本文以民國93年至96年共33家公開發行之綜合證券商為研究對 象,首先利用資料包絡分析法,來探討各樣本公司之相對經營效率, 再以差額變數分析提出券商未來改善經營的建議,第二階段則利用 Tobit迴歸分析,探討本研究所選取之四項外生變數與第一階段所估 計之效率值間的關係。 實證結果顯示:(一)受評單位在這四年間的平均總技術效率值為 69.9%,顯示仍有30.1%的資源浪費,因此台灣綜合券商的整體經營效 率尚有相當大的改善空間;(二)Tobit迴歸分析結果顯示,在總技術 效率方面,證券公司設立年數及公司的資本額與經營效率呈現顯著正 相關,而證券公司是否屬於金控架構及證券商的分公司家數則無顯著 相關。
The objective of this study is to utilize the Data Envelopment Analysis (DEA) approach to measure the relative efficiency of 33 listed security companies in Taiwan during 2004 to 2007. The slack variable analysis is then applied to each inefficient company to provide information on future inefficiency improvement. The second stage of the study is to employ the Tobit regression to DEA efficiency measures. Factors influencing security companies’ overall technical inefficiency will be identified. Major findings of the study are as follows. First, average overall technical efficiency of all DMUs during the four-year periods is about 0.699, indicating there are more than 30% resources are wasted in Taiwan’s security companies. Firms can improve their inefficiency by changing their inputs and/or outputs mix. Second, results of Tobit regression show that the life and the equity capitals of the security companies have significantly positive impact on firm’s efficiency. Whether firms belong to financial holding company or not and the number of branches of a firm are not significantly related to firm’s efficiency. Keywords: