標題: 連續多期DEA應用在銀行業經營績效評估之研究
Study on the Performance Evaluation of Banking with Multi-periods by Data Envelopment Analysis
作者: 周皎如
Chiao-ju Chou
梁馨科
王淑芬
Shing-Ko Liang
Sue-Fung Wang
管理科學系所
關鍵字: 資料包絡分析法;績效評估;連續多期;Data Envelopment Analysis;Performance Evaluation;Multi-periods
公開日期: 2001
摘要: 在金融自由化的趨勢下,國內銀行家數及其分支機構數快速的增加,使得我國銀行業的競爭更為激烈。如何就銀行業之經營績效進行評估,實為政府主管機關、銀行之經營者、投資人等利害關係人所關心之重要課題。本研究以四十八家本國一般銀行及中小企業銀行為分析對象,以資料包絡分析法(Data Envelopment Analysis, DEA)作為衡量經營績效之工具。研究結果如下: (1)全體樣本銀行之總效率為0.9521、純粹技術效率為0.9715及規模效率為 0.98。90年度之總效率與技術效率比89年度為低,但90年度之規模效率比89年度為高。 (2)以多階段效率評選方式,採用經濟學中的等效率曲線概念及相對比較的基礎,多階段的效率評選方式來評估銀行經營績效,可以顯示出各銀行相對效率的等級情形,使效率評估更為合理。 (3)連續3年之效率分類比較,可以看出各銀行之效率變動呈現進步、不變、上下波動或衰退現象。在四十八家樣本銀行中,效率變動呈現進步現象的有15家、維持不變的有17家、呈現上下波動現象的有5家及呈現衰退現象的有11家。這些結果可作為經營者在經營管理上的警惕。
Under the trend of liberalization of financial service, there is a rapid increase in domestic banks and their branches. This caused the financial market more competitive. How to evaluate the performance of banking is the most crucial question to the government, bankers and investors. In this research we take forty-eight banks including general commercial banks and small-medium commercial banks as samples by Data Envelopment Analysis (DEA) to evaluate the performance of banking. Three major results are as following: (1)In 2001, the average of total efficiency of all samples is 0.9521. The average of technical efficiency of all samples is 0.9715. The average of scale efficiency of all samples is 0.98. The total efficiency and technical efficiency in 2001 are worse than those in 2000. But the scale efficiency in 2001 is better than that in 2000. (2)The multi-steps evaluation of efficiency method is based on the concept of efficient frontier. The multi-steps evaluation of efficiency method is not only more reasonable but also shows the levels of comparative efficiency between banks. (3)Sorting by 3-year-peroids efficiency, there are four clusters of banks. Those banks which remain in the same level are called consistent. Those banks with increasing level are called upward. Those banks with unstable levels are called fluctuating. Those banks with declining levels are called downward. This research shows that in forty-eight sample banks, fifteen banks are upward, seventeen banks are consistent, five banks are fluctuating and eleven banks are downward.
URI: http://140.113.39.130/cdrfb3/record/nctu/#NT900458012
http://hdl.handle.net/11536/69098
Appears in Collections:Thesis