1、<p>　　1800單詞，10500英文字符，中文2700字</p><p>　　出處：Barr R S, Killgo K A, Siems T F, et al. Evaluating the productive efficiency and performance of US commercial banks[J]. Managerial Finance, 2002, 28(8):3-25.&l

2、t;/p><p>　　Evaluating the Productive Efficiency and Performance of U.S. Commercial Banks</p><p>　　Richard S. Barr；Kory A. Killgo；Thomas F. Siems；Sheri Zimmel</p><p>　　Abstract: In this

3、 study, we use a constrained multiplier, input-oriented, data envelopment analysis (DEA) model to evaluate the productive efficiency and performance of U.S. commercial banks from 1984 to 1998. We find strong and consiste

4、nt relationships between efficiency and our inputs and outputs, as well as independent measures of bank performance. Further, our results suggest that the impact of varying economic conditions is mediated to some extent

5、by the relative efficiencies of the banks t</p><p>　　Keywords: Banks; Efficiency; Performance; Benchmarking; Data envelopment analysis</p><p>　　Introduction</p><p>　　Over the past t

6、wo decades, substantial research by financial economists in government and academia from all over the world has gone into evaluating the efficiencies of financial institutions. Berger and Humphrey (1997) survey 130 studi

7、es that apply frontier efficiency analysis to financial institutions in 21 countries. The vast majority of these studies were published in the 1990s, highlighting the importance and greater frequency of this research in

8、recent years.</p><p>　　Not coincidentally, this research and literature has expanded and evolved at a time of great change in world financial markets. A number of forces have fundamentally changed the world

9、in which financial services providers compete, including technology, regulations, and economic changes. For U.S. commercial banks, recent years have witnessed sweeping changes in the regulatory environment, huge growth i

10、n off-balance sheet risk management financial instruments, the introduction of e-commerce and on</p><p>　　In competitive industries, production units can be separated by some standard into those that perform

11、 relatively well and those that perform relatively poorly. Financial economists have done this “separation” by applying nonparametric and parametric frontier efficiency analyses. Berger and Humphrey explain that informa

12、tion obtained from such studies can be used for a variety of reasons. They can inform government policy by assessing the effects of various regulatory changes on efficiency. Resea</p><p>　　Success in competi

13、tive markets demands achieving the highest levels of performance through continuous improvement and learning. Comparative analyses and benchmarking information can alert institutions to new paradigms and new practices, l

14、eading to significant increases in firm efficiency and effectiveness. Frontier analysis methodologies are essentially sophisticated ways to benchmark institutions to determine the relative performance or efficiency among

15、 competing firms. Such analyses can identi</p><p>　　In this paper, we use a constrained-multiplier, input-oriented data envelopment analysis (DEA) model to quantifiably benchmark the productive efficiency of

16、 U.S. commercial banks. Using the parsimonious DEA model developed by Siems and Barr (1998), we measure relative productive efficiency of these institutions over the 15-year period from 1984 to 1998. We find strong and

17、consistent relationships between efficiency and our inputs and outputs, as well as independent measures of bank performance. </p><p>　　2. The efficiency of financial institutions</p><p>　　The fi

18、nancial institution efficiency literature is now both large and recent. Berger and Humphrey (1997) report that of the 130 studies that apply frontier analysis to determine financial institution efficiency, 116 were publi

19、shed from 1992 to 1997. Berger and Humphrey also report that there are now enough frontier analysis studies to draw some tentative comparisons of average efficiency levels both across measurement techniques and across co

20、untries, as well as outline the primary results of the</p><p>　　Previous studies have examined efficiency and associated effects on financial institution performance from several different perspectives. Thes

21、e include the effects of mergers and acquisitions (see Berger, Demsetz, and Strahan, 1999, and Resti, 1998), institution failure (see Barr, Seiford, and Siems, 1993, and Cebenoyan, Cooperman, and Register, 1993), and der

22、egulation issues (see Humphrey and Pulley, 1997, and DeYoung, 1998), among many others. Frontier efficiency models are employed by these</p><p>　　There are at least four frontier analysis methodologies used

23、 to compute financial institution efficiency, and there is no consensus among researchers on which method is best. The approaches differ mainly in how they handle random error and their assumptions regarding the shape of

24、 the efficient frontier. The three main parametric methodologies include the stochastic frontier approach (SFA), the thick frontier approach (TFA), and the distribution-free approach (DFA). In general, parametric approac

25、</p><p>　　DEA has proven to be a valuable tool for strategic, policy, and operational problems, particularly in the service and nonprofit sectors. Its usefulness to benchmarking is adapted here to provide an

26、 analytical, quantitative benchmarking tool for measuring relative productive efficiency. That is, DEA generally focuses on technological, or productive, efficiency rather than economic efficiency.</p><p>　　

27、3. Mathematical foundations for DEA</p><p>　　DEA generalizes the Farrell (1957) single-output/single-input technical efficiency measure to the multiple-output/multiple-input case. DEA optimizes on each indiv

28、idual observation with the objective of calculating a discrete piecewise linear frontier determined by the set of Pareto-efficient decision making units (DMUs). Using this frontier, DEA computes a maximal performance mea

29、sure for each DMU relative to all other DMUs. The only restriction is that each DMU lie on the efficient (extremal) f</p><p>　　Hypotheses</p><p>　　Our overall hypothesis in this study is that m

30、ore efficient institutions differ significantly from less efficient institutions (as determined by our DEA model) in measurable ways, and these results can be used for benchmarking. As one would expect, some measurable d

31、ifferences should manifest themselves particularly in the DEA model’s inputs and outputs. But we also expect to see differences in a variety of bank performance measures and between strong and weak institutions as determ

32、ined by bank e</p><p>　　Model results</p><p>　　Our DEA model was applied to publicly available year-end data reported by U.S. commercial banks from 1984 through 1998. For the purposes of our ana

33、lysis, de novo institutions (defined as those institutions that were less than three years old) are not included, as such institutions tend to have cost structures that differ significantly from more established institu

34、tions. Also excluded are banks that reported nonpositive values for any of the input or output measures, as such values are frequentl</p><p>　　To isolate the relative input and output characteristics of bank

35、s for further analysis, the banks that met our criteria are separated into quartiles by their derived efficiency score. These four groups serve as the basis for our comparison of more and less efficient banks vis-à

36、-vis the DEA models’ individual inputs. To control for banks of varying sizes, we employ a weighted ratio for each of the eight components using the appropriate asset measure: quarter-end assets for balance sheet-relate

37、d</p><p>　　6. Conclusion</p><p>　　In this study, we employ a constrained-multiplier, input-oriented DEA model to evaluate the relative productive efficiency of U.S. commercial banks across a 15-

38、year period. The DEA model offers numerous benefits, including the ability to target areas of relative efficiency between banks. Perhaps most importantly, it allows analysis of multiple aspects of a financial institution

39、’s performance, unlike more common benchmarking methodologies that focus on only one of many interrelated measures at a </p><p>　　We divide commercial banks into quartiles based on their DEA-derived efficien

40、cy score, and find that in each year each quartile has significantly higher efficiency scores than the quartile beneath it. A similar, rank-distinct relationship is discovered between efficiency quartiles on the weighte

41、d measures of noninterest income, other noninterest expense, and purchased funds (all three inversely related to efficiency), as well as earning assets and return on average assets (both positively relat</p><p

42、>　　The level of nonperforming loans to total loans is significant and negatively related to the efficiency scores of the most and least efficient quartiles from 1984 through 1993. The relationship of efficiency to sal

43、ary expense is similar from 1984 through 1994. It is likely that nonperforming loans and salary expense lose their predictive power vis-à-vis efficiency as a result of the same external forces: the improving economy

44、, improving conditions of financial institutions after the difficulties</p><p>　　References</p><p>　　Barr, R.S., Seiford, L.M., Siems, T.F., 1993. An Envelopment-Analysis Approach to Measuring t

45、he Managerial Efficiency of Banks. Annals of Operations Research 45, 1-19.</p><p>　　Bauer, P.W., Berger, A.N., Ferrier, G.D., Humphrey, D.B., 1998. Consistency Conditions for Regulatory Analysis of Financial

46、 Institutions: A Comparison of Frontier Efficiency Methods. Journal of Economics and Business 50(2), 85-114.</p><p>　　Bean, M.L., Duncan-Hodge, M., Ostermiller, W.R., Spaid, M., Stockton, R.S., 1998. Managin

47、g the Crisis: The FDIC and RTC Experience (Federal Deposit Insurance Corporation, Washington).</p><p>　　Berger, A.N., Demsetz, R.S., Strahan, P.E., 1999. The Consolidation of the Financial Services Industry:

48、 Causes, Consequences, and Implications for the Future. Journal of Banking and Finance 23, 135-194.</p><p>　　Berger, A.N., Humphrey, D.B., 1997. Efficiency of Financial Institutions: International Survey and

49、 Directions for Future Research. European Journal of Operational Research 98(2), 175-212.</p><p>　　Berger, A.N., Mester, L.J., 1997. Inside the Black Box: What Explains Differences in the Efficiencies of Fin

50、ancial Institutions? Journal of Banking and Finance 21, 895-947.</p><p>　　Bowlin, W.F., 1998. Measuring Performance: An Introduction to Data Envelopment Analysis (DEA). Journal of Cost Analysis, 3-27.</p&

51、gt;<p>　　Cebenoyan, A.S., Cooperman, E.S., Register, C.A., 1993. Firm Inefficiency and the Regulatory Closure of S&Ls: An Empirical Investigation. Review of Economics and Statistics 75, 540-545.</p><

52、;p>　　Charnes, A., Cooper, W.W., 1962. Programming with Linear Fractional Functionals. Naval Research Logistics Quarterly 9(3/4), 181-185.</p><p>　　Charnes, A., Cooper, W.W., Golany, B., Seiford, L., Stutz

53、, J., 1985. Foundations of Data Envelopment Analysis for Pareto-Koopmans Efficient Empirical Production Functions. Journal of Econometrics 20, 91-107.</p><p>　　Charnes, A., Cooper, W.W., Lewin, A.Y., Seiford

54、, L.M., 1994. Data Envelopment Analysis: Theory, Methodology and Applications (Kluwer Academic Publishers, Norwell, MA).</p><p>　　Charnes, A., Cooper, W.W., Rhodes, E., 1978. Measuring the Efficiency of Deci

55、sion Making Units. European Journal of Operational Research 2(6), 429-444.</p><p>　　Cole, R.A., Gunther, J.W., 1995. A CAMEL Rating’s Shelf Life. Financial Industry Studies, 13-20.</p><p>　　DeYo

56、ung, R., 1998. Management Quality and X-Inefficiency in National Banks. Journal of Financial Services Research 13(1), 5-22.</p><p>　　Farrell, M.J., 1957. The Measurement of Productive Efficiency. Journal of

57、the Royal Statistical Society, Series A, General, Part 3, 253-281.</p><p>　　Humphrey, D.B., Pulley, L.B., 1997. Banks’ Responses to Deregulation: Profits, Technology, and Efficiency. Journal of Money, Credit

58、, and Banking, 73-93.</p><p>　　Resti, A., 1998. Regulation Can Foster Mergers, Can Mergers Foster Efficiency? The Italian Case. Journal of Economics and Business 50(2), 157-169.</p><p>　　Siems,

59、T.F., Barr, R.S., 1998. Benchmarking the Productive Efficiency of U.S. Banks. Financial Industry Studies, Federal Reserve Bank of Dallas, 11-24.</p><p>　　美國商業(yè)銀行的生產效率和績效評估</p><p><b>　　摘要<

60、;/b></p><p>　　我們采用基于約束乘數以及投入的數據包絡分析模型來評估從1984到1998年美國商業(yè)銀行的生產效率。我們發(fā)現效率和獨立措施之間有強烈的一致性。此外，我們發(fā)現，在一定程度上，變化的經濟條件產生的影響受到在這些條件下運轉的銀行的相對效率的制約。最后，我們發(fā)現了效率與由銀行審單員信用評級決定的穩(wěn)固性之間的緊密聯系。對于那些用基準問題測試與其他機構以及監(jiān)管機構的聯系，以此作為銀行審查過程

61、中的監(jiān)控工具的補充的銀行，這種模型將會起到一定作用。</p><p>　　關鍵詞：銀行;效率;性能;基準;數據包絡分析</p><p><b>　　1.引言</b></p><p>　　在過去的二十年中，有大量的研究涉及了金融機構的效率評估。伯杰和漢弗萊（1997）研究了近期130個運用前沿效率分析法分析分布于21個國家的金融機構的案例。這個研

62、究以及理論成果在世界金融市場經歷巨大變化的時期進一步發(fā)展，這并非偶然。美國商業(yè)銀行見證了監(jiān)管環(huán)境翻天覆地的變化，資產負債表金融工具風險管理的巨大增長，電子商務和網上銀行的引入，以及意義深遠的金融行業(yè)整合。所有這些都使美國的銀行產業(yè)更具競爭力。</p><p>　　在競爭激烈的行業(yè)中，根據一些標準可以將生產經營單位分成經營較好以及較差的兩種。金融方面的經濟學家運用前沿效率分析方法進行了這一種“區(qū)分”。伯杰漢弗萊表示

63、從這些研究中所獲得的信息可以使用于多個方面。他們可以通過評估效率管理變化所產生的影響來為政策導向提供信息。通過描述一個行業(yè)的效率可以決定研究議題。此外，通過確定高效率的最“好”和低效率的最“壞”的生產實踐能夠提高管理水平。</p><p>　　在本文中，我們使用基于約束乘數以及投入的數據包絡分析模型，以量化的基準，測量美國商業(yè)銀行的生產效率。我們使用DEA方法，是因為它著眼于生產或者技術上的效率。DEA規(guī)定了一系

64、列關于最好的實踐模式的觀察值并形成了可以評估所有機構的分段線型前沿。</p><p>　　如果我們的DEA模型顯示了銀行效率和銀行獨立措施之間一致的聯系——包括銀行審單員所作的信用評級，那么這個模型就能夠作為監(jiān)控工具的補充而對銀行以及監(jiān)管機構起作用，從而輔助銀行審查。對比分析以及基準信息可以提醒機構注意新的方法，由此促進公司效率及效力的顯著增長。機構可以被定位并得到效率值和排名，這些將有益于決策者，業(yè)內分析師，競

65、爭公司的管理者。</p><p>　　通過使用由Siems和Barr（1998）開發(fā)的DEA模型，我們測量了美國從1984年至1998年經營超過15年的商業(yè)銀行的相對生產效率。我們發(fā)現效率和投入產出以及銀行獨立措施之間強有力的聯系。此外，我們的研究結果表明在一定程度上，變化的經濟條件產生的影響受到在這些條件下運轉的銀行的相對效率的制約。最后，我們發(fā)現了效率與由銀行審單員信用評級決定的穩(wěn)固性之間的緊密聯系。<

66、/p><p><b>　　2．金融機構的效率</b></p><p>　　之前對金融機構的研究金融機構從不同角度對其效率以及績效進行了探索。這些探索包括兼并收購的影響，機構倒閉，撤銷管制等等。前沿分析模型之所以能夠被研究者采用主要是因為他們能夠提供對相對績效客觀量化的測量，從而減少外因的影響。因此研究者可以將注意力集中于成本，投入，產出，收入，利潤等量化的測量，由此將效率

