Professional Documents
Culture Documents
Access to this document was granted through an Emerald subscription provided by emerald-srm:534301 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.
positive approach uses ratios to predict future inputs to produce multiple outputs where the
performance such as earnings and also to form of the production is not known or speci-
predict bankruptcy and assess the riskiness of fied as is the case with the parametric
the firm. approach. DMUs refer to the collection of
Both the normative and positive approach- firms, departments, divisions, or administra-
es have had some success. However, there tive units with the same goals and objectives
have been numerous methodological prob- which have common inputs and outputs.
lems pointed out by Barnes (1987), Smith Examples of DMUs include hospitals,
(1990), and Fernandez-Castro and Smith schools, courts, banks, and so on.
(1994). DEA is based on the pioneering work of
One of the prime reasons for using ratios is Farrell (1957). However, it was the work of
to control for the effect of size on the financial Charnes et al. (1978) which developed this
variables being studied so that one can com- technique. They generalized Farrells frame-
pare different firms or compare a firm to an work to include multiple incommensurate
industry average. However, this control for inputs and multiple incommensurable out-
size depends on the assumption of there being puts and reformulated as a mathematical
a proportionality between numerator and programming model to assess the compara-
denominator. This proportionality assump- tive efficiency of DMUs.
tion may not be true in many cases, thus The original DEA model developed by
Downloaded by University of Ghana At 16:32 27 June 2016 (PT)
leading to erroneous conclusions being drawn Charnes et al. (1978) is a fractional and non-
from ratio analysis. linear model. The objective function in that
Another problem with using ratio analysis model was to maximize the ratio of weighted
is that choice of a single ratio does not provide outputs to weighted inputs for a particular
enough information about the various dimen- DMU. This is done subject to the constraints
sions of performance of a firm. Considering (one for each DMU) that the ratio of weighted
many ratios is costly and there may be con- outputs to weighted inputs is equal to or less
flicting signals emerging from competing than 1. The decision variables are output
ratios. Aggregation of ratios is usually avoided weights (one for each type of output) and
because it requires weighting the ratios in input weights (one for each type of input).
some fashion and any such weighting is ulti- The DEA approach uses a linear program-
mately arbitrary. ming (LP) model to construct a hypothetical
A problem which also arises with ratio composite unit based on all units in the refer-
analysis is the choice of a benchmark against ence group. That is, the performance of each
which to compare a univariate or multivariate DMU is measured relative to the performance
score from ratio analysis. The choice of of all other DMUs. The unit being evaluated
benchmark depends on the costs to a user of can be judged relatively inefficient if the com-
an error in prediction and different users may posite unit requires less input to obtain the
require different benchmarks for different output achieved by the unit being evaluated,
purposes. However, this is not explicitly or judged relatively efficient if the composite
considered in most studies using ratio analy- unit requires as much input as the unit being
sis. evaluated does. A composite unit is a hypo-
The problems in ratio analysis pointed out thetical best-practice unit made up of a subset
above have prompted researchers to find new of units that should be emulated by a given
ways of addressing the issue of efficiency inefficient unit in order to improve the efficien-
analysis of firms. One such technique which cy of its operation. DEA results help in identi-
has been used in recent years is a mathemati- fying the relatively inefficient DMUs and
cal programming technique known as data providing insights into ways to improve pro-
envelopment analysis (DEA). ductivity of these relatively inefficient units
without reducing quality of service and while
maintaining or even increasing the volume of
Modeling with data envelopment
services provided by DMUs.
analysis
DEA analysis has been used in a number of
DEA is a nonparametric mathematical pro- settings to evaluate the performance of organi-
gramming model used to evaluate the relative zations including rate collection departments
efficiency of a group of entities or decision- (Thanassoulis et al., 1987), school districts
making units (DMUs) in their use of multiple (Bessent et al., 1983), hospitals (Sherman,
6
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
1984), court systems (Lewin et al., 1982), and evaluate the performance of banks in Jordan.
university accounting departments (Tomkins In contrast to the original, Charnes et al.s
and Green, 1988). (1978) model, no inputs are specified. The
DEA has also been used in a number of model can be written as:
bank studies. Sherman and Gold (1985) used Maximize Z0
DEA to evaluate bank branch operating
N =1
efficiency for a savings bank in the USA with
14 branch offices. They located inefficient
subject to n rin z0 ri 0 i = 1, , m
n =1
branches by explicitly considering the mix of
N
services provided and the resources used to n =1
provide these bank services. Vassiloglou and n =1
Giokas (1990) used DEA to assess the relative
Z 0 0; n 0 ( n = 1, ..., N )
efficiency of bank branches at the Commer-
cial Bank of Greece. Drake and Howcroft where Z0 indicates the ratio efficiency score of
(1994) used DEA to assess the relative effi- bank 0, n represents the weight placed on
ciency of the branches of a UK bank. each of the banks in forming the efficiency
frontier for bank 0, and ri refers to the
observed ratios for bank 0. N represents the
Corporate performance is number of banks which are appraised on m
Downloaded by University of Ghana At 16:32 27 June 2016 (PT)
Table I
ratios), subject to the condition of the model the various performance aspects of a bank.
that a high ratio be better than a low ratio, in Ratios 1 and 2 are generally considered good
order to get multidimensional indicators of indicators of the profitability of a firm. The
the performance of the bank. It was decided third ratio, earnings per share, is particularly
to leave out leverage ratios because it cannot useful for investors. Ratios 4 and 5 are good
be unambiguously concluded that a high activity measures for a bank. Ratio 6 is a good
leverage ratio is good for the firm. A high indicator of the liquidity of the bank. A high
leverage ratio may be good for the firm if it ranking in the above ratios is generally consid-
can trade successfully on the equity and take ered an indicator of good financial perfor-
advantage of the tax shield provided by debt. mance.
On the other hand a high leverage ratio can be
an indicator of high risk and may therefore be
bad for a firm. Fernandez-Castro and Smith Solutions to the DEA model
(1994) also indicated in their study that lever- The DEA assessment results for the years
age ratios may lead to confusing results. 1991-94 are given in Tables II-IV. The effi-
The ratios chosen for the study were as ciency scores, ranks, and composite DMUs
follows: are given in Table II. In 1991, out of the
(1) Return on investment = (Net income sample of 16 banks, three (18.75 percent)
after taxes 100) / Total assets were found to be relatively efficient (Z = 100
(2) Return on equity = (Net income after percent) and 13 (81.25 percent) were rated as
taxes 100) / Shareholders equity relatively inefficient (Z < 100 percent). In
(3) Earnings per share = Net income after 1992, 1993, and 1994, the number of relative-
taxes / Number of subscribed shares ly efficient banks went up to four (25 percent)
(4) Credits to total assets = (Credits granted whereas the relatively inefficient ones dropped
100) / Total assets to 12 (75 percent). The lowest efficiency score
(5) Credits to deposits = (Credits granted was 52.61 percent (bank 16) in 1991, 54.12
100) / Total deposits percent (bank 14) in 1992, 52.43 per cent
(6) Cash and portfolio investment to deposits (bank 14) in 1992, and 49.84 (bank 14) in
= (Cash and portfolio investment)/ 1994. It is also interesting to note that three
Deposits banks (bank 1, bank 6, and bank 13) contin-
ued to be relatively the most efficient (Z = 100
The ratios chosen above cover all the tradition- per cent) ones compared with the rest of
al categories of ratios used in the assessment of banks in the set from 1991 through 1994.
9
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
Table II
The efficiency rating does not only rank order the efficiency of bank 1 in 1991, the bank
the banks, but also suggests the degree of was more reliant on ratio 3 (credits to total
inefficiency of a bank compared with its effi- assets) than other ratios in the attainment of
ciency reference set. Next to each of the its performance. One can also note from the
relatively inefficient banks appears the corre- data for 1991 and 1992 given in Table III
sponding efficiency composite DMUs (refer- that the return on investment (ratio 1) con-
ence subset) and their associated shadow straint was binding for all banks in those
prices (weights). For example, bank 2 is 76.54 particular years. This also means that it
per cent as efficient as its reference set (bank contributes, with varying degrees, to the
1, bank 16, and bank 13). This means that it performance of all banks in the sample in
would have to increase its outputs by 23.46 1991 and 1992.
per cent to become as efficient as its reference The analysis of weightings for inefficient
subset. banks reveals additional insights. A zero
The study also reports further analysis of weight for a particular ratio implies a strictly
the reasons of efficiency, given by the weight- positive slack in the inequalities and suggests
ings and slacks of financial ratios for the inves- poor performance of the bank on the relevant
tigated banks. These results are given in Table ratio. This also suggests that the constraint of
III. For efficient firms, a zero slack in the that ratio is nonbinding. For instance, a
inequalities is associated with a zero and non- glance at Table III shows that ratio 5 (credits
zero weights and suggests good performance to deposits) assumed the position of the prin-
on the relevant ratio. A zero slack also sug- cipal determinant of the poor performance, as
gest that the constraint is significantly bind- it was nonbinding (positive slacks) for 12 of
ing. The determinants of efficiency are ratios the 16 banks for the years 1991-94. Along
with binding (zero slack) constraints. For with ratio 5, ratio 3 (credits to total assets)
instance, the results for 1991 given in Table assumed the position of the principal determi-
III indicate that the ratios that contributed to nant of the poor performance for 12 out of 16
the efficiency of bank 1 in 1991 are the ratios banks in the years 1992 through 1994. In
that assumed non-zero weightings (zero 1991, ratio 3 was nonbinding for only seven of
slacks). These ratios were return on invest- 16 banks, whereas ratio 6 (cash and portfolio
ment, earnings per share, and credits to total investments to deposits) was nonbinding for
assets. Their weights, respectively, were: ten banks in 1991.
0.30, 0.29, and 0.42. These results suggest In addition to identifying the magnitude of
that although ratio 1 (return on investment) inefficiency, efficient subset of banks, slacks,
and ratio 2 (earnings per share) contribute to and weights of the financial ratios, the DEA
11
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
Table III
1993
Bank 1 0.00 0.00 0.00 0.00 0.00 0.00
(0.00) (0.00) (0.19) (0.81) (0.00) (0.00)
Bank 2 55.90 0.00 97.60 0.00 0.49 0.00
(0.00) (0.94) (0.00) (0.01) (0.00) (0.05)
Bank 3 0.00 0.99 70.41 0.00 5.56 0.00
(0.39) (0.00) (0.00) (0.15) (0.00) (0.46)
Bank 4 0.00 0.00 62.32 0.10 0.73 4.58
(0.05) (0.95) (0.00) (0.00) (0.00) (0.00)
Bank 5 6.87 0.00 5.59 0.15 0.43 5.77
(0.00) (1.00) (0.00) (0.00) (0.00) (0.00)
Bank 6 0.00 0.00 0.00 0.00 0.00 0.00
(0.62) (0.00) (0.00) (0.01) (0.00) (0.36)
Bank 7 62.34 0.00 115.05 0.00 1.25 0.00
(0.00) (0.93) (0.00) (0.02) (0.00) (0.05)
Bank 8 0.00 0.00 77.48 0.17 1.04 6.87
(0.05) (0.95) (0.00) (0.00) (0.00) (0.00)
Bank 9 1.53 0.00 134.01 0.00 151.00 0.00
(0.00) (0.83) (0.00) (0.01) (0.00) (0.15)
Bank 10 0.00 0.00 0.00 0.00 0.00 0.00
(0.00) (0.99) (0.00) (0.00) (0.01) (0.01)
Bank 11 0.00 3.67 112.65 0.24 4.80 0.00
(0.48) (0.00) (0.00) (0.00) (0.00) (0.52)
Bank 12 0.00 7.61 128.91 0.00 3.55 0.00
(0.54) (0.00) (0.00) (0.02) (0.00) (0.44)
Bank 13 0.00 0.00 0.00 0.00 0.00 0.00
(0.17) (0.00) (0.00) (0.04) (0.00) (0.79)
Bank 14 0.00 2.51 138.69 0.31 0.99 1.08
(1.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Bank 15 41.90 0.00 123.21 0.20 1.53 0.00
(0.00) (0.81) (0.00) (0.00) (0.00) (0.19)
Bank 16 0.67 0.00 42.00 0.05 0.26 0.00
(0.00) (0.97) (0.00) (0.00) (0.00) (0.03)
(Continued)
13
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
Table III
methodology provides a much more detailed bank 6, and bank 13. The 74.39 percent
analysis of the actual and target (composite) composite efficiency score was computed as
financial ratios as given in Table IV. The follows:
financial ratios of composite DMUs are (0.00 62.85) + (0.29 176.56)
derived by multiplying the actual financial + (0.71 33.05) = 74.39 percent
ratios of each bank by the associated weights
of the reference set (taken from Table II). The
Summary and conclusions
banks with a nonzero shadow price are the
efficiency reference set. For instance, the DEA is becoming an increasingly popular
target score for the first financial ratio (return modeling tool for measuring the relative
on investment) for bank 2 in 1991 was 74.39 performance of DMUs. This paper described
percent whereas the actual ratio was 56.94 the application of DEA methodology to the
percent. This means that there was a 17.45 modeling and evaluation of the relative effi-
percent deficiency in the return on investment ciency of 14 banks. A modified LP model, in
ratio compared with composite of bank 1, which no inputs were specified, was utilized.
