Vous êtes sur la page 1sur 6

Finance Management

Measuring Efficiency in Indian Commercial Banks


with DEA and Tobit Regression
Vikas Adhegaonar
vikasadh@rediffmail.com

ABSTRACT
Last 25 years Indian commercial banks has been observing deregulation, technological innovation
and increased opportunities to finance Indian economy and emerging competition from private sector
accompanied with foreign ownership banks. Government approach to liberalization is to spur
competition which further influences to efficiency in Indian commercial banking sector. In this paper
attempt has been made to examine the efficiency in Indian commercial banks particularly focus on
nationalized banks. 19 nationalized banks are selected for this study. The period of the study is 2010,
2011 and 2012; the required data is collected from IBA website. The data analysis is processed with
the help of R- software. To examine the efficiency in nationalized banks data envelopment analysis
technique in used with constant return to scale and variable return to scale. For data envelopment
analysis the inputs selected are interest expenses, operating expenses and for output variable selected
is interest income, non interest income. For further analysis tobitregression model is considered with
fiveindependent variables deposit share of bank, operating profit, non performing asset, credit deposit
ratio , investment deposit ratio and dependent variable is technical efficiency score.

Keywords- Nationalized banks, Data Envelopment Analysis, CSR,VSR, Technical Efficiency ,


Tobit Regression

1) Introduction
The banking sector plays a very crucial role in the economic growth in India. The efficient
banking sector is thus the fundamental requirement for smooth functioning of any economy (Arora
et.al.).Soundness is key for Indian financial system and soundness is synonymous for stability,
profitability, efficiency, productivity and a shock free environment ( A.K. Mishra et.al. ,2013 ).If
banks intermediate efficiently it positively affects to economic growth and banking failures results
into systematic crises, so bank performances are at vital interest for depositors, regulators, customers
and investors(M Duygun- Fethi,2009).Measuring operational efficiency of financial institutions is
pivotal for academic researcher and policy makers, as aim of both is to assess the impact of market
structure on financial system and improve efficiency of financial system (ShaikSaleemet.al.,
2014).The main objective of liberalization Indian banking is stability, stand against external shocks
and remain internally sound and sensible. As efficiency in Indian commercial banks increases it leads
to reduction in spreads, this will stimulate industrial loan demand (lead to higher economic growth)
and greater mobilization of savings ( Majid Karimjade,2012).Competition in banks and banking
system forces commercial banks to perform efficiently (I.A. Shah, 2012).
Indian commercial banking comprises of scheduled commercial bank and non scheduled
commercial banks. Scheduled commercial bank includes SBI group, nationalized banks, private sector
banks, foreign sector banks and regional rural banks. This paper is focused on nationalize bank group.
Indian government nationalizes commercial bank in 1969 and 1980 to pursue objectives indentified
with social orientation. These banks are under government control and faced various challenges from
regulated interest rates, reserve requirement, directed lending. By 1991, liberalization with
deregulated interest rates allowed these banks to pursue its own course of action, devise strategy to
compete with other banks and perform efficiently. But still government expectation in social objective
is not eliminated; these banks have to support government in schemes like Jan DhanYojana, LPG
subsidy. These schemes drags nationalize banks into additional admin work, staff, administrative
expenditure. Also nationalize banks faces political interference into distribution of credit which

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 239
Finance Management

resulted into NPA. Nationalized banks have huge branch network which resulted into substantial
amount of staff remuneration cost. The objective of banking is to work as intermediary between
savers and borrowers and this job need to be done with efficiency. But as above mentioned issues like
political interference, government interference, employee unions forces these banks to work in less
efficiency. So this paper address efficiency issue in Indian nationalized banks and after liberalization
in banking policies such analysis is required to measure how these banks are performing.
The objective of this paper is to measure the efficiency in Indian commercial banks during 2012-
2014 and this paper focuses on nationalized banks particularly. The first part of this paper provides
introduction to the topic, second part discusses on literature review, third part introduces to the
research methodology adopted for study and then data analysis with findings are supported. Final part
of the paper concludes this study.

2) Literature Review
Avkiran (1999), DEA gives an intuition on the areas that need to be improved but it lacks
information on what are the ways in which the improvement need to be done.Vinod R.R. (2013)
adopted intermediation approach to study efficiency in Indian private sector banks and found that only
3 private sector banks out 12 are consistent during 2008-2012. i.e. only 25% of sample banks are
consistently technically efficient (ING Vysya Bank Ltd, Karur Vysya Bank Ltd, Nainital Bank Ltd).
Abhiman Das et.al. (2006) have adopted intermediation approach, value-added approach and
operating approach to examine how efficiency score vary changes in input and output. The findings of
the study indicate that banks that have less non performing loans are technically more efficient.(Don
U. Galgedara,2010) examined the efficiency of Indian commercial banks during 1995-2002,
observed that no significant growth in productivity in private sector banks and public sector banks
demonstrated modest positive change. (Amit Kumar Dwivedi et.al., 2011) studied the efficiency in
all Indian commercial banks with the model which includes loans and non interest income as output
and no of branches , operating expenses and deposit as input. His study found that national banks,
new private banks and foreign banks showed higher efficiency as compare to private banks, SBI
group and nationalized banks during study period 2006-2010.(Mukesh Kumar et.al.,2012 )
examined Indian commercial banking sector from 1996-2010 and observed economic reforms and
global financial crisis. The study concluded that sector banks are faintly doing better than private
sector banks in terms of technical efficiency and scale efficiency. The observed increased return to
scale in public sector banks indicates that sustaintial gains could be obtained with altering scale of
operations either with internal growth or consolidation in sector.(Arora et.al.,2014) found lest overall
technical efficiency score for foreign banks where private sector banks better performed to public
sector banks. (A.K. Mishra et.al., 2013) found that private sector banks are found efficient compare
to public sector banks. (ShaikSaleem et.al., 2014) foundforeign banksare more efficient than private
and public banks. So to improve operational efficiency in public sector banks , PSB should develop
management team and highly qualified personnel with sufficient skills in assessing the opportunities
and threats.(MilindSathye) found that mean efficiency score compares well with world mean
efficiency socre and private sector banks as group perform better than public and foreign sector
banks.(A. Armugam, 2014) found that nationalized banks and SBI group are less efficient as
compare to the private and foreign sector banks during 2002-2011, but in recent times all banks are
found increased their efficiency socres.(Sunil kumar,2008) found that proposition that larger the
bank more it is efficient does hold in Indian sector banks. So the banks to become more efficient need
to choose appropriate input and output mix with optimum scale of size.

3) Research Methodology
i) Objectives –
a) To measure efficiency in Indian commercial banks with focus on nationalized banks.
b) To predict efficiency score with Tobit regression model.
ii) Period of study is 3 years 2012, 2013, 2014.
iii) Data is collected from IBA website.

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 240
Finance Management

iv) Sample for study is 19 nationalized banks.


v) Data analysis is processed with the help of statistical package -R software.
vi) Techniques used for data analysis
1) Data envelopment analysis
In banking researchers consider production approach and intermediation approach of data
envelopment analysis. The production view of DEA consider number of accounts of deposits or loans
as inputs and outputs respectively. This approach assumes that banks produce loans and deliver
financial services and assumed that banks are intermediaries between depositors and borrowers. The
intermediation approach assumes banks as financial intermediary and so considers volume of
deposits, loans and other variables as inputs and outputs. In this paper intermediation approach of
DEA is used with input orientation model which assumes that efficient consumption of resources
while holding output constant. CCR model of DEA work out as constant return to scale (CRS)
assumption, provides technical efficiency which is overall efficiency of banks. The BCC model
assumes that variable return to scale (VRS), which permits to the calculation of technical efficiency
and scale efficiency. TE score obtained underCRS model measures overall efficiency due to input
output configuration and size of operations i.e. it is gross efficiency score and it comprises of scale
efficiency and technical efficiency aggregated into one. The efficiency measure under VRS model
represents pure technical efficiency due to only managerial performance or inefficiency due to
managerial underperformance. The relationship between TE under CRS model and VRS model is
scale efficiency.

Input variable Output variable


Interest expenses Interest income
Non interest expenses Non interest income
Table No.1- DEA model used in the study
DEA is operation research based non parametric technique which indicates efficiency when TE
score =1 and inefficiency otherwise. This paper examined input output oriented model of DEA which
assumes: By how much quantities can proportionately be reduced without altering quantities
produced?
2) Tobit regression model
In second stage, censored Tobit regression model is considered for further analysis and this study
examines following Tobit model.
TE= β0+ β 1 dep+ β 2 oprofit+ β 3 idr+β 4 cdr + β 5 npa+ε
Dependent variable = TE score
Independent variables
Dep= deposit share of i bank in total bank deposit; oprofit= operating profit to total asset;
idr=investment deposit ratio; cd=credit deposit ratio;npa= Non performing asset; ε= error term

4) Data analysis and interpretation of results


Following part of the paper discusses on data analysis and interpretation of the findings. First
discussion is provided on technical efficiency scores obtained with VRS model, CRS model and then
results of Tobit regression model are provided.
Appendix 1 provides detailed TE score, PTE score and scale efficiency scores. In 2012,
Allahabad bank, Dena bank, Syndicate Bank, UCO Bank is found efficient and rest banks are
inefficient with corporation bank and Indian bank found lowest score. Also average efficiency is
found 74% with standard deviation 25%. In 2013, Bank of Maharashtra, Punjaband Sind Bank,
Syndicate Bank, UCO Bank found efficient and rest banks are inefficient with Andhra bank , Indian
bank are lowest efficient. Also average efficiency in 2013 is found 75% with standard deviation 26%.
In 2014, Andhra Bank, Dena Bank, Punjab & Sind Bank are found efficient banks and rest banks are
found efficient with United Bank of India as lowest efficient banks. Also average efficiency in 2014 is
found 70% with standard deviation 21%.

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 241
Finance Management

Call:
tobit(formula = te ~ dep + oprofit + idr + cdr + npa, data = a)
Observations:
Total Left-censored Uncensored Right-censored
51 0 51 0
Coefficients:
Estimate Std. Error z valuePr(>|z|)
(Intercept) 1.790065 0.321938 5.560 2.69e-08 ***
dep -1.792775 0.476880 -3.759 0.000170 ***
oprofit 14.799312 3.968888 3.729 0.000192 ***
idr -0.009366 0.004532 -2.066 0.038782 *
cdr -0.009816 0.003488 -2.814 0.004893 **
npa -0.002271 0.014705 -0.154 0.877248
Log(scale) -2.515367 0.099015 -25.404 < 2e-16 ***
---
Signif.codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Scale: 0.08083
Gaussian distribution
Number of Newton-Raphson Iterations: 4
Log-likelihood: 55.92 on 7 Df
Wald-statistic: 24.51 on 5 Df, p-value: 0.00017296

Table no.2-Tobit model results


Table no.2 shows the result of tobit regression model. All independent variables are found
statistical significant except NPA where dependent variable is TE scores. The banks which have
higher market share in deposit are adversely affecting to TE scores and deposit is found statistically
significant. This indicates that banks which have easy access to deposit mobilization are less efficient.
The banks which have higher operating profit have higher TE score and operating profit is found
statistically significant. The banks which have higher investment deposit ratio focuses on investment
to increase income than credit distribution so it resulted into less efficiency and investment deposit
ratio is statistically significant. Also banks which have high credit deposit ratio are found performing
less efficiently and this relationship is found statistically significant. NPA is found statistically
insignificant but the relationship indicates negative relationship between NPA and TE score. Chart
no.1 shows that actual te score and te score predicted by tobit model with residual. The model has
predicted te scores accurately with presence of small variation in residual.
1.2
Actual TE Score Predicted TE Scoe Residual
1

0.8

0.6

0.4

0.2

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51
-0.2

-0.4
Chart no.1- Actual TE score, predicted TE score, Residual

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 242
Finance Management

5) Findings and conclusion


In this paper attempt has been made to understand efficiency in Indian commercial banks with
focus particularly on nationalized banks. The study found that overall average technical efficiency in
2012, 2013, 2014 is 74%, 75%, 70% respectively with CRS model. Average technical efficiency in
2012, 2013, 2014 is found 87%,84%,88% respectively with VRS model. Average scale efficiency in
2012, 2013, 2014 is found 82%, 85%, 78% respectively with VRS model. Tobit regression model has
found all the independent variables significant except NPA. Only operating profit variable is
positively related with TE score. The study found banks with high deposit ratio need to focus on their
overall efficiency by identifying and removing factors adversely affecting their performance. The
banks with high investment deposit ratio are less efficient, these banks need to reduce their investment
and focus on their core activities or alter their investment portfolio with high return securities. Banks
with high CD ratio are also inefficient, these banks need to focus on credit distribution by avoiding
NPA issues or increase investment in other securities to increase efficiency.

References
1) A. Armugam, G. Selavalakshmi(2014), “Impact Of Banking Sector Reforms In India The
Post- Reforms Era”, Indian Journal of Research, Vol.3, Issue No.4, pp.no.14-18
2) Abhiman Das, SaibalGhosh, (2006), “Financial Deregulation And Efficiency : An Empirical
Analysis Of Indian Banks During The Post Reform Period ”, Review of Financial
Economics, Vol.15, pp.no.193-221
3) A.K. Mishra, J. Gadhia, (2013), “Are Private Sector Banks Are More Sound And Efficienty
Than Public Sector Banks ?Assessments Based On Camel And Data Envelopment Analysis
Approaches”, Research Journal of Recent Sciences, Vol.2(4), pp.no.28-35
4) Amit Kumar Dwivedi, D. Charyulu, (2011), “Efficiency Of Banking Industry In The Post
Reform Area”, Research And Publication Of Indian Institute of Management w.p. no. 2011-
03-01, pp.no.1-15
5) Arora, GurpreetKaur, (2014), “Evaluating Efficiency In Indian Banking Sector Using Data
Enevelopment Analysis”, International Journal of Economic, Commerce and Management,
Vol.II, Issue No.8, pp.no.1-13
6) Avkiran, N. K. , (1999), The evidence of efficiency gains: The role of mergers and the
benefits to the public, Journal of Banking and Finance 23, 991-1013
7) Don U. Galgedara, P. Edirisuriya, (Performance Of Indian Commercial Banks (1995-2002:
Applicatin Of Data Envelopment Analysis And Malmquist Productivity Index),Retrieved
from http://128.118.178.162/eps/fin/papers/0408/0408006.pdf, pp.no.1-31
8) I.A. Shah, S. Shah, H.Ahmad, 2012, “Comparing The Efficiency Islamic Versus
Conventional Banking : Through Data Envelopment Analysis”, Africal Journal of Business
Management, Vol.3, Issue 6,pp.no.787-797
9) M Duygun- Fethi, (2009), “Assessing Bank Performance With Operational Research And
Artificial Intelligence Techniques : A Survey”, University Of Bath School Of Management,
Working Paper Series 2009.02 , pp.no.1-65
10) MajidKarimzadeh ,(2012), “Efficiency analysis by using Data Envelopment Analysis Model:
Evidence from Indian banks”, International Journal of Latest Trends In Financial ,
Economics Science, Vol.2,No.3, pp.no.228-237
11) MilindSathye, “Efficiency Of Banks In A Developing Economy : A Case Of India”,
Retrieved
from https://crawford.anu.edu.au/acde/asarc/pdf/papers/conference/CONF2001_13.pdf.
12) Mukesh Kumar, Vincent Charles, (2012), “Evaluating The Performance Of Indian Banking
Sector Using Data Envelopment Analysis During Post Reform And Global Financial Cisis ”,
Retrieved from http://centrum.pucp.edu.pe/pdf/working_paper_series/CEFE_WP2012-09-
0007.pdfworking paper series no. 2012-09-0007,pp.no.1-28
13) ShaikhSaleem, M. Reddy, (2014), “A Study Of Operational Performance Of Public, Private
And Foreign Sector Banks”, Tactful Management Research Journal, Vol.2 , Issue 4, pp.no.1-6

