Bank marketing investments and bank performance

Description
The purpose of this paper is to examine empirically the effects of investments by US
banks in advertising and promotion on their performance in the areas of profits and market share.

Journal of Financial Economic Policy
Bank marketing investments and bank performance
Donald J . Mullineaux Mark K. Pyles
Article information:
To cite this document:
Donald J . Mullineaux Mark K. Pyles, (2010),"Bank marketing investments and bank performance", J ournal
of Financial Economic Policy, Vol. 2 Iss 4 pp. 326 - 345
Permanent link to this document:http://dx.doi.org/10.1108/17576381011100856
Downloaded on: 24 January 2016, At: 21:39 (PT)
References: this document contains references to 44 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1241 times since 2010*
Users who downloaded this article also downloaded:
Ahmed Arif, Ahmed Nauman Anees, (2012),"Liquidity risk and performance of banking
system", J ournal of Financial Regulation and Compliance, Vol. 20 Iss 2 pp. 182-195 http://
dx.doi.org/10.1108/13581981211218342
Ram N. Acharya, Albert Kagan, Srinivasa Rao Lingam, (2008),"Online banking applications and
community bank performance", International J ournal of Bank Marketing, Vol. 26 Iss 6 pp. 418-439 http://
dx.doi.org/10.1108/02652320810902442
Subrata Sarkar, Rudra Sensarma, (2010),"Partial privatization and bank performance:
evidence from India", J ournal of Financial Economic Policy, Vol. 2 Iss 4 pp. 276-306 http://
dx.doi.org/10.1108/17576381011100838
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
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.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Bank marketing investments
and bank performance
Donald J. Mullineaux
Gatton College of Business and Economics, University of Kentucky,
Lexington, Kentucky, USA, and
Mark K. Pyles
College of Charleston, Charleston, South Carolina, USA
Abstract
Purpose – The purpose of this paper is to examine empirically the effects of investments by US
banks in advertising and promotion on their performance in the areas of pro?ts and market share.
Design/methodology/approach – The model presented in the paper is motivated by the theory of
the pro?t function. We estimate a base model with a ?xed-effects panel including an AR(1) disturbance
over the period 2002-2006. To test for selection bias, we also estimate a Heckman model.
Findings – It is found that bank pro?ts and market share increase signi?cantly with increased
spending on advertising and promotion. Also, signi?cant evidence is found of increasing returns to scale
in this type of marketing expenditure. It is also found that increased expenditures on branching result in
higher pro?ts and increased market share, but without scale effects. The results are robust, the inclusion
of variables is not suggested by pro?t function theory and corrected for prospective selection bias.
Originality/value – The extant literature does not include research on the effectiveness of bank
marketing from the viewpoint of its impact on pro?t performance. The ?ndings should be of interest to
academics in ?nance and marketing and to banking practitioners.
Keywords Banks, Advertising, Pro?ts, Market share
Paper type Research paper
1. Introduction
Like other ?rms, commercial banking organizations spend substantial sums annually
on marketing their various products and services to actual and potential customers.
Very little is known about the effectiveness or productivity of these investments,
however[1]. Indeed, this situation is not unique to bank marketing. Writing in one of
the leading academic journals in marketing, Rust et al. (2004, p. 76) observe that:
Marketing practitioners and scholars are under increased pressure to be more accountable for
and to show how marketing expenditure adds to shareholder value. The perceived lack of
accountability has undermined marketing’s credibility, threatened marketing’s standing
within the ?rm, and even threatened marketing’s existence as a distinct capability within the
?rm[2].
Our goal in this somewhat interdisciplinary paper is to analyze whether “marketing
pays” in banking and, if so, to what extent. The answer to this question should be of
interest to a wide audience, including academics in ?nance and marketing, bank
managers and bank shareholders.
Why else focus such a study on the banking industry? One answer is that data on
marketing expenditures are more widely available in this industry. The few studies
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JFEP
2,4
326
Journal of Financial Economic Policy
Vol. 2 No. 4, 2010
pp. 326-345
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381011100856
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
that address marketing’s ?nancial effectiveness rely on Compustat data, which is
con?ned to public ?rms. We gather our data from reports banks are required to ?le
quarterly (“Call Reports”) with the bank regulatory agencies. Many of these banks are
either privately held or subsidiaries of bank holding companies that own the shares
of the subsidiary banks. Consequently, we can obtain a larger sample than those
commonly employed in non-bank-related studies of this issue. While only a subset of
banks report their marketing expenditures, we nonetheless obtain sizeable samples
that cover the period from 2002 to 2006.
We study the impact of marketing on accounting pro?ts, employing a
well-established analytical framework known as the “bank pro?t function,” which
has been widely employed in the literature to study the production characteristics of the
banking industry[3]. While prior research in this area takes account of investments in
bank branch networks, it ignores investments in brand equity via advertising and
promotion. Branchnetworks are themselves critical components of a marketingstrategy
since they represent perhaps the primary mechanism for delivering bank products and
services[4]. Our empirical model permits us to estimate the relative rates of return on
each type of investment as re?ected in the relevant pro?t elasticity.
We also examine the relationship between market share of bank deposits and
marketing investments. Increased market share could be one of the channels through
which marketing affects bank performance. In fact, some studies in the marketing
literature (incorrectly, in our view) include both marketing and market share-related
variables as explanatory factors in models of either pro?t or capital market performance.
One drawback of our approach is that we are unable to relate marketing
investments to value creation in banking. Most of the institutions in our sample are
either private institutions or owned by bank or ?nancial holding companies.
Consequently, there is no trading in the equity of these banks. Holding company stocks
do trade in most cases, but, unfortunately, to our knowledge marketing expenditures are
not reported at the holding company level for our time period studied. Nor can we simply
add up the marketing expenditures of the bank subsidiaries since many holding
companies also own non-bank subsidiaries that presumably spend on marketing as
well[5]. There are available data on marketing for holding companies starting in 2005
and we plan to extend this research to the holding company arena in a subsequent
research.
We ?nd, as hypothesized, that pro?ts and market shares increase with increased
expenditures on advertising and promotion and on opening newbranches. The ?ndings
are robust to controlling for ?rm-speci?c effects. We also ?nd evidence of increasing
returns to scale in advertising and promotion investment for both pro?ts and market
share. Branch expansion is not characterized by increasing returns to scale, however.
Instead, the evidence indicates constant or decreasing returns to scale, the latter
particularly in the model for market share. Our ?ndings on the returns to marketing are
robust in all cases to the inclusion of factors in the model, such as accounting measures
of risk, market concentration, and capital or liquidity variables. Also, the results do not
appear to be a product of selection bias.
We organize the discussion in the following way: Section 2 offers a brief review of
prior research. Section 3 describes the data and the models estimated. Section 4 presents
the results for the estimated pro?t function and Section 5 for the market share model.
Section 6 concludes.
Bank marketing
investments
327
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
2. Brief review of the literature
We are aware of only one refereed paper that focuses on the impact of marketing
expenditures on ?nancial institution performance. Hasan et al. (2000) ?nd no
relationship over the period 1985-1989 between increased promotion expenditures by
thrift institutions in the Southeast US and thrift pro?tability, measured as return on
assets. Marketing expenditures are positively related to the ratio of non-interest income
scaled by assets, however. DeYoung and Ors (2004) study the determinants of
advertising spending at US thrift institutions over the period 1994-2000, along with the
in?uence of advertising on deposit rates.
Despite calls for increasedefforts to link?nance and marketingresearch, relativelyfew
papers witha marketingorientationhave appearedin?nance journals[6]. One exceptionis
a studypublishedbyChauvinandHirschey(1993) inFinancial Management that reports a
positive relationship between advertising spending and ?rm value that becomes larger
with ?rm size. Marketing researchers have been somewhat more active in exploring the
?nancial effects of marketing investments. A “Retrospective” on the ?nance/marketing
interface appeared in the Journal of Marketing Science in 2005 (Vol. 33, No. 4), including a
survey by Conchar et al. (2005) of a number of studies on the link between advertising and
promotion spending and market valuation, including event studies and efforts to estimate
the relationshipbetweenmarketingandeconomic value addedand/or various measures of
market capitalization.
Ben Zion (1978) was the ?rst to uncover a positive relation between market value
and advertising expenditures. Lane and Jacobson (1995) ?nd that brand extension
announcements produce abnormal equity returns. Cornwell et al. (2005) show that
?rms announcing sponsorships of major athletic teams or leagues also experience
positive abnormal returns[7]. Balasubramanian et al. (2005) report that ?rm’s receiving
a Malcolm Baldridge National Quality Award experience positive excess equity
returns. Singh et al. (2005) ?nd that ?rms with higher advertising expenditures have a
lower cost of capital, con?rming a result reported earlier by Gleason et al. (2000).
Conchar et al. (2005) performa meta-analysis of 15 published studies that focus on the
relationship between marketing activities and various measures of ?rm valuation[8].
Based on an analysis of 88 different models estimated in these studies, they ?nd an
average positive (and signi?cant) relationship between advertising and promotion
expenses and ?rm value. They note that some of the studies suffer from endogeneity
problems (market share is often used an explanatory variable, for instance). Compustat
is the data source in about two thirds of the studies and all the ?rms studied are public
entities. None of these valuation-based studies focuses on banks.