14
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
Table IV
Note:
Target financial ratios in parentheses
16
Modeling the operating efficiency of banks Logistics Information Management
Minwir Al-Shammari and Anwar Salimi Volume 11 Number 1 1998 517
The DEA is most useful when a comparison is school district, Management Science, Vol. 23,
sought against best-practice DMUs. pp. 1355-67.
The solution to the DEA models of investi- Charnes, A., Cooper, W.W., and Rhodes, E. (1978), Mea-
gated banks provide a measure of the relative suring the efficiency of decision-making units,
efficiency of the target bank, the efficiency European Journal of Operational Research, Vol. 2
No. 6, pp. 429-44.
reference set, the weights of ratios leading to
bank efficiency, and the target ratios for ineffi- Drake, L. and Howcroft, B. (1994), Relative efficiency in
the branch network of a UK bank: an empirical
cient banks. The majority of banks (81.25 study, OMEGA, Vol. 22 No. 1, pp. 83-90.
percent in 1991 and 75 percent in 1992
Farrell, M.J. (1957), The measurement of productive
through 1994) have showed poor perfor- efficiency, Journal of the Royal Statistical Society,
mance compared with the subset of best- Series A, Vol. 120 No. 3, pp. 253-90.
practice ones. The efficient banks are known Fernandez-Castro, A. and Smith, P. (1994), Towards a
as the peer group for the inefficient banks. general non-parametric model of corporate perfor-
Bank 1, bank 6, and bank 13 are the peers for mance, OMEGA, Vol. 22 No. 3, pp. 237-49.
the rest of banks throughout the period 1991- Lewin, A.Y., Morey, R.C. and Cook T.J. (1982), Evaluating
94. A fairly detailed analysis was applied to the administrative efficiency of courts, OMEGA,
the composite of best-practice set and its Vol. 10, pp. 401-11.
weights, the major determinants of banks Ricketts, D. and Stover, R. (1978), An examination of
relative performance, and to the target finan- commercial bank financial ratios, Journal of Bank
Research, Summer, pp. 121-4.
Downloaded by University of Ghana At 16:32 27 June 2016 (PT)
17
This article has been cited by:
1. Apostolos G. Christopoulos, Ioannis G. Dokas, Sofia Katsimardou, Konstantinos Vlachogiannatos. 2015. Investigation of the
relative efficiency for the Greek listed firms of the construction sector based on two DEA approaches for the period 2006
2012. Operational Research . [CrossRef]
2. Dimitrios Giokas, Nicolaos Eriotis, Ioannis Dokas. 2015. Efficiency and productivity of the food and beverage listed firms in
the pre-recession and recessionary periods in Greece. Applied Economics 47:19, 1927-1941. [CrossRef]
3. Wei-Kang Wang, Wen-Min Lu, Yu-Han Wang. 2013. The relationship between bank performance and intellectual capital
in East Asia. Quality & Quantity 47:2, 1041-1062. [CrossRef]
4. Wei-Kang Wang, Wen-Min Lu, Yi-Ling Lin. 2012. Does corporate governance play an important role in BHC performance?
Evidence from the U.S. Economic Modelling 29:3, 751-760. [CrossRef]
5. Wen-Min Lu, Mei-Hui Chen. 2011. A benchmark-learning roadmap for the Military Finance Center. Mathematical and
Computer Modelling 53:9-10, 1833-1843. [CrossRef]
6. Jos Humberto Ablanedo-Rosas, Hongman Gao, Xiaochuan Zheng, Bahram Alidaee, Haibo Wang. 2010. A study of the
relative efficiency of Chinese ports: a financial ratio-based data envelopment analysis approach. Expert Systems 27:5, 349-362.
