經營環境及隨機干擾調整後之兩岸銀行效率分析

Abstract

本研究使用Fried et al. (2002)三階段資料包絡分析法分析2000年到2004年台灣29家及中國大陸16家全國型銀行的經營效率。由於資料包絡法所估算的技術效率值會受到人為管理技術影響、銀行所處環境影響以及隨機干擾影響，因此本研究使用三階段資料包絡分析法的目的即要區分這三種因素對於效率的影響。採用存款、員工人數及固定資產為投入變數，並以投資及放款作為產出變數，兩岸銀行資料均經過以2000年為基期的GDP平減轉成實質變數，並均以億美元為單位進行研究。第一階段使用原始投入及產出，以資料包絡分析法分析兩岸銀行的經營效率；第二階段使用隨機邊界迴歸模型估計分支機構數、成立年限、放款/存款比率、貪腐印象分數與經濟自由度等外生變數對各項投入差額變數的顯著性分析，並分離人為管理、環境因素及隨機干擾的影響，接著再進行投入項調整；第三階段使用調整後的投入項資料，再次以資料包絡分析法估算銀行效率。 研究結果發現：分支機構數、成立年限及貪腐印象分數皆與存款差額變數、員工人數差額變數及固定資產差額變數呈現顯著正向影響；放款/存款比率與存款差額變數呈顯著負向影響；經濟自由度則與員工人數差額變數呈顯著負向影響。而在經由環境因素及隨機干擾項調整過後的兩岸銀行整體平均效率皆有明顯的提升，亦即兩岸銀行現階段受經營環境影響很深，應即早做好準備，以減少不必要的浪費。

This thesis applies a three-stage data envelopment analysis (DEA) approach proposed by Fried et al. (2002), in order to analyze the operational efficiency of 45 nationwide banks across the Taiwan Strait (29 in Taiwan and 16 in mainland China) from 2000 to 2004. Deposit, numbers of employees, and fixed assets are the three input variables, while investment and loans the two outputs. All nominal variables are transformed into real variables by GDP deflators using 2000 as the base year. At the first stage, the DEA with original input and output is used to compute the operational efficiency and input slacks. At the second stage, the stochastic frontier regression model to estimate if the exogenous variables have significant effects on the slack variables. The artificial management, the environmental factors, and statistical influences are the decomposed. All the inputs are adjusted based on slacks. At the third stage, the adjusted inputs are used to re-run DEA again. Our major empirical findings are as follows: The number of branches, period since establishment, and Corruption Perception Index have significantly positive effects on slacks of deposit, number of employees, and fixed asset. The loan/deposit ratio has a significantly negative effect on the slack of deposit. The Index of Economic Freedom has a significantly negative effect on slacks of number of employees. After environmental effects and statistical noises are adjusted, the operational efficiency of the banks in both Taiwan and mainland China become significantly higher, implying that environmental variables do matter for bank efficiency across the Taiwan Strait.