While research on the impact of bank marketing on market share is sparse, a
sizeable literature addresses the relative ef?ciency of banks in relation to market
concentration and/or market share (Smirlock, 1985; Berger and Hannan, 1989; Berger,
1995a; Scholtens, 2000). The discussion focuses primarily on the reasons for the
observed positive relationship between market concentration and pro?tability.
Smirlock (1985), building upon Demsetz (1973) and Peltzman (1977), ?nds evidence in
support of the “ef?cient structure hypothesis,” which implies that market share proxies
for ef?ciency. These authors claim that superior ef?ciency of the leading ?rms in the
market accounts for their high market share. Accordingly, any relationship between
concentration and pro?tability is spurious and is rendered mute when market share is
included in the model.
JFEP
2,4
328
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
An alternative viewis the “structure-conduct-performance hypothesis,” which posits
that concentration and pro?tability will be related in non-competitive markets because
of pricing power. Berger and Hannan (1989) provide some support for this view.
Scholtens (2000) ?nds only a weak relationship between concentration and pro?tability,
however, and concludes that concentration does not result in monopoly pro?ts. Berger
(1995a) presents an in-depth examination of both the ef?cient structure and
structure-conduct-performance hypotheses, ?nding weak support for both, but
concludes than neither is of great importance in determining bank pro?ts.
3. Data and model
We gather data from the Call Reports submitted by banks to the regulatory agencies
on a quarterly basis. Under the category “other non-interest expense,” banks are
required to report “advertising, promotional public relations, marketing and business
development expenses” if these exceed 1 percent of the sum of interest income
plus non-interest income. The proportion of banks reporting marketing expenditures
averages 60.9 percent over the period 2002-2006 and ranges from 57.4 (2002) to
63.0 (2005) percent[9]. Unfortunately, we cannot discriminate among the various
subcategories of marketing expenditures, but conversations with bankers indicate
the ?gures primarily represent dollars spent on advertising and promotion. We
disaggregate our sample according to bank asset size, using the classi?cations
employed by the Federal Deposit Insurance Corporation (FDIC) in its Quarterly Banking
Pro?le: assets ,$100 million, assets between $100 million and $1 billion, assets $1 to
$10 billion, and assets .$10 billion. In most years, the percentage of banks reporting
marketing increases with bank size.
Most of the remainder of the data used in the estimations comes from the same
source. The exception is the number of branches and the market share for each
institution, which we obtain from the FDIC’s web site. Tables I and II present some
descriptive statistics on bank marketing expenses and other data used to estimate the
models over the period 2002-2006. The data represent averages per bank. Advertising
and promotion spending per bank have grown at a compound annual rate of 4.7 percent
over the sample period, whereas the average number of branches has increased more
slowly and declined in 2006 from 2005. Not surprisingly, larger institutions spend more
money on marketing. The banks in the sample that exceed $10 billion in assets spent
an average of $153 million annually, on average, over the sample period vs $3.6 million
for banks in the $1-$10 billion size class. The largest banks spend about 43 times as
much on advertising and promotion as the banks with assets over $1 but ,$10 billion
dollars in assets[10].
There are also sizeable differences in spending within each of the asset size classes.
For example, in the largest size group, the minimum spending on this category of
marketing in 2006 was $7 million, but the maximum was over $1.6 billion. Relative to
assets, banks spend similar amounts in the three smaller classes (0.11-0.13 percent), but
the largest banks spend about three times more than this. When marketing is scaled by
revenues, the smallest banks and the largest ones have higher ratios than the
institutions “in the middle.” The main factor accounting for the relatively high ratio at
the smallest banks is that these institutions generate much less non-interest income
than their larger counterparts. A very high percentage of their fee income comes from
service charges on deposits, which have been trending downwards in recent years.
Bank marketing
investments
329
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
A
v
e
r
a
g
e
s
n
4
,
6
5
6
4
,
7
9
4
4
,
9
6
8
4
,
9
0
0
4
,
6
6
8
4
,
7
9
7
M
k
t
1
,
5
7
7
1
,
5
3
7
1
,
6
7
5
1
,
7
5
4
1
8
9
4
1
,
6
8
7
S
a
l
B
e
n
E
x
p
1
0
,
4
2
2
1
1
,
0
3
8
1
1
,
6
6
9
1
2
,
9
9
0
1
4
,
0
0
7
1
2
,
0
2
6
N
u
m
E
m
p
1
9
6
1
9
2
2
0
9
2
1
4
2
2
3
2
0
7
B
r
a
n
c
h
e
s
6
.
3
7
6
.
4
9
7
.
0
9
7
.
3
8
6
.
9
4
6
.
8
6
T
A
6
5
0
,
7
8
1
6
9
5
,
4
6
8
7
7
6
,
4
5
5
8
4
5
,
2
9
6
8
8
1
,
3
0
7
7
7
0
,
7
9
1
S
T
A
s
s
e
t
s
4
1
,
1
7
5
4
5
,
1
5
8
4
2
,
2
9
9
4
3
,
8
1
3
4
7
,
6
0
6
4
4
,
0
0
9
I
n
c
o
m
e
v
a
l
u
e
s
I
n
t
I
n
c
3
8
,
3
3
6
3
5
,
4
1
4
3
6
,
8
6
1
4
5
,
1
8
2
5
6
,
3
7
8
4
2
,
3
5
6
N
o
n
I
n
t
I
n
c
1
7
,
4
9
3
1
7
,
6
7
7
1
8
,
5
7
9
1
9
,
5
1
8
2
1
,
2
4
8
1
8
,
8
9
9
R
e
a
l
E
s
t
a
t
e
I
n
c
1
2
,
8
3
4
1
3
,
1
4
8
1
3
,
6
8
1
1
8
,
0
8
4
2
3
,
2
0
8
1
6
,
1
6
8
I
n
d
i
v
i
d
u
a
l
I
n
c
7
,
2
3
5
6
,
4
9
5
7
,
0
8
2
7
,
8
2
4
8
,
9
5
0
7
,
5
1
0
C
o
m
I
n
d
I
n
c
4
,
9
8
5
4
,
1
9
3
3
,
8
1
3
5
,
0
1
6
6
,
4
0
3
4
,
8
6
6
T
o
t
R
e
v
5
5
,
8
3
0
5
3
,
0
9
0
5
5
,
4
4
0
6
4
,
7
0
0
7
7
,
6
2
6
6
1
,
2
5
5
D
e
p
o
s
i
t
v
a
l
u
e
s
T
o
t
D
e
p
4
0
5
,
3
4
7
3
5
0
,
4
5
1
3
9
9
,
1
5
1
4
2
9
,
8
5
6
4
3
4
,
1
5
0
4
0
3
,
8
2
9
D
e
m
D
e
p
4
7
,
6
6
2
4
7
,
4
8
4
5
1
,
4
6
5
5
0
,
2
7
1
5
0
,
0
7
8
4
9
,
3
9
4
I
n
t
D
e
p
7
,
2
8
9
5
,
5
9
7
5
,
3
0
9
8
,
6
9
9
1
3
,
5
1
9
8
,
0
6
2
M
k
t
S
h
r
e
D
e
p
0
.
3
4
0
.
3
9
0
.
3
3
0
.
3
4
0
.
3
1
0
.
3
4
E
x
p
e
n
s
e
v
a
l
u
e
s
I
n
t
E
x
p
1
2
,
5
0
6
9
,
8
4
7
9
,
7
8
3
1
6
,
0
5
3
2
5
,
2
4
8
1
4
,
6
1
5
N
o
n
I
n
t
E
x
p
2
4
,
1
6
9
2
4
,
7
7
4
2
6
,
8
7
1
2
8
,
9
5
8
3
0
,
9
2
9
2
7
,
1
4
3
T
o
t
E
x
p
3
6
,
6
7
5
3
4
,
6
2
0
3
6
,
3
5
4
4
5
,
0
1
1
5
6
,
1
7
7
4
1
,
7
5
8
L
o
a
n
v
a
l
u
e
s
R
e
a
l
E
s
t
a
t
e
L
o
a
n
s
2
0
6
,
1
2
4
2
3
2
,
0
8
9
2
6
8
,
4
9
3
3
0
9
,
2
6
9
3
1
5
,
4
3
0
2
6
7
,
0
1
8
I
n
d
i
v
i
d
u
a
l
L
o
a
n
s
8
3
,
4
2
8
8
6
,
6
3
5
9
5
,
9
5
4
1
0
3
,
0
7
0
1
0
4
,
5
8
0
9
4
,
8
7
8
C
o
m
I
n
d
L
o
a
n
s
7
5
,
3
1
9
6
9
,
0
8
2
7
1
,
2
9
3
7
7
,
2
5
3
8
2
,
6
6
2
7
5
,
0
9
4
L
o
a
n
C
h
a
r
g
e
O
f
f
s
6
,
1
4
9
5
,
4
9
6
4
,
5
8
5
5
,
0
3
2
4
,
3
7
7
5
,
1
1
9
N
o
n
-
P
f
m
i
n
g
L
n
s
1
,
8
6
8
2
,
3
2
7
2
,
0
1
8
2
,
4
1
5
2
,
9
5
0
2
,
3
1
5
L
o
a
n
R
e
c
8
6
8
9
3
0
1
,
0
1
0
1
,
1
4
8
1
,
0
5
0
1
,
0
0
3
T
o
t
L
o
a
n
s
4
2
0
,
0
0
3
4
4
5
,
6
1
2
5
0
8
,
2
0
2
5
5
6
,
6
6
0
6
0
0
,
2
0
5
5
0
6
,
7
1
9
C
a
l
c
u
l
a
t
e
d
v
a
l
u
e
s
R
a
t
e
R
E
7
.
