Description
The purpose of this paper is to examine the impact of bank branch location on the likelihood
of bank failure during the most recent financial crisis
Journal of Financial Economic Policy
Bank branch location and stability during distress
J ill M. Hendrickson Mark W. Nichols Daniel R. Fairchild
Article information:
To cite this document:
J ill M. Hendrickson Mark W. Nichols Daniel R. Fairchild , (2014),"Bank branch location and stability during
distress", J ournal of Financial Economic Policy, Vol. 6 Iss 2 pp. 133 - 151
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-07-2013-0026
Downloaded on: 24 January 2016, At: 21:49 (PT)
References: this document contains references to 43 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 201 times since 2014*
Users who downloaded this article also downloaded:
Wenling Lu, David A. Whidbee, (2013),"Bank structure and failure during the financial crisis", J ournal of
Financial Economic Policy, Vol. 5 Iss 3 pp. 281-299http://dx.doi.org/10.1108/J FEP-02-2013-0006
Vighneswara Swamy, (2014),"Testing the interrelