Download CHAPTER 6 ECONOMETRIC RESULTS OF COST FUNCTIONS

CHAPTER 6
ECONOMETRIC RESULTS OF COST FUNCTIONS ±
CROSS-SECTIONAL ANALYSIS
The cross-section
three
years,
data was drawn for 28 public sector banks
viz. 1985, 1986 and 1987.
For the
for
analysis,
the
average of these years was used.
The
dependent
variable,
cost,
is
defined
expenses (TCE) net of interest expenses.
namely,
the
as
total
current
Two components of
establishment (labour) expenses and
other
(non-labour) expenses have also been used as dependent
TCE
current
variables
to investigate the sources of economies of scale results.
Six different alternative output measures have been used.
These
are:
i)
ii)
iii)
iv)
v)
vi)
Out
of
Volume of Business (VB)
Total Assets (TA)
Earning Assets (EA)
Total Deposits (TD)
Total Operating Earnings (TOE)
Total Number of Branches (TB)
these six output variables, first five are
in
(Rs.) terms and the last one is in physical (number of
term
and
has
been used to find out the
branch
monetary
branches)
economies
and
diseconomies in banking.
Like
the time series analysis, we have estimated four
the cost functions viz., cubic, quadratic, linear and
for
each cost variable and for each of six output
115
forms
of
double-log
variables.The
best-fit
for
empirical results (estimated cost
equations)
selected
all the types of costs and outputs are given in Appendix
V,
Table No. 1 to 6.
The
estimated cost equations are discussed in detail
then the cost
6.1
6.1.1
1.
below
and
elasticities have been derived at.
The Analysis of Cost Equations
Dependent Variable : Total Current Expenses (TCE)
Out
of
the total six cost equations
output
variables,
equation is
five
selected
for
six
cubic,
and
one
equations are
quadratic in functional form.
The quadratic
cost equation is for the TB output variable.
2.
R
(0.98)
output
3.
There
is maximum in the cost equation
TA
as
variable.
are only two cost equations which list
variables.
These
output
mix
variables
equations with TA and TOE output variables.
mix
with
variables
output-mix
are
The
with signs of their coefficients
for
cost
outputare
as
and
RB
follows:
4.
None
Positive
Negative
COMM/TOE
FA/TA
OA/TA
of the estimated cost equations include TB
variables.
116
6.1,2
1.
Dependent Variable:
Out
of the total six cost equations, five are cubic
one is
2.
3.
The
Establishment Expenses (EC)
and
quadratic in functional form.
maximum
—2
R
(0.996)
is found to
be
in
equations with VB and TD as
output varibles.
The
are
output-mix
significantly
equations
variables
in
found
to
only three cost equations,
with VB, TOE and TB as output
the
be
cost
listed
viz.
cost
variables
with
the following signs of their coefficients :
Positive
Negative
BORR./VB
COMM/TOE
4.
RB/TB
SU/TB
U/TB
Except in the case of the cost equation with TA as output
variable,
all
variable
with
coefficients.
branch
other
cost
equations
have
’a
priori'
positive
signs
This
finding
suggests
that
listed
TB
of
the
there
are
diseconomies existing in the case of labour
cost
in public sector banks.
6.1.3
1.
Dependent Variable : Other Current (Non-labour) Expenses (0C)
Except for the cost equation with TB as output
variable,
which
equations
is quadratic in output, all other
are in cubic functional form.
117
cost
2.
The
maximum R^
(0,992) is for cost equation with VB
as
output variable.
3.
None
of
the non-labour cost equations
list
output-mix
variables.
4.
The
ratio of rural branches to total number of
(RB)
has emerged as an important
listed
in
five
equations.
independent
The
branches
variable,
coefficients
of
this
variable are with negative signs implying thereby that an
increase in RB is non-labour cost advantageous.
6.1.4
1)
Conclusions
The estimated cost equations with all the cost components
and
output variables are found to be statistically
significant with high values of R
2)
o
very
and R .
Fifteen out of total eighteen cost equations are cubic in
functional form.
3)
Total
Asset
(TA) has turned out to be the
best
output
variable for the Indian Banking Industry.
4)
Output-mix
variable included in the cost
functions
found to be not having significant influence over
are
either
of the cost components.
5)
In the case of labour cost, total number of branches (TB)
is
listed in all the cost equations with positive
of
its
coefficients, implying thereby that
expansion leads to increasing wage bills.
118
the
signs
branch
6)
Cost
equations
with
non-labour
cost
as
a
dependent
variable have listed the independent variable, the
of
rural branches to total branches (RB)
sign.
with
ratio
negative
This finding indicates the economical use of
non­
labour cost in the rural branches.
6.2
The Elasticity Results
The
output
elasticities of different types of costs
calculated
in
analysis.
Further,
enjoying
the way as discussed earlier in
banks
the
having elasticities
have
been
time-series
below
one
economies of scale, those with above one are
suffering
from
economies of scale and those having around
one ( +
have
reached
of output
the
optimum size both,in respect
are
0.5
)
and
branches.
The
sample of the cross-section analysis consists of
sector
banks and elasticities have been calculated for
individual
costs
28'
sample banks.
public
all
The elasticities for various types
the
of
and output have been shown in Appendix VI,Table Numbers
7
to 9.
The results on cost elasticities of various outputs are found
to
be sensitive to the output measures used in the cross - sectional
analysis,
unlike
time-series
elasticity
results, five monetary based output variables can
divided into two sets.
variables,
namely
elasticities.
analysis.
On
the
basis
of
be
The first set, consisting of three output
TA, EA and TOE giving more
or
less
Another set of output includes VB and TD,
119
similar
having
similar elasticities but different from the former set.
The cost
elasticity
(TB)
totallly
case,
vith
respect
to total number of
branches
different from these two sets of elasticities. In
for policy inference, the following
two criteria
is
such
can
be
on
the
used :
1.
Selection
of
basis
of statistical inferences and cost
of
the best-fit cost/output equation
that particular
sectional
asset
the
(TA)
output variable.
analysis,
turns
as
out
discussed
to be
best
elasticity
In the
cross-
earlier,
total
output
basis of statistical inference and the
measure
on
hypothesised
theoretical preference.
2.
Using the surrogative measure i.e. simple average of
five
To
make
elassticities.
the policy inferences more significant, the
elasticities
similar.
values
from these two criteria have turned out to be
Hence,
elasticities
the
for
policy
can be used.
measures,
any
of
of
very
these two
For this analysis we have chosen
the
surrogative measure of elasticity for policy inference.
Further,
the error term turned out to be rather large
for
banks, namely PNB, BARODA, CNTRL, and CANRA, when EA was used
four
as
output variable and for four banks, namely INDIN, PNB, CNTRL, and
CANRA,
when TOE was used as output variable.
These
banks
with
larger error terms are termed as out-liers in regression analysis
and in order to miligate their consequences on elasticities,
corresponding elasticities have been ignored in the analysis.
120
the
The
results
elasticity
component
of
of
measure
of
elasticity
have been presented here
results for all the
with
twenty
and
cost
the
cost
eight
These elasticity results are presented
three bank groups:
banks
the
branch
elasticity
sector banks.
for
surrogative
public
separately
(A) banks having economies of scale
having diseconomies of scale and
(B)
(C) banks operating
optimum level, in Table No. 1, 2, and 3
at
respectively.
The
elasticity results and findings are given below ;
6.2.1
Banks Having Economies of Scale
Table No. 1
Elasticity Results of Banks Having Scale Economies
Total Branch el, of
Surrogative el. of
BANKS
TCE
CANRA
PNB
CNTRL
INDIA
SYNDI
BRODA
UCO
UNION
IOB
UNITED
INDIAN
0.42
0.44
0.45
0.46
0.48
0.50
0.62
0.69
0.75
0.87
0.92
EC
OC
TCE
EC
OC
0.59
0.62
0.54
0.64
0.57
0.55
0.57
0.50
0.65
0.59
0.64
0.53
0.61
0.72
0.79
0.65
0.67
0.92
0.87
0.89
1.24
0.79
0.82
1,24
1.32
0.89
0.72
0.94
1.12
1.16
0.73
0.86
0.91
0.96
1.37
1.36
0.96
0.87
1.04
1.22
1.30
0.91
1.00
1.12
0.81
1.31
1.68
1.20
0.80
1.15
1.54
1.48
0.83
1.25
1.09
Findings and Policy Implications
1.
Eleven
out
economies
of
of
28
public
scale.
All
sector
of
banks
them
are
belonged
enjoying
to
the
nationalised bank group.These banks are advised to strive
for expansion of the volume of business.
121
2.
Both
the cost components of total cost viz., labour
non-labour costs are also showing the economies of
(except
for
elasticities
UNITED)
of
labour
as
indicated
and
by
scale
the
non-labour
and
output
costs.
These
elasticities are significantly less than unity.
3.
Though,
all these eleven nationalised banks are
the fruits of conoaies of scale, four of them,
reaping
mentioned
below, are suffering from branch diseconomies.
i)
ii)
i i i)
iv)
PNB
CNTRL
UCO
UNION
The
above four banks would be better-off
the
turnover
of their existing
by
increasing
branches.
The
policy
implication is that the average size per branch should be
increased.
Further, both the cost components
cost
showing
are
economies.
increasing
with
the
Hence,
the
branch
diseconomies
utlimate
objective
of
total
but,
their policies should be
output and reducing the
of
tuned
branch
optimizing
scale
for
network
the
cost
allocation.
4.
Remaining
seven banks, listed below, are enjoying
both,
scale as well as branch economies, though, very marginal.
i)
ii)
iii)
iv)
V)
vi)
vii)
CANRA
INDIA
SYNDI
BRODA
IOB
UNITED
INDIAN
122
These hanks are ideal for launching all out programs
expansion,
both
turnover.
branch
in
terms
of
branches
and
for
business
Hence, they should be chosen for granting
opening
licenses by the Reserve
Bank
of
new
India
whereever possible.
6.2.2
Banks Having Diseconomies of Scale
Table No. 2
Elasticity Results of Banks Experiencing Diseconomies of Scale
Surrogative el. of
Total Branch el. of
BANKS
SB I
PATLA
TRVCR
INDOR
ALLAHA
NAHA
PN&SB
NEWBK
CORPN
ORNTL
ANDRA
TCE
EC
OC
TCE
EC
4.17
1.40
1.16
1. 13
1.08
1 .09
1.09
1.12
1.15
1.15
1.21
2.03
0.86
0.67
0.63
0.69
0.65
0.64
0.61
0.70
0.74
0.77
2.80
1.36
1,16
1.04
1.33
1.51
1.05
1.31
0.98
0.95
1.21
1.61
0.82
0.62
0.62
1.52
1.11
0.77
0.75
0.69
0.71
1.06
OC
1.47 .
1.09 '
1.09
0.85
1.12
1.26
1,00
0.92
0.97
1.02
1.35
1.52
1.07
1.67
0.85
2.13
1.83
1.01
1.21
0.83
0.79
1.32
Findings and Policy Implications
1.
Diseconomies
of scale are found in eleven public
banks, given in Table No. 2.
follow
a
emphasising
policy
on
of
improved
sector
It is recommended that they
consolidation
quality
of
of
business,
lending,
deposit
mobilisation etc.
2.
The source of diseconomies of scale in all these banks is
major
cost component, the non-labour cost.
123
The
labour
cost
is
except
shoving
SBI.
This
economies of scale in
implies
that
economise on non-labour expenses
needs
to
all
these
the
hanks
hanks
should
in future, where as SBI
economise both, on the labour
and
non-labour
components of cost.
3.
Seven,
out of a total of eleven banks with
diseconomies
of scale, are enjoying branch economies. These banks arei)
vi)
vii)
PATLA
TRVCR
INDOR
PN & SB
NEW BK
CORPN
ORNTL
These
banks
ii)
i i i)
iv)
V)
the
expand
to
needs
need
number of branches.
be
total
to contract their
planned
The
in a manner
output
branch
soi
as
and
expansion
to
transfer
business from some of the very big branches to the
newly
opened branches.
4.
Four
banks,
given below, are suffering from
both,
the
output
and
diseconomies of scale and branches:
i)
ii)
iii)
iv)
SBI
ALLAHA
MAHA
ANDRA
These
banks are too large in terms of both,
branch
network.
To economise on costs,
embark on a consolidation program.
they
need
to
Two or three branches
situated in a compact business area scould be merged into
one.
New business need to be accepted very
after overall
scrutiny.
124
selectively
6.2.3
Banks Operating at the Optimum
Table No. 3
Elasticity Results of Banks with Output Optimality
Surrogative el. of
Total Branch el. of
BANKS
“ h* l-»
O
• *
n
H*
r»4
o (0
vO43 vo
oo r—in oo oo
m
<
i-*oo
00 -4 -4 CO -4 00
OC
4 h \0 O'h h
EC
in ,-t cs
r*-
oooooo
*«4
o ©
in
in O' r" r» in vo
i
^4
vo O'
0.98
1.03
1.03
1.05
1.04
1.00
TCE
»
OC
oin oin oinoindvooin
BK&JP
HYDER
MYSOR
SRSTR
VIJYA
DENA
EC
H H O O H O
O — *4 vfl O £
•4 O K5 U M M
TCE
Findings and Policy Implications
1.
Only
six
banks,
operating
2.
These
out of 28
public
sector
are
at the optimum capacity.
banks (except DENA) are having
Hence,
banks,
branch
to gain further on cost efficiency,
economies.
they
increase their branch network and transfer some
from
heavily loaded branches to the newly
Dena
Bank
is
suffering
from
very
should
business
opened
marginal
ones.
branch
diseconomies.
3.
These
banks with optimum size are found to
economies
of scale in labour cost
scale
non-labour
control
in
the
and
cost. Hence,these
enjoying
diseconomies
banks
inefficient use of non-labour
bringing about reduction in non-labour cost.
125
be
need
factors
of
to
for
6*3
1.
Overal1 Findings
Out of the total 28 public sector banks,eleven banks have
economies
of
diseconomies
scale and the same number
of
banks
have
of scale and the remaining six
banks
have
reached the optimality
2.
All
the
point,
public sector banks (except SBI)
economies of scale in labour cost.
banks
labour
cost.
oldest
one
As
But, majority of
SBI
which is the biggest
also, has
diseconomies.
126
bank
diseconomies of scale
the
in
non­
and
the
in
labour
in overall cost.
many as nine public sector banks are
branch
enjoying
are suffering from diseconomies of scale
cost, non-labour cost as well as
3.
are
suffering
from