2
1
6
.
5
9
6
.
1
7
6
.
5
2
8
.
7
7
7
.
0
3
R
a
t
e
I
n
d
9
.
5
5
8
.
6
7
8
.
4
3
8
.
6
5
9
.
1
8
8
.
8
8
(
c
o
n
t
i
n
u
e
d
)
Table I.
Summary statistics
JFEP
2,4
330
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
2
0
0
2
2
0
0
3
2
0
0
4
2
0
0
5
2
0
0
6
A
v
e
r
a
g
e
s
R
a
t
e
C
o
m
I
n
d
7
.
8
2
7
.
1
0
6
.
7
9
7
.
5
0
8
.
5
1
7
.
5
3
A
v
g
C
o
s
t
D
e
p
2
.
3
6
1
.
9
8
1
.
6
8
2
.
1
9
3
.
1
6
2
.
2
7
A
v
g
W
a
g
e
s
4
7
.
8
6
4
9
.
8
2
5
1
.
4
8
5
3
.
6
8
5
6
.
9
3
5
1
.
9
6
C
O
R
a
t
e
3
.
5
9
3
.
5
1
2
.
5
6
2
.
2
1
1
.
7
2
2
.
5
2
N
P
R
a
t
e
8
.
2
6
7
.
6
7
6
.
1
9
5
.
7
5
5
.
5
3
6
.
6
6
H
H
I
6
.
3
7
8
.
5
4
7
.
1
1
7
.
0
9
7
.
2
1
7
.
2
7
C
a
p
i
t
a
l
1
0
.
9
9
1
1
.
0
2
1
1
.
2
3
1
1
.
8
2
1
2
.
3
7
1
1
.
4
9
L
i
q
u
i
d
i
t
y
1
6
.
0
3
1
6
.
2
5
1
4
.
1
2
1
5
.
4
5
1
6
.
8
1
1
5
.
7
1
N
o
t
e
s
:
T
h
i
s
t
a
b
l
e
p
r
e
s
e
n
t
s
s
u
m
m
a
r
y
s
t
a
t
i
s
t
i
c
s
f
o
r
b
a
n
k
s
t
h
a
t
r
e
p
o
r
t
e
d
a
v
a
l
u
e
f
o
r
m
a
r
k
e
t
i
n
g
e
x
p
e
n
s
e
s
i
n
e
a
c
h
q
u
a
r
t
e
r
o
f
e
a
c
h
r
e
s
p
e
c
t
i
v
e
y
e
a
r
.
T
h
e
v
a
r
i
a
b
l
e
s
a
r
e
a
v
e
r
a
g
e
s
p
e
r
b
a
n
k
i
n
e
a
c
h
y
e
a
r
.
M
k
t
i
s
t
h
e
r
e
p
o
r
t
e
d
a
d
v
e
r
t
i
s
i
n
g
,
p
r
o
m
o
t
i
o
n
a
l
p
u
b
l
i
c
r
e
l
a
t
i
o
n
s
,
m
a
r
k
e
t
i
n
g
a
n
d
b
u
s
i
n
e
s
s
d
e
v
e
l
o
p
m
e
n
t
e
x
p
e
n
s
e
s
f
o
r
e
a
c
h
y
e
a
r
.
S
a
l
B
e
n
E
x
p
i
s
t
h
e
t
o
t
a
l
s
a
l
a
r
y
a
n
d
b
e
n
e
?
t
e
x
p
e
n
s
e
s
f
o
r
e
a
c
h
y
e
a
r
.
N
u
m
b
E
m
p
i
s
t
h
e
a
v
e
r
a
g
e
n
u
m
b
e
r
o
f
e
m
p
l
o
y
e
e
s
o
v
e
r
t
h
e
f
o
u
r
q
u
a
r
t
e
r
s
o
f
e
a
c
h
y
e
a
r
.
B
r
a
n
c
h
e
s
i
s
t
h
e
n
u
m
b
e
r
o
f
b
r
a
n
c
h
e
s
e
a
c
h
?
r
m
h
a
s
o
n
a
v
e
r
a
g
e
,
d
u
r
i
n
g
e
a
c
h
r
e
s
p
e
c
t
i
v
e
y
e
a
r
.
T
A
i
s
t
h
e
a
v
e
r
a
g
e
t
o
t
a
l
a
s
s
e
t
s
o
v
e
r
t
h
e
f
o
u
r
q
u
a
r
t
e
r
s
o
f
e
a
c
h
y
e
a
r
.
I
n
t
I
n
c
(
N
o
n
I
n
t
I
n
c
)
i
s
t
o
t
a
l
i
n
t
e
r
e
s
t
(
n
o
n
i
n
t
e
r
e
s
t
)
i
n
c
o
m
e
f
o
r
e
a
c
h
y
e
a
r
.
S
T
A
s
s
e
t
s
i
s
s
u
m
o
f
f
e
d
f
u
n
d
s
s
o
l
d
,
r
e
v
e
r
s
e
-
R
P
s
,
a
n
d
s
e
c
u
r
i
t
i
e
s
w
i
t
h
a
m
a
t
u
r
i
t
y
o
f
l
e
s
s
t
h
a
n
1
y
e
a
r
.
R
e
a
l
E
s
t
a
t
e
I
n
c
,
I
n
d
i
v
i
d
u
a
l
I
n
c
a
n
d
C
o
m
I
n
d
I
n
c
i
s
t
h
e
t
o
t
a
l
i
n
t
e
r
e
s
t
i
n
c
o
m
e
f
o
r
e
a
c
h
y
e
a
r
f
r
o
m
r
e
a
l
e
s
t
a
t
e
,
i
n
d
i
v
i
d
u
a
l
,
a
n
d
c
o
m
m
e
r
c
i
a
l
a
n
d
i
n
d
u
s
t
r
i
a
l
p
r
o
d
u
c
t
s
l
o
a
n
s
,
r
e
s
p
e
c
t
i
v
e
l
y
.
T
o
t
R
e
v
i
s
t
h
e
s
u
m
o
f
t
h
e
r
e
p
o
r
t
e
d
i
n
t
e
r
e
s
t
a
n
d
n
o
n
i
n
t
e
r
e
s
t
i
n
c
o
m
e
f
o
r
e
a
c
h
y
e
a
r
.
T
o
t
D
e
p
(
D
e
m
D
e
p
)
i
s
t
h
e
a
v
e
r
a
g
e
o
f
t
o
t
a
l
(
d
e
m
a
n
d
)
d
e
p
o
s
i
t
s
o
v
e
r
t
h
e
f
o
u
r
q
u
a
r
t
e
r
s
o
f
e
a
c
h
y
e
a
r
.
I
n
t
D
e
p
i
s
t
h
e
t
o
t
a
l
a
m
o
u
n
t
o
f
i
n
t
e
r
e
s
t
o
n
d
e
p
o
s
i
t
s
f
o
r
e
a
c
h
y
e
a
r
.
M
k
t
S
h
r
e
D
e
p
i
s
t
h
e
w
e
i
g
h
t
e
d
a
v
e
r
a
g
e
d
o
f
t
h
e
p
e
r
c
e
n
t
a
g
e
s
o
f
t
o
t
a
l
m
a
r
k
e
t
d
e
p
o
s
i
t
s
,
w
h
e
r
e
a
m
a
r
k
e
t
i
s
d
e
?
n
e
d
a
s
a
s
t
a
t
e
a
n
d
t
h
e
w
e
i
g
h
t
s
a
r
e
d
e
t
e
r
m
i
n
e
d
b
y
t
h
e
p
e
r
c
e
n
t
a
g
e
o
f
t
h
e
b
a
n
k
s
t
o
t
a
l
a
n
n
u
a
l
d
e
p
o
s
i
t
s
i
n
e
a
c
h
s
t
a
t
e
f
o
r
e
a
c
h
y
e
a
r
.
I
n
t
E
x
p
(
N
o
n
I
n
t
E
x
p
)
i
s
t
h
e
s
u
m
o
f
i
n
t
e
r
e
s
t
(
n
o
n
i
n
t
e
r
e
s
t
)
e
x
p
e
n
s
e
f
o
r
e
a
c
h
y
e
a
r
.
T
o
t
E
x
p
i
s
t
h
e
t
o
t
a
l
i
n
t
e
r
e
s
t
a
n
d
n
o
n
i
n
t
e
r
e
s
t
e
x
p
e
n
s
e
f
o
r
e
a
c
h
y
e
a
r
.
R
e
a
l
E
s
t
a
t
e
L
o
a
n
s
,
I
n
d
i
v
i
d
u
a
l
L
o
a
n
s
,
a
n
d
C
o
m
I
n
d
L
o
a
n
s
i
s
t
h
e
a
v
e
r
a
g
e
o
f
r
e
a
l
e
s
t
a
t
e
,
i
n
d
i
v
i
d
u
a
l
,
a
n
d
c
o
m
m
e
r
c
i
a
l
a
n
d
i
n
d
u
s
t
r
i
a
l
l
o
a
n
s
,
r
e
s
p
e
c
t
i
v
e
l
y
,
o
v
e
r
t
h
e
f
o
u
r
q
u
a
r
t
e
r
s
o
f
e
a
c
h
y
e
a
r
.
L
o
a
n
C
h
a
r
g
e
O
f
f
s
i
s
t
h
e
t
o
t
a
l
a
m
o
u
n
t
o
f
l
o
a
n
c
h
a
r
g
e
o
f
f
s
f
o
r
t
h
e
y
e
a
r
.
N
o
n
P
f
m
g
L
n
s
i
s
t
h
e
t
o
t
a
l
a
m
o
u
n
t
o
f
n
o
n
-
p
e
r
f
o
r
m
i
n
g
l
o
a
n
s
f
o
r
t
h
e
y
e
a
r
.
L
o
a
n
R
e
c
i
s
t
h
e
a
m
o
u
n
t
o
f
l
o
a
n
r
e
c
o
v
e
r
i
e
s
f
o
r
t
h
e
y
e
a
r
.
T
o
t
L
o
a
n
s
i
s
t
h
e
a
v
e
r
a
g
e
o
f
t
o
t
a
l
l
o
a
n
s
o
v
e
r
t
h
e
f
o
u
r
q
u
a
r
t
e
r
s
o
f
e
a
c
h
y
e
a
r
.
R
a
t
e
R
E
,
R
a
t
e
I
n
d
,
a
n
d
R
a
t
e
C
o
m
I
n
d
i
s
t
h
e
r
a
t
e
o
f
r
e
t
u
r
n
o
n
r
e
a
l
e
s
t
a
t
e
,
i
n
d
i
v
i
d
u
a
l
,
a
n
d
c
o
m
m
e
r
c
i
a
l
a
n
d
i
n
d
u
s
t
r
i
a
l
l
o
a
n
s
,
r
e
s
p
e
c
t
i
v
e
l
y
,
f
o
r
e
a
c
h
y
e
a
r
.
R
a
t
e
s
a
r
e
c
a
l
c
u
l
a
t
e
d
a
s
(
R
e
a
l
E
s
t
a
t
e
I
n
c
/
R
e
a
l
E
s
t
a
t
e
L
o
a
n
s
)
*
1
0
0
,
(
I
n
d
i
v
i
d
u
a
l
I
n
c
/
I
n
d
i
v
i
d
u
a
l
L
o
a
n
s
)
*
1
0
0
,
a
n
d
(
C
o
m
I
n
d
I
n
c
/
C
o
m
I
n
d
L
o
a
n
s
)
*
1
0
0
,
r
e
s
p
e
c
t
i
v
e
l
y
.
A
v
g
C
o
s
t
D
e
p
i
s
t
h
e
a
v
e
r
a
g
e
c
o
s
t
o
f
d
e
p
o
s
i
t
s
f
o
r
e
a
c
h
y
e
a
r
,
c
a
l
c
u
l
a
t
e
d
a
s
(
I
n
t
D
e
p
/
T
o
t
D
e
p
)
*
1
0
0
.
A
v
g
W
a
g
e
s
i
s
t
h
e
a
v
e
r
a
g
e
w
a
g
e
r
a
t
e
f
o
r
e
a
c
h
y
e
a
r
,
c
a
l
c
u
l
a
t
e
d
a
s
(
S
a
l
B
e
n
E
x
p
/
N
u
m
b
E
m
p
)
.
C
O
R
a
t
e
i
s
c
a
l
c
u
l
a
t
e
d
a
s
(
L
o
a
n
C
h
a
r
g
e
O
f
f
s
/
T
o
t
L
o
a
n
s
)
*
1
,
0
0
0
a
n
d
N
P
R
a
t
e
i
s
c
a
l
c
u
l
a
t
e
d
a
s
(
N
o
n
P
f
m
g
L
n
s
/
T
o
t
L
o
a
n
s
)
*
1
,
0
0
0
.
H
H
I
i
s
t
h
e
H
e
r
?
n
d
a
h
l
-
H
i
r
s
c
h
m
a
n
I
n
d
e
x
,
c
o
m
p
u
t
e
d
a
s
a
w
e
i
g
h
t
e
d
a
v
e
r
a
g
e
o
f
e
a
c
h
s
t
a
t
e
’
s
H
H
I
w
h
e
r
e
a
b
a
n
k
t
a
k
e
s
d
e
p
o
s
i
t
s
,
w
h
e
r
e
t
h
e
w
e
i
g
h
t
s
a
r
e
c
a
l
c
u
l
a
t
e
d
a
s
t
h
e
p
e
r
c
e
n
t
a
g
e
o
f
t
o
t
a
l
d
e
p
o
s
i
t
s
t
a
k
e
n
i
n
e
a
c
h
r
e
s
p
e
c
t
i
v
e
s
t
a
t
e
,
m
u
l
t
i
p
l
i
e
d
b
y
1
0
0
.
C
a
p
i
t
a
l
i
s
c
a
l
c
u
l
a
t
e
d
a
s
(
T
E
/
T
A
)
*
1
0
0
a
n
d
L
i
q
u
i
d
i
t
y
i
s
c
a
l
c
u
l
a
t
e
d
a
s
(
S
T
A
s
s
e
t
s
/
T
A
)
*
1
0
0
Table I.
Bank marketing
investments
331
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
,100M 100M , .1B 1B , .10 B .10B
Panel A: 2002
n 2,516 1,927 176 37
TA 49,759 268,080 2,702,892 51,690,441
Mkt 52.47 320.82 4,337.31 157,497.00
MktMin 1 2 187 6,426
MktMax 7,648 51,661 89,400 1,376,000
MkttoTA 0.13 0.13 0.15 0.53
MkttoRev 2.06 1.40 1.64 3.03
MkttoExp 2.64 3.08 3.88 6.81
Branches 1.90 6.18 31.63 194.43
BranchesMax 1 1 1 1
BranchesMin 10 67 133 1,069
Panel A: 2003
n 2,416 2,136 199 43
TA 51,324 274,471 2,631,888 48,838,607
Mkt 47.23 318.94 3,398 137,107
MktMin 1 1 212 4230
MktMax 5,710 32,563 45,300 1,541,000
MkttoTA 0.10 0.11 0.12 0.38
MkttoRev 2.34 1.50 1.65 2.81
MkttoExp 2.57 3.05 3.59 6.51
Branches 1.92 6.05 29.50 173.63
BranchesMax 1 1 1 1
BranchesMin 11 37 131 990
Panel A: 2004
n 2,390 2,303 226 49
TA 51,779 275,552 2,544,613 51,510,204
Mkt 46.89 296.56 3,639.05 136,864.86
MktMin 1 1 118 1
MktMax 1,188 20,764 91,283 2,186,000
MkttoTA 0.09 0.11 0.13 136,865
MkttoRev 2.30 1.58 1.83 2.41
MkttoExp 2.62 3.12 3.87 5.77
Branches 1.90 6.02 27.90 208.51
BranchesMax 1 1 1 1
BranchesMin 9 160 161 1,325
Panel A: 2005
n 2,268 2,343 243 46
TA 51,938 284,134 2,593,334 59,309,785
Mkt 51.77 329.40 3,495.12 149,018.26
MktMin 1 1 1 1
MktMax 3,272 40,037 103,993 2,003,000
MkttoTA 0.10 0.11 0.13 0.25
MkttoRev 2.48 1.53 1.72 2.07
MkttoExp 2.77 3.23 3.97 5.12
Branches 1.91 5.98 29.55 221.41
BranchesMax 1 1 1 1
BranchesMin 9 194 172 1,367
Panel A: 2006
n 2,073 2,321 243 36
(continued)
Table II.
Marketing expenditures
and branches summary
statistics disaggregated
by bank size
JFEP
2,4
332
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Marketing as a proportion of total non-interest expense, however, increases
monotonically with bank size. Marketing accounts for about 6 percent of expenses
at the largest banks, vs 3 percent or less at the smaller banks. The largest organizations
are more likely to hire global advertising and public relations ?rms, spend in the
national media on brand building and/or to pursue expensive sponsorships of sports
teams, which presumably accounts for the substantial differences we observe in these
data.
Academics point to multiple channels through which marketing investments can
affect ?rm ?nancial performances. Resources are expended to develop product,
promotion, and channel delivery strategies. The strategies underpin tactical marketing
efforts such advertising campaigns, service quality improvement programs, branding
initiatives, and loyalty programs designed to retain existing customers and attract new
ones. From a ?nance perspective, resource expenditures in these areas can serve to
reduce information asymmetries. And since “brand’ is frequently linked to ?rm
reputation, marketing investments might reduce moral hazard problems. Indeed, the
?rm’s brand, its market networks and the intellectual capital behind them represent
intangible assets that serve to generate cash ?ows in the same sense as tangible assets.
“Brand equity” is core concept in marketing science and Keller (1993) notes that its
value arises from the incremental discounted cash ?ow from the sale of additional
,100M 100M , .1B 1B , .10 B .10B
TA 51,767 290,620 2,510,467 77,556,041
Mkt 70.04 376.87 3,118.70 196,639.83
MktMin 1 1 80 7,000
MktMax 30,946 57,057 91,272 1,654,000
MkttoTA 0.13 0.12 0.12 0.25
MkttoRev 2.09 1.43 1.49 2.06
MkttoExp 2.86 3.30 3.84 5.20
Branches 1.87 5.93 29.42 208.97
BranchesMin 1 1 1 1
BranchesMax 10 39 181 1,397
Panel A: total
n 11,663 11,030 1,087 211
TA 51,280 279,085 2,589,476 57,141,583
Mkt 53.23 329.01 3,561.58 153,380
MktMin 1 1 1 1
MktMax 30,946 57,057 103,993 2,186,000
MkttoTA 0.11 0.12 0.13 0.33
MkttoRev 2.25 1.49 1.67 2.46
MkttoExp 2.68 3.16 3.84 5.86
Branches 1.90 6.03 29.51 201.82
BranchesMin 1 1 1 1
BranchesMax 11 194 181 1,397
Notes: This table presents summary statistics for marketing expenditures, segmented by ?rm size (in
TA. Mkt and Branches are as de?ned in Table I. MktMin (MktMax) is the minimum and maximum
value of Mkt in each size segment and BranchesMin (BranchesMax) represent the same for values of
Branches. MkttoTa, MkttoRev, and MkttoExp are ratios of the annual marketing expense to TA, total
revenue, and total non-interest expenses, respectively. Each ratio is multiplied by 100 Table II.
Bank marketing
investments
333
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
products or services occasioned by the brand promise or signal of quality and
reliability[11].
We focus initially on the prospective linkage between marketing expenditures and
bank pro?ts and ground our model speci?cation in the theory of the pro?t function.
The theory of pro?t function was developed by McFadden (1978) and has been applied
most often to the analysis of relative ef?ciency at ?nancial (Akhavein et al., 1997a, b)
and agricultural ?rms. The pro?t function expresses the maximized pro?t for a ?rm in
a competitive situation as a function of the prices of outputs and variable factor inputs
and the quantities of the ?xed factors of production.
The theory shows that pro?ts are increasing in output prices, decreasing in input
prices, and increasing in the ?xed factors. Most studies of the relationship between
marketing expenditures and various measures of pro?tability in the literature involve
ad hoc speci?cations and almost all are plagued by endogeneity problems. Our
strategy is to treat a marketing investment as a quasi-?xed factor of production. By
de?nition, a quasi-?xed factor is one that is partly ?xed and partly variable. Oi (1962)
emphasized that labor should be viewed as a quasi-?xed factor of production, since a
?rm commonly incurs hiring and training costs that make its individual workers more
valuable to that ?rm relative to others. Klein et al. (1978) and Williamson (1979)
generalized this argument, noting that any two parties who have incurred investment
costs that are “transaction speci?c” will be better off trading with each other than with
other parties[12]. Expenditures designed to develop a “brand promise” that mitigates
search and other information costs associated with quality assessments are explicitly
designed to produce “relationship speci?c investments” between sellers and buyers.
Accordingly, they represent quasi-?xed factors. This hypothesis, from the point of
view of the theory of the pro?t function, would be con?rmed by a positive coef?cient
on the advertising and promotion expenditure variable.
We also examine whether there are increasing or decreasing returns to marketing
by creating some interaction terms that allow for non-linearities in the relationship
between advertising and promotion and bank pro?ts[13]. In particular, we create
separate dummy variables for each of the three larger bank size classi?cations and
interact these dummies with advertising and promotion expenditures. We offer no
hypothesis about prospective non-linearities, but simply allow the data to speak to the
issue. We do the same for the branching variables.
The pro?t function is derived from microeconomic theory based on the assumption
of perfect competition in a non-industry speci?c context. There may be variables
relevant to predictions of pro?t and market share that are speci?c to ?rms in the banking
industry, however. Likewise, banking markets may be imperfectly competitive.
Consequently, we augment the original speci?cation with several additional variables.
These include the logs of the ratio of nonperforming to total loans, the Tier 1 leverage
ratio (Tier 1 capital/assets), a liquidity ratio, and the Her?ndahl-Hirschman index (HHI)
as a measure for the degree of market competition.
4. Estimates of the pro?t function model
We initially estimate a bank pro?t function. Following prior literature, we use the
average prices of various types of loans as proxies for output prices, the average price
of deposits and the average wage rate as input price proxies, and the number of
branches as a ?xed or quasi-?xed factor of production[14]. In addition, we also include
JFEP
2,4
334
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
marketing expenditures as a quasi-?xed factor. Bank pro?ts thus should increase with
loan rates, decrease with deposit rates and average wages, and increase with the
number of branches and with advertising and promotion expenditures.
The argument that the wage coef?cient should be negatively signed, however,
is based on a view of labor as a variable factor of production. As noted above, Oi (1962)
suggests that labor is better represented as a “quasi-?xed” factor of production. Since
labor may not be a variable factor of production for banks, the anticipated sign on the
coef?cient of this variable is ambiguous.
The primary variables of interest in the model are advertising and
promotion expenditures and the number of branches, both of which we assume
represent quasi-?xed investments in a pro?t-function framework. The very purpose of
marketing spending in the formof advertising and promotion is to “marry” the customer
to the ?rm, resulting in a relationship-speci?c investment. And since branches represent
one means to deliver banking products, expenditures on the branch network can
likewise be viewedas a marketinginvestment. Consequently, buildinga newbranchalso
represents a transaction-speci?c investment, much like we argued for advertising
and promotion[15]. The fact that many bank customers cite location convenience as
a primary factor in opening and maintaining an account is consistent with this
interpretation. Berger et al. (1997) ?nd there are about twice the number of branches
relative to the number that would minimize costs. They posit that, fromthe standpoint of
pro?tability, banks feel a need to attract customers and opening newbranches facilitates
this process. Hence, by the theory of the pro?t function, the coef?cients of the marketing
and branching variables should be positive. All of our variables are measured annually
and, as is common in estimating pro?t functions, we take the log of all variables in our
estimations. Consequently, the coef?cients can be interpreted as pro?t elasticities.
We examine the prospect of differential returns to scale for both advertising and
promotion expenditures and branching activity by interacting these variables with
dummy variables that re?ect bank size.
Introducing notation, the model we initially estimate is:
Dep ¼ b
0
þb
Firm
þ b
1
LnMkt þ b
2
Mkt2 þ b
3
Mkt3 þb
4
Mkt4
þ b
5
LnBranches þ b
6
Branches2 þ b
7
Branches3 þb
8
Branches4
þ b
9
LnRateRE þ b
10
LnRateInd þ b
11
LnRateCI þb
12
LnAvgCostDep
þ b
13
LnAvgWages þ 1
ð1Þ
where Dep is LnRevMExp, de?ned as the natural log of the bank revenues less
expenses (excluding marketing and occupancy costs). LnMkt is the natural logarithm
of reported advertising and promotion expense for each year. LnBranches is the
natural logarithm of the number of branches the ?rm has on June 30 of each year[16].
In addition, we segment the sample based upon size, measured by total assets (TA).
The four designations again correspond to those used in the FDIC’s Quarterly Banking
Pro?le. The categories are assigned variables, TA1, TA2, TA3, and TA4, respectively.
Thus, Mkt2, Mkt3, and Mkt4 are interaction variables calculated as the product of the
natural log of marketing expenditures and each respective size dummy variable.
Likewise, Branches2, Branches3, and Branches4 are interaction variables calculated as
the product of the natural log of the number of branches and each respective size
dummy variable.
Bank marketing
investments
335
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
RateRE, RateInd, and RateCI are the average rates on real estate, individual, and
commercial and industrial loans, respectively, for each year. AvgCostDep is the average
cost of deposits for each year, calculated as interest from deposits (IntDep) divided by
total deposits (TotDep), and multiplied by 100. Finally, AvgWages is the average wage
rate for each year, calculated as total salary and bene?t expenses for each year
(SalBenExp) divided by the average number of employees throughout each year
(NumbEmp).
In a second stage estimation, we include more bank speci?c variables in the model,
primarily as control variables. We posit negative signs for the coef?cients of the
non-performing loan (NPRate) and liquidity ratios (Liquidity), which is measured as the
sum of fed funds sold, reverse-RPs, and securities with a maturity of less than 1 year as
a percentage of assets. Hays et al. (2009) include similar variables in their study[17].
Pro?ts should decline, ceteris paribus, as non-performing loans increase and as the
balance sheet becomes more liquid[18]. We posit no sign on the capital ratio (Capital )
coef?cient and a positive sign on the HHI. Berger (1995b) extensively examines the
relationship between capital and pro?ts, illustrating the importance of controlling for
the potential in?uence. Intuitively, a higher level of capital reduces ?rm risk, which
should result in a lower required return. However, he ?nds that during the 1980s the
relationship between capital and ROE to be positive. Hutchison and Cox (2007) ?nd a
positive relationship between equity capital and return on assets.
HHI is the sum of squared market shares for all participants in a given market[19].
The HHI is a widely accepted measure of market concentration and has been used
extensively in the literature. Markets become more concentrated with the HHI and
the prospect of pricing power in a concentrated market could increase revenue.
However, banks are multi-product ?rms and our HHI measure is deposit driven. The
HHI-revenue relationship may be attenuated if deposit market concentration is not
highly correlated with the degree of concentration in other bank product markets.
Descriptive statistics on all these variables are presented in Table I.
We initially estimate a “base” pro?t function model for a ?xed-effects panel with an
AR(1) disturbance to control for serial autocorrelation over our sample period (Table III,
“Base” Column)[20]. The results show a positive and signi?cant impact of advertising
and promotion expenditures on pro?tability. Since our coef?cient estimates are
elasticities, a 10 percent increase in marketing expenditures produces almost a 2
percent increase in pro?ts. We also ?nd increasing returns to advertising with the scale
of the organization. Each of the interaction coef?cients is positive and signi?cant and
the returns monotonically increase for banks up to $10 billion in assets. We perform
F-tests to determine if the differences in the interaction coef?cients are signi?cant and
?nd there are higher marginal returns as bank size grows, save for the case where size
moves above $10 billion.
Bank pro?ts also increase with the number of branches and the estimated elasticity
is similar to that for advertising and promotion. But the evidence for increasing returns
with branching is more limited in this base model estimation. The marginal bene?t
increases as institutions grow from above $100 million to $1 billion, but the coef?cients
for the larger size groups are either negative or insigni?cant.
The remaining coef?cients in the pro?t function model are statistically signi?cant
in all cases. Pro?ts increase with loan rates and decline with deposit costs. The
coef?cient of the wage variable is positive, implying that labor acts more as a quasi-?xed
JFEP
2,4
336
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
LnRevMExp
Base Extended Selection
Coef t-stat Coef t-stat Coef t-stat
Intercept 2.93 71.58 1.54 38.96 4.09 26.16
LnMkt 0.18 19.46 0.17 17.57 0.30 23.24
Mkt2 0.02 3.73 0.03 4.03 0.17 27.14
Mkt3 0.09 6.68 0.07 4.88 0.36 19.26
Mkt4 0.07 2.50 0.01 0.21 0.37 5.83
LnBranches 0.20 8.39 0.17 6.94 0.35 17.80
Branches2 0.05 2.07 0.02 0.81 20.04 21.94
Branches3 20.08 22.11 20.04 21.15 20.26 25.93
Branches4 20.00 20.03 0.15 1.24 20.09 20.61
LnRateRE 0.46 17.12 0.51 17.12 0.25 5.52
LnRateInd 0.11 7.77 0.09 5.53 0.08 4.00
LnRateCI 0.20 13.52 0.21 13.45 0.02 0.86
LnAvgCostDep 20.17 214.05 20.17 213.43 0.03 1.29
LnAvgWages 0.61 29.83 0.65 27.96 0.18 5.51
LnNPRate 20.00 21.77
LnCapital 20.56 218.70
LnLiquidity 20.04 25.02
LnHHI 20.02 23.10
Difference tests
Mkt2 vs Mkt3 0.0000 0.0018 0.0000
Mkt2 vs Mkt4 0.0837 0.6842 0.0029
Mkt3 vs Mkt4 0.5181 0.1508 0.9187
Br2 vs Br3 0.0001 0.0527 0.0000
Br2 vs Br4 0.3851 0.2822 0.7556
Br3 vs Br4 0.2174 0.0948 0.2795
n 16,007 13,698 22,248
R
2
0.7787 0.7130 –
p . 0.0000 0.0000 0.0000
Notes: This table presents results from the following model:
LnRevMExp ¼ b
0
þ b
Firm
þ b
1
LnMkt þ b
2
Mkt2 þ b
3
Mkt3 þ b
4
Mkt4 þ b
5
LnBranches
þ b
6
Branches2 þ b
7
Branches3 þ b
8
Branches4 þ b
9
LnRateRE þ b
10
LnRateInd
þ b
11
LnRateCI þ b
12
LnAvgCostDep þ b
13
LnAvgWages þ b
14
NPRate
þ b
15
Capital þ b
16
Liquidity þ b
17
LnHHI þ 1
where LnRevMExp is the natural log of the revenues less expenses (excluding marketing and
occupancy costs). We segment the sample based upon size (measured by TA). The four designations
are (1) those ?rms with TA ,$100M, (2) those ?rms with TA between $100M and $1B, (3) those ?rms
with TA between $1B and $10B, and (4) those ?rms with TA . $10B. The categories are assigned
variables, TA1, TA2, TA3, and TA4, respectively. Thus, Mkt2, Mkt3, and Mkt4 are interaction
variables calculated as the product of the natural log of marketing expenditures and each respective
size dummy variables. Likewise, Branches2, Branches3, and Branches4 are interaction variables
calculated as the product of the natural log of the number of branches and each respective size dummy
variable. All other variables are the natural log of each respective variable, all as previously de?ned in
Table I. We implement a ?xed effects model with AR(1) disturbances to control for serial
autocorrelation in the base and extended models. For the model controlling for selection bias, we
implement a maximum likelihood Heckman method, with standard errors clustered by bank.
p . refers to p . F for base and extended models, and p . x
2
for the selection model
Table III.
Total sample results:
pro?t models
Bank marketing
investments
337
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
factor in producing banking services rather than a traditional input. The evidence
reveals that “marketing matters” and that the returns to advertising and promotion
increase with the size of an organization up to an asset level of $10 billion.
Are these ?ndings robust to the inclusion of additional model variables that recognize
that our ?rms are banks andthat also allowfor less thanperfect competition? The results
in the column labeled “Extended” are the coef?cient estimates when bank capital,
liquidity, and non-performing loan ratios enter the model, along with the HHI measure
for the degree of market concentration. As predicted, the impact of increased liquidity
and non-performing loans on pro?ts is negative, but only the former coef?cient is
signi?cant. Bank pro?ts decrease with capital and decline as market concentration
increases. The latter ?nding runs contrary to the expected effect, but could re?ect the
prospect that concentration in deposit market shares fails to capture market pricing
power across the range of banking products and services. Most importantly, the
inclusion of these variables has very little impact on our estimates of returns to
advertising and promotion and to branching. Both types of investments continue to
yield increased pro?ts and advertising spending continues to show increasing returns
with scale up to an asset size of $10 billion. The scale effects again are less relevant for
branching.
Since only some banks are required to report their marketing expenditures, it is
possible our results may be contaminated by selection bias. We examine this possibility
by re-estimating the model employing the Heckman (1979) technique. Speci?cally, we
cluster the standard errors by ?rm and use Heckman maximum likelihood. The results
for the impact of one of the main variables of interest-advertising and promotion – is
larger in size, as well as more signi?cant, than in the prior two model estimations.
Similarly, the observed increases in returns to scale are much larger. A 10 percent
increase in marketing spending for banks with assets over $1 billion yields about a 6.6
percent gain in pro?ts, other things equal. This is more than twice as large as the
estimated return to advertising at smaller institutions.
The results for branching in the selection model show no evidence of increasing
returns as all three interaction coef?cients have negative signs. One of the coef?cients is
signi?cant and another marginally so. The results are more consistent with negative
marginal returns to spending on branches, at least for banks up to $10 billion in size.
The coef?cients of the remaining variables are generally smaller and less signi?cant
than in prior estimations. Thus, taking account of prospective selection bias does
not affect our inferences on the main variables of interest, but it does yield somewhat
different results with respect to the impact of scale. In the case of advertising and
promotion spending, the estimated returns to scale are larger andmore signi?cant. In the
case of branching, the returns to scale estimates are somewhat weaker and in one case
signi?cantly negative.
5. Estimates of the market share model
Much of the marketing literature focuses on how marketing affects the evolution of
market share. And, as noted above, there is a body of banking-focused research on this
market share, though with a somewhat different focus than ours. Economic theory
suggests that a ?rm’s capacity to in?uence prices in its market depends on its market
power, which should be positively correlated with market share. Since increases in
output prices are associated with higher pro?ts, ceteris paribus, market share and ?rm
JFEP
2,4
338
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
pro?ts should be positively related. Consequently, we next estimate a model that
examines the relationship between bank marketing expenditures and market share.
While the speci?cation of a bank pro?t model can be theory-motivated under a perfect
competition assumption, the underlying logic of a relationship between market share,
marketing, and pro?tability must assume less than perfect competition. We are unaware
of any theory that would motivate a fairly precise market-share model speci?cation, so
we posit an admittedly ad hoc model that relates market share to the same variables that
appear in the pro?t function model.
One rationale for such an investigation is the strong interest among practitioners, and
across textbooks and the practical literature, on the effects of marketing investments on
market share. If marketing works in the manner taught in most marketing courses, there
should be a positive relationship between marketing expenditures, either as advertising
and promotion or through opening additional branches, and measured market share.
Another motivating factor is the prospect that our pro?t function results in some way
re?ect the effects of marketing on market share. To the extent we ?nd somewhat
consistent results for the market share and pro?t models, we have at least implicit
evidence of a link between market share and pro?tability.
Since banks are multi-service ?rms operating across widely varying geographical
areas, measuring market share is not a straightforward proposition. We collect market
share data for each sample bank in each state where they have a branch. Thus, our
“market” classi?cation is on a state level. Many banks have market shares in multiple
states. In this case, we compute a weighted average of each market share based upon
the percentage of the banks total annual deposits in each state. The natural log of this
variable (LnMktShreDep) serves as the dependent variable in the regressions.
We estimate the model using the same econometric techniques as for the pro?t
function model. The results are reported in Table IV. The results are remarkably
similar to the pro?t function model results. Advertising and promotion expenditures
are positively and signi?cantly related to deposit market share and there is again
signi?cant evidence of increasing returns to scale in the sense that the impact of
advertising on market share increases with assets size, in this case in monotonic
fashion. But while building or buying more branches results in signi?cant increases in
deposit market share, the ?ndings for the interaction dummies suggest there are
signi?cant decreasing returns to scale with respect to branching. The results suggest
that the observed results for bank pro?ts may be at least somewhat driven by the
impact of market share on pro?tability.