[CrossRef]
7. Norma Md. SaadDepartment of Economics, Kulliyyah of Economics and Management Sciences, International Islamic
University Malaysia, Kuala Lumpur, Malaysia M. Shabri Abd. MajidDepartment of Economics, Kulliyyah of Economics
and Management Sciences, International Islamic University Malaysia, Kuala Lumpur, Malaysia Salina KassimDepartment of
Economics, Kulliyyah of Economics and Management Sciences, International Islamic University Malaysia, Kuala Lumpur,
Malaysia Zarinah HamidDepartment of Economics, Kulliyyah of Economics and Management Sciences, International Islamic
Downloaded by University of Ghana At 16:32 27 June 2016 (PT)
University Malaysia, Kuala Lumpur, Malaysia Rosylin Mohd. YusofDepartment of Economics, Kulliyyah of Economics and
Management Sciences, International Islamic University Malaysia, Kuala Lumpur, Malaysia. 2010. A comparative analysis of
the performance of conventional and Islamic unit trust companies in Malaysia. International Journal of Managerial Finance
6:1, 24-47. [Abstract] [Full Text] [PDF]
8. Yi-De Liu. 2009. Implementing and Evaluating Performance Measurement Initiative in Public Leisure Facilities: An Action
Research Project. Systemic Practice and Action Research 22:1, 15-30. [CrossRef]
9. Roma Mitra DebnathIGSM, Greater Noida, Uttar Pradesh, India Ravi ShankarSchool of Management, Asian Institute of
Technology, Pathumthani, Thailand. 2009. Assessing performance of management institutions. The TQM Journal 21:1,
20-33. [Abstract] [Full Text] [PDF]
10. Mohamed M. Mostafa. 2009. A probabilistic neural network approach for modelling and classifying efficiency of GCC banks.
International Journal of Business Performance Management 11:3, 236. [CrossRef]
11. Roma Mitra DebnathInstitute for Integrated Learning in Management, New Delhi, India Ravi ShankarSchool of
Management, Asian Institute of Technology, Pathumthani, Thailand. 2008. Benchmarking telecommunication service in
India. Benchmarking: An International Journal 15:5, 584-598. [Abstract] [Full Text] [PDF]
12. Yi-De Liu. 2008. Profitability measurement of UK theme parks: an aggregate approach. International Journal of Tourism
Research 10:3, 283-288. [CrossRef]
13. Constantin ZopounidisAthanasios G. NoulasDepartment of Accounting and Finance, University of Macedonia,Thessaloniki,
Greece Niki GlaveliAristotle University of Thessaloniki, Thessaloniki, Greece Ioannis KiriakopoulosNational Bank of Greece,
Athens, Greece. 2008. Investigating cost efficiency in the branch network of a Greek bank: an empirical study. Managerial
Finance 34:3, 160-171. [Abstract] [Full Text] [PDF]
14. Roma Mitra Debnath, Ravi Shankar. 2008. Measuring performance of Indian banks: an application Data Envelopment
Analysis. International Journal of Business Performance Management 10:1, 57. [CrossRef]
15. YI-DE LIU. 2007. Profitability Measurement of United Kingdom Theme Parks: An Aggregate Approach. Anatolia 18:2,
367-372. [CrossRef]
16. Mohamed MostafaGulf University for Science and Technology, Hawally, Kuwait. 2007. Modeling the efficiency of GCC
banks: a data envelopment analysis approach. International Journal of Productivity and Performance Management 56:7, 623-643.
[Abstract] [Full Text] [PDF]
17. Mohamed MostafaGulf University for Science and Technology, Hawally, Kuwait. 2007. Benchmarking top Arab banks'
efficiency through efficient frontier analysis. Industrial Management & Data Systems 107:6, 802-823. [Abstract] [Full Text]
[PDF]
18. Can Deniz Kksal, A. Akin Aksu. 2007. Efficiency evaluation of A-group travel agencies with data envelopment analysis
(DEA): A case study in the Antalya region, Turkey. Tourism Management 28:3, 830-834. [CrossRef]
19. Hussein A Hassan Al-Tamimi, Ahmad M Lootah. 2007. Evaluating the operational and profitability efficiency of a UAE-
based commercial bank. Journal of Financial Services Marketing 11:4, 333-348. [CrossRef]
20. Semih nt, Selin Soner. 2007. Analysis of energy use and efficiency in Turkish manufacturing sector SMEs. Energy
Conversion and Management 48:2, 384-394. [CrossRef]
21. Mohammed AlHawariCentral Queensland University, Queensland, Australia Tony WardCentral Queensland University,
Queensland, Australia. 2006. The effect of automated service quality on Australian banks' financial performance and the
mediating role of customer satisfaction. Marketing Intelligence & Planning 24:2, 127-147. [Abstract] [Full Text] [PDF]
22. Marianna Sigala. 2004. Using Data Envelopment Analysis for Measuring and Benchmarking Productivity in the Hotel Sector.
Journal of Travel & Tourism Marketing 16:2-3, 39-60. [CrossRef]
Downloaded by University of Ghana At 16:32 27 June 2016 (PT)