本文主要使用2009年至2014年,台灣40家綜合證券商與中國大陸80家證券公司資料來進行效率與生產力分析。第一階段使用資料包絡分析法進行台灣與中國各別的技術效率分析,第二階段使用生產面Malmquist生產力變動指數與其分解出各項效率變動進行台灣與中國各別國家的生產力分析,最後使用共同邊界法,利用兩國的群組邊界以及共同邊界,進行效率與生產力的分析。 研究結果顯示台灣綜合證券商的效率值普遍低落,還有很大的進步空間。台灣非上市櫃綜合證券商效率值顯著大於上市櫃綜合證券商;非金控綜合證券商的規模效率顯著大於金控綜合證券商。中國證券公司效率值容易受到股市影響。中國非上市證券公司之規模效率顯著大於上市證券公司;位於京滬深地區的證券公司效率值顯著大於非京滬深地區。麥氏生產力指數的部分,台灣全期間有6.3%的衰退;中國全期間有2.7%的衰退。共同邊界麥氏生產力指數顯示,台灣與中國全期間的效率變動率雖然呈現收斂的情形,但因為群組的技術變動速度落後了共同邊界下的技術變動速度,因此全期間的catch-up值都沒有呈現收斂的情形。
This study aims at investigating the efficiency and productivity of Taiwan's securities firms and China’s securities firms from 2009 to 2014. First, we use Data Envelopment Analysis (DEA) to assess technical efficiencies of each firm in Taiwan and China. Second, we apply input-oriented Malmquist Productivity Index (MPI) to estimate the components of productivity change in Taiwan's securities industry and China's securities industry. Finally, we use Metafrontier method to analyze the efficiency and productivity between the group frontier and the metafrontier. The results show the efficiency of securities firms in Taiwan is generally low. The efficiency of non-listed and non-OTC securities firms in Taiwan is significantly greater than that of listed and OTC securities firms. And the scale efficiency of non-financial holding company (NFHC) securities firms is significantly greater than that of financial holding company (FHC). Non-listed securities firms’ scale efficiency in China is significantly greater than that of listed securities firm. The efficiency of securities firms located in Beijing, Shanghai and Shenzhen is significantly greater than firms that are not in Beijing, Shanghai and Shenzhen. Malmquist Productivity Index shows that there is a 6.3% recession in Taiwan’s securities firms from 2009 to 2014, but there are growing year by year; China’s securities has a 2.7% recession in the same period. Meta Malmquist Productivity Index shows that, despite the fact that the efficiency of securities firms in Taiwan and China shows convergence, because the technological change of the group frontier lags behind the technological change of the metafrontier, the catch-up does not show convergence.