The majority of the coef?cients of the remaining variables in the model are not
statistically signi?cant. Where we have signi?cance, the market share of deposits tends
to increase with rates on real estate loans and to decline with average wages. When we
add the additional bank speci?c variables the results for the coef?cients of advertising
and promotion again remain robust and consistent with increasing returns to scale. In
the case of branching, the evidence for decreasing returns to scale becomes somewhat
weaker, as the in?uence of marketing expenditures on banks with assets of .$100
million and ,$1 billion is insigni?cantly different from that of the smallest banks[21].
We again take account of potential sample selection problems by estimating a
Heckman model. The estimated impacts of increased spending on advertising and
promotion and/or branching become much stronger in each case. The predicted returns
to scale for advertising are likewise stronger than in the prior models, but the result for
Bank marketing
investments
339
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
the effects of scale on branching are statistically insigni?cant across the size
groupings. In the Heckman, four of the ?ve remaining variables in the model are
signi?cant with plausible signs. The results for these alternative models that remain
most consistently robust across the different speci?cations are those re?ecting the
impact of adverting and promotion on market share.
LnMktShreDep
Base Extended Selection
Coef t-stat Coef t-stat Coef t-stat
Intercept 22.88 230.97 23.16 221.55 28.02 237.36
LnMkt 0.07 6.16 0.07 4.98 0.20 12.63
Mkt2 0.05 7.01 0.03 3.40 0.10 9.01
Mkt3 0.10 6.47 0.08 4.73 0.23 8.60
Mkt4 0.16 4.99 0.26 4.59 0.20 3.29
LnBranches 0.32 12.52 0.26 8.47 0.38 11.09
Branches2 20.06 22.22 0.01 0.29 0.05 1.30
Branches3 20.20 24.84 20.13 22.79 20.12 21.74
Branches4 20.30 24.41 20.52 23.24 20.01 20.07
LnRateRE 0.07 2.37 0.08 2.22 0.06 0.96
LnRateInd 0.00 0.09 20.02 21.02 0.07 2.03
LnRateCI 20.03 21.46 20.01 20.47 20.09 22.40
LnAvgCostDep 20.02 21.10 20.04 22.59 0.22 4.81
LnAvgWages 20.07 22.53 20.05 21.36 20.21 24.41
LnNPRate 0.00 0.92
LnCapital 20.15 23.66
LnLiquidity 0.00 0.22
Difference tests
Mkt2 vs Mkt3 0.0004 0.0012 0.0000
Mkt2 vs Mkt4 0.0003 0.0000 0.0889
Mkt3 vs Mkt4 0.0410 0.0011 0.6013
Br2 vs Br3 0.0002 0.0007 0.0073
Br2 vs Br4 0.0003 0.0009 0.6274
Br3 vs Br4 0.1197 0.0124 0.4345
n 16,178 13,743 22,561
R
2
0.4890 0.4512 –
p . 0.0000 0.0000 0.0000
Notes: This table presents results from the following model:
LnMktShreDep ¼ b
0
þ b
Firm
þ b
1
LnMkt þ b
2
Mkt2 þ b
3
Mkt3 þb
4
Mkt4 þ b
5
LnBranches
þb
6
Branches2 þ b
7
Branches3 þ b
8
Branches4 þb
9
LnRateRE
þb
10
LnRateInd þ b
11
LnRateCI þ b
12
LnAvgCostDep þ b
13
LnAvgWages
þb
14
NPRate þb
15
Capital þ b
16
Liquidity þb
17
LnHHI þ 1
where LnMktShreDep is the natural log of the bank’s weighted average percentage of deposits in each
bank’s market, where a market is de?ned as a state and the weights are determined by the percentage
of the banks total deposits in each market. All other variables are the natural log of each respective
variable, all as previously de?ned in Tables I and III. We implement a ?xed effects model with AR(1)
disturbances to control for serial autocorrelation in the base and extended models. For the model
controlling for selection bias, we implement a maximum likelihood Heckman method, with standard
errors clustered by bank. p . refers to p . F for base and extended models, and p . x
2
for the
selection model
Table IV.
Total sample results:
market share models
JFEP
2,4
340
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
6. Conclusion
Our paper ?lls a gap in the literature by developing quantitative estimates of the impact
of increased investments in advertising and promotion on bank pro?t performance and
market share generation. Using the pro?t function framework, the panel estimation
results show that enhanced marketing investments either in the form of brand-focused
advertising and promotion or additional spending on branch-based delivery systems
generates increased pro?ts. The results also showevidence of increasing returns to scale
in advertising and promotion expenses. There is no evidence of scale effects in the case of
branching, however. Rather, in some estimations, the results are more consisting with
decreasing returns to scale. The overall results are robust to including bank-speci?c
variables in the model, as well as to taking account of the variation in concentration
across markets and to the prospect of sample selection problems.
We also examine how these same factors affect variation in deposit market shares
across institutions and over time. The results are quite similar to the estimated results
for the pro?t function model. Investments in brand equity via increased advertising
and/or in expanding the branch network have favorable effects on market share. Once
again, advertising and promotion shows positive scale effects, whereas the results are
more consistent with decreasing returns to scale for branching. The results are
similarly robust to alternative speci?cations and to addressing sample selection
problems. Our overall conclusion is that for the representative banking institution,
“marketing pays.”
Notes
1. Marketing expenditures in the form of advertising and promotion can be viewed as
investments since they re?ect a given amount of current spending designed to produce cash
?ows over some, presumably long, future period. Kasanen (1993) notes, for instance, that:
“Any investment, especially in strategic projects such as new technology, brand name or
company image, may generate future investment opportunities.”
2. Srinivasan and Hanssens (2009) provide a thorough discussion of the existing literature
pertaining to the in?uences of marketing on ?rm value.
3. More speci?cally, estimations of the pro?t function have been used to study whether
banking is characterized by economies of scale (Mullineaux, 1978) and/or economies of scope
(Berger et al., 1993).
4. Investments in alternative delivery mechanisms such as ATM machines, internet banking,
mobile banking (via cell phones), and remote capture of deposits via specialized terminals
have grown rapidly in recent years, but unfortunately our data source contains no speci?c
information on expenditures on these mechanisms. Rather, they are embedded in the broad
category “other expenses.”
5. An additional complication is that the parent might itself spend resources on marketing, but
again such expenditures are not reported in any speci?c manner at the parent level.
6. There is a wealth of papers in the ?nance literature on the link between ?rm value and R&D
expenditures. For examples, see Zantout and Tsetsekos (1994), Szewczyk et al. (1996), and
Chan et al. (2001). Since marketing and R&Dcan both be viewed as investments in intangible
assets, the relatively heavy focus in ?nance on just one type of these expenditures seems
anomalous.
7. Announcing sponsorship by a celebrity endorser, a la Michael Jordan, produces a like result
(Mathur et al., 1997).
Bank marketing
investments
341
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
8. These studies use either market capitalization (typically relative to sales, book value or
replacement costs) or, less frequently, equity returns.
9. This data item was ?rst included on March 2001 call reports.
10. As a reference point, the average size of the banks in the largest size group is about 22 times
the average size of the banks in the $1-$10 billion range over our sample period.
11. Tybout and Carpenter (2000) indicate that Interbrand estimated the brand equity of Home
Depot at $84 billion in 1999, for example.
12. Williamson (1979) observes that site speci?city (nearby locations), asset speci?city (highly
specialized inputs), or human asset speci?city (learning by doing) can be the root of
relationship-speci?c investments.
13. Verma (1980) and Nguyen (1987) show the sales-advertising relationship is nonlinear, and
while sales do not necessarily result in pro?ts, the correlation between the two is typically
high.
14. Bank branches can be closed or sold to other institutions and hence are probably best treated
as quasi-?xed rather than ?xed factors of production. In either case, the hypothesized sign is
positive on this variable.
15. Alternatively, branches could be viewed as a component of “property and equipment”
investment, which is commonly treated as a ?xed cost in the short run, which again argues
for a positive coef?cient.
16. This variable was used by Mullineaux (1978) in estimating the bank pro?t function.
17. Speci?cally, Hays et al. (2009) include the liquidity ratio and the ratio of net charge offs to
loans, which is a different measure of asset risk than the non-performing loan rate.
In unreported analysis, we replace NPRate with CORate, which is total loan charge-offs
minus recoveries divided by total loans. The results are unchanged.
18. Liquid assets have lower returns than illiquid assets, provided the yield curve is upwards
sloping, as it was over our sample period.
19. We take each state as a potential market and use deposit shares to measure concentration.
Where banks operate in multiple states, we calculate the weighted average HHI for such
institutions, where the weights are the percentages of the institutions total deposits
generated in each state.
20. The data, like many panel data sets, is biased by issues of autocorrelation, as evidenced by
the Wooldridge (2002) test. To ensure the reported model has adequately addressed the issue
of autocorrelation, we also calculate the Baltagi and Wu (1999) least biased instrument
statistic for each regression. These values indicate that correlation is no longer a biasing
in?uence following the AR(1) control.
21. The reported results do not have HHI included due to the fact that the calculated HHI is a
weighted average, based upon their deposits in each state, just as market share. Thus, there
is a very high correlation between the two variables and the coef?cient on HHI is highly
signi?cant. However, in unreported results, the inclusion of HHI in the model does not
signi?cantly alter the primary results in any way.