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 243
Finance Management

14) Sunil kumar, R. Gulati, (2008), “An Examination Of Technical, Pure Efficiency And Scale
Efficiencies In Indian Public Sector Banks Using Data Envelopment Analysis”, Eurosian
Journal of Business And Economics , Vol.1, Issue 2, Pp.No.33-68
15) Vinod R.R., (2013), “Efficiency Of Old Private Sector Banks In India : A DEA Approach”,
International Journal of Management And Social Sciences Of Research, Vol.2 No.6,
pp.no.78-82

Appendix 1 – DEA Efficiency scores of nationalized banks 2012-2014


CRS(TE) VRS CRS(TE) VRS CRS(TE) VRS
Nationalise Banks TE TE SE TE TE SE TE TE SE
Allahabad Bank 1.00 1.00 1.00 0.74 0.74 1.00 0.72 1.00 0.72
Andhra Bank 0.55 0.85 0.65 0.04 0.45 0.09 1.00 1.00 1.00
Bank of Baroda 0.76 1.00 0.76 0.75 0.90 0.83 0.65 0.97 0.67
Bank of India 0.65 0.78 0.83 0.72 0.84 0.86 0.56 0.75 0.75
Bank of Maharashtra 0.92 1.00 0.92 1.00 1.00 1.00 0.74 0.82 0.90
Canara Bank 0.67 0.81 0.83 0.67 0.73 0.92 0.63 0.82 0.77
Central Bank of India 0.74 0.83 0.89 0.72 0.73 0.99 0.65 0.85 0.76
Corporation Bank 0.18 0.49 0.37 0.82 1.00 0.82 0.53 0.66 0.80
Dena Bank 1.00 1.00 1.00 0.97 1.00 0.97 1.00 1.00 1.00
Indian Bank 0.07 0.44 0.16 0.10 0.37 0.27 0.67 0.90 0.74
Indian Overseas Bank 0.77 0.82 0.94 0.74 0.74 1.00 0.65 0.86 0.76
Oriental Bank of Commerce 0.84 0.90 0.93 0.82 0.85 0.96 0.79 1.00 0.79
Punjab & Sind Bank 0.79 0.80 0.99 1.00 1.00 1.00 1.00 1.00 1.00
Punjab National Bank 0.75 1.00 0.75 0.71 0.85 0.84 0.68 1.00 0.68
Syndicate Bank 1.00 1.00 1.00 1.00 1.00 1.00 0.73 0.97 0.75
UCO Bank 1.00 1.00 1.00 1.00 1.00 1.00 0.87 1.00 0.87
Union Bank of India 0.72 0.85 0.85 0.72 0.78 0.92 0.64 0.86 0.74
United Bank of India 0.86 1.00 0.86 0.91 1.00 0.91 0.04 0.59 0.07
Vijaya Bank 0.86 1.00 0.86 0.83 1.00 0.83 0.74 0.74 1.00
Average 0.74 0.87 0.82 0.75 0.84 0.85 0.70 0.88 0.78
Standard Deviation 0.2536 0.17 0.222 0.2664 0.187 0.249 0.2125 0.13 0.205

******

ISSN : 2230-9667 Chronicle of the Neville Wadia Institute of Management Studies & Research 244

Vous aimerez peut-être aussi