67、和最好的實踐機構聯系起來。</p><p>　　目前至少有四個前沿分析方法用于計算金融機構的效率，而研究者們對于最好的方法并沒有達成一致。這些方法的主要區(qū)別在于如何處理隨機誤差以及關于效率邊界的猜測。三個主要參數包括隨機前沿方法（SFA），厚前沿方法（TFA），和自由分布方法（DFA）。</p><p>　　總的來說，參數方法詳細說明了在成本，利潤或者在輸入輸出以及環(huán)境因素之間的生產關系的

68、功能形式，并允許存在隨機誤差。這種主要的非參數方法是數據包絡分析。由Charnes, Cooper, and Rhodes (1978)發(fā)明的數據包絡分析通過使用多重輸入以及多重輸出來計算由個人決斷的單位的技術（生產）效率。</p><p>　　我們將DEA方法作為我們發(fā)展效率前沿的選擇，這是因為DEA的僅僅著眼于生產效率，不需要明確的規(guī)范基本生產關系的形式。DEA方法已被證明是一個有價值的工具，用于戰(zhàn)略，政策和

69、操作問題，特別是在服務和非營利性部門。它對測量基準的作用也適用于這里，提供了一個分析，定量的工具用于銀行之間相對生產效率的測量。</p><p>　　3．DEA方法的數學基礎</p><p>　　DEA將法雷爾（1957）的技術效率測量方法進一步發(fā)展概括于多重輸入/輸出的案例中。 DEA以計算離散分段線性前沿為目標，優(yōu)化對每個個體的觀察報告。使用這個前沿，DEA計算出每個DMU相對于其他D

70、MU的最大性能測量。唯一的缺陷是每個DMU要依靠效率前沿或者限制在這個前沿之內，那些依賴這個前沿的DMU是最好的實踐機構。</p><p><b>　　4．假說</b></p><p>　　我們的總體假設是，效率較高的機構與效率較低的機構之間有顯著差別，這些差別是可以被測量出來的。我們希望看到各個銀行業(yè)績評估的顯著差別，各個強弱機構之間的顯著差異，以及該模型的輸入和輸

71、出的不同。具體來說，更高效的機構應具有較高的盈利水平，較少的問題貸款和較低（強）銀行信用評級。</p><p><b>　　5．結果</b></p><p>　　根據美國1984年到1998年商業(yè)銀行的報告，我們的模型可公開運用于年終數據統計。那些經營還少于3年的機構不包括在內，因為這種機構傾向于擁有與那些已存在的機構截然不同的成本結構。根據報告，那些投入或者產出為負

72、值的銀行業(yè)不包括在內，因為那些數值通常暗示報告的失誤以及操作的紊亂。</p><p>　　根據得出的效率分數，我們將符合我們標準的銀行分成四個部分。這四個組將作為我們這四個組作為我們比較更高效率或更低效率銀行的基準。通過我們控制銀行規(guī)模雇用的八個組成部分，每年的加權比例，使用適當的資產措施：季度末的資產，資產負債表相關項目及相關收入和支出項目的平均資產。此外，分析銀行業(yè)績：平均資產的回報率，貸款總額的不良貸款比率

73、，貸款總額占總資產的比例一般措施。</p><p><b>　　6. 結論</b></p><p>　　在本文中，我們采用基于約束乘數以及投入的數據包絡分析模型來評估美國商業(yè)銀行15年內的生產效率。DEA模型允許關于金融機構績效多個方面的分析，不像更多共同的基準方法僅側重于同一時期內相互關聯的測量中的一種。DEA創(chuàng)造了一種更為廣闊的分析法，這種分析法不需以犧牲洞察的深

74、度，且更相關，更適用于真實世界中復雜的金融機構的運行。</p><p>　　根據DEA導出的效率分數將銀行分成四種類別時，我們發(fā)現每年每一種類別都會比下面的一種擁有更高的效率分數。同樣的，在效率和非利息收入，其他的非利息支出和購買基金之間的等級關系。盡管效率和利息收入及支出之間的聯系也許不像市場競爭的結果那樣普遍，但是我們可以看到的是效率和利息收入之間正相關以及和支出之間負相關，這樣一個顯著的趨勢。此外，在最高效