References
Akhavein, J., Berger, A. and Humphrey, D. (1997a), “The effects of megamergers on ef?ciency
and prices: evidence from a bank pro?t function”, Review of Industrial Organization,
Vol. 12, pp. 95-139.
JFEP
2,4
342
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Akhavein, J., Swamy, P., Taubman, S. and Singamsetti, R. (1997b), “A general method of deriving
the inef?ciencies of banks from a pro?t function”, Journal of Productivity Analysis, Vol. 8,
pp. 71-93.
Balasubramanian, S., Mathur, I. and Thakur, R. (2005), “The impact of high-quality ?rm
achievements on shareholder value: focus on Malcolm Baldridge and J.D. Power and
Associates awards”, Journal of the Academy of Marketing Science, Vol. 33, pp. 413-22.
Baltagi, B. and Wu, P. (1999), “Unequally-spaced panel data regressions with AR(1)
Disturbances”, Econometric Theory, Vol. 15, pp. 814-23.
Ben Zion, Y. (1978), “The investment aspect of nonproduction expenditures: an empirical test”,
Journal of Economics and Business, Vol. 30, pp. 224-9.
Berger, A. (1995a), “The pro?t-structure relationship in banking-tests of market-power and
ef?cient structure hypotheses”, Journal of Money, Credit, and Banking, Vol. 27, pp. 404-31.
Berger, A. (1995b), “The relationship between capital and earnings in banking”, Journal of
Money, Credit, and Banking, Vol. 27, pp. 432-56.
Berger, A. and Hannan, T. (1989), “The price-concentration relationship in banking”, The Review
of Economics and Statistics, Vol. 71, pp. 291-9.
Berger, A., Hancock, D. and Humphrey, D. (1993), “Bank ef?ciency derived from the pro?t
function”, Journal of Banking & Finance, Vol. 17, pp. 317-47.
Berger, A., Leusner, J. and Mino, J. (1997), “The ef?ciency of bank branches”, Journal of Monetary
Economics, Vol. 40, pp. 141-62.
Chan, L., Lakonishok, J. and Sougiannis, T. (2001), “The stock market valuation of research and
development expenditures”, The Journal of Finance, Vol. 56, pp. 2431-56.
Chauvin, K. and Hirschey, M. (1993), “Advertising, R&D expenditures, and the market value of
the ?rm”, Financial Management, Vol. 22, pp. 128-40.
Conchar, M., Crask, M. and Zinkhan, G. (2005), “Market valuation models of the effect of
advertising and promotional spending: a review and meta-analysis”, Journal of the
Academy of Marketing Sciences, Vol. 33, pp. 445-60.
Cornwell, T., Pruitt, S. and Clark, J. (2005), “The relationship between major league sports’ of?cial
sponsorship announcements and the stock prices of sponsoring ?rms”, Journal of the
Academy of Marketing Science, Vol. 33, pp. 401-12.
DeYoung, R. and Ors, E. (2004), “Advertising and pricing at multiple output ?rms: evidence from
US thrift institutions”, Working Paper (2004-35), Federal Reserve Bank of Chicago,
Chicago, IL.
Demsetz, H. (1973), “Industry structure, market rivalry, and public policy”, Journal of Law &
Economics, Vol. 16, pp. 1-9.
Gleason, K., Mathur, L. and Mathur, I. (2000), “The interrelationship between culture, capital
structure, and performance: evidence from European retailers”, Journal of Business
Research, Vol. 50, pp. 185-91.
Hasan, I., Hunter, W. and Mathis, R. (2000), “Promotional expenditures, market competition, and
thrift behavior”, Journal of Business Research, Vol. 50, pp. 177-84.
Hays, F., de Lurgio, S. and Gilbert, A. (2009), “Ef?ciency ratios and community bank
performance”, Journal of Finance and Accountancy, Vol. 1.
Heckman, J. (1979), “Sample selection as a speci?cation Error”, Econometrica, Vol. 47, pp. 153-61.
Hutchison, D. and Cox, R. (2007), “The causal relationship between bank capital and
productivity”, Annals of Financial Economics, Vol. 3, pp. 40-54.
Bank marketing
investments
343
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Kasanen, E. (1993), “Creating value by spawning investment opportunities”, Financial
Management, Vol. 22, pp. 251-8.
Keller, K. (1993), “Conceptualizing, measuring, and managing customer-based brand equity”,
Journal of Marketing, Vol. 57, pp. 1-22.
Klein, B., Crawford, R. and Alchian, A. (1978), “Vertical integration, appropriable rents and the
competitive contracting process”, Journal of Law & Economics, Vol. 21, pp. 297-326.
Lane, V. and Jacobson, R. (1995), “Stock market reactions to brand extension announcements: the
effects of brand attitude and familiarity”, Journal of Marketing, Vol. 59, pp. 63-77.
McFadden, D. (1978), “Cost, revenue and pro?t functions: a cursory review”, in Fuss, M. and
McFadden, D. (Eds), Production Economics: A Dual Approach to Theory and Applications,
North Holland, Amsterdam.
Mathur, L., Mathur, I. and Rangan, N. (1997), “The wealth effects associated with a celebrity
endorser: the Michael Jordan phenomenon”, Journal of Advertising Research, Vol. 37,
pp. 67-73.
Mullineaux, D. (1978), “Economies of scale and organizational ef?ciency in banking:
a pro?t-function approach”, Journal of Finance, Vol. 33, pp. 259-80.
Nguyen, D. (1987), “Advertising, random sales response, and brand competition: some theoretical
and econometric implications”, Journal of Business, Vol. 60, pp. 159-279.
Oi, W. (1962), “Labor as a quasi-?xed factor”, Journal of Political Economy, Vol. 70, pp. 538-55.
Peltzman, S. (1977), “The gains and losses from industrial concentration”, Journal of Law &
Economics, Vol. 20, pp. 229-63.
Rust, R., Ambler, T., Carpenter, G., Kumar, V. and Srivastava, R. (2004), “Measuring marketing
productivity: current knowledge and future directions”, Journal of Marketing, Vol. 68,
pp. 76-89.
Scholtens, B. (2000), “Competition, growth, and performance in the banking industry”, Center for
Financial Institutions Working Papers 00-18, Wharton School Center for Financial
Institutions, University of Pennsylvania, Philadelphia, PA.
Singh, S., Faircloth, S. and Nejadmalayeri, A. (2005), “Capital market impact of product
marketing strategy: evidence from the relationship between advertising expenses and the
cost of capital”, Journal of the Academy of Marketing Science, Vol. 33, pp. 432-44.
Smirlock, M. (1985), “Evidence on the (non)relationship between concentration and pro?tability”,
Journal of Money, Credit, and Banking, Vol. 17, pp. 69-83.
Srinivasan, S. and Hanssens, D. (2009), “Marketing and ?rm value: metrics, methods, ?ndings,
and future directions”, Journal of Marketing Research, Vol. 46, pp. 293-312.
Szewczyk, S., Tsetsekos, G. and Zantout, Z. (1996), “The valuation of corporate R&D
expenditures: evidence from investment opportunities and free cash ?ow”, Financial
Management, Vol. 25, pp. 105-10.
Tybout, A. and Carpenter, G. (2000), “Creating and managing brands”, in Iocabucci, D. (Ed.),
Kellogg on Marketing, Wiley, New York, NY.
Verma, V. (1980), “A price theoretical approach to the speci?cation and estimation of the sales
advertising function”, Journal of Business, Vol. 55, pp. 115-37.
Williamson, O. (1979), “Transaction-cost economics: the governance of transactional relations”,
Journal of Law & Economics, Vol. 22, pp. 233-61.
Wooldridge, J. (2002), Econometric Analysis of Cross Section and Panel Data, MIT Press,
Cambridge, MA.
JFEP
2,4
344
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
Zantout, Z. and Tsetsekos, G. (1994), “The wealth effects of announcements of R&D expenditure
increases”, Journal of Financial Research, Vol. 17, pp. 205-16.
Further reading
Flannery, M. (1982), “Retail bank deposits as quasi-?xed factors of production”, American
Economic Review, Vol. 72, pp. 527-36.
Hyman, H. and Mathur, I. (2005), “Retrospective and prospective views on the marketing/?nance
interface”, Journal of the Academy of Marketing Science, Vol. 33, pp. 390-400.
Corresponding author
Donald J. Mullineaux can be contacted at: [email protected]
Bank marketing
investments
345
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)
This article has been cited by:
1. Sittichai Meesrijan, Wanno Fongsuwan. 2015. Structural Equation Model of Factors Affecting Thailand's
Commercial Bank Branch Expansion. Research Journal of Business Management 9, 319-334. [CrossRef]
2. Sittichai Meesrichan, Wanno Fongsuwan. 2014. An Analysis of Thai Commercial Banks Branch
Expansion Factors Including Leadership, Location, Cost and Economics. Research Journal of Business
Management 8, 552-565. [CrossRef]
D
o
w
n
l
o
a
d
e
d

b
y

P
O
N
D
I
C
H
E
R
R
Y

U
N
I
V
E
R
S
I
T
Y

A
t

2
1
:
3
9

2
4

J
a
n
u
a
r
y

2
0
1
6

(
P
T
)

doc_460431985.pdf