze what kind of financial characteristics influences enterprise to issue DR. We find low performance, high firm size, low leverage, low overseas sales ratio, low fixed assets ratio, high capital expenditure ratio, high stock liquidity, younger, TSE listed companies are positive effective factors to affect enterprise issuing DR. Simultaneously, we forecast the issuing DR probability through logistic regression. Finally, we use issuing DR or not and issuing DR probability through 2SLS Model to analyze the influence of company performance. The empirical results reveal that issuing DR and company performance both have significantly negative influence. Meaning that firms with good performance tend not to issue DR, firms with bad performance tend to issue DR, and firms issued DR show a negative influence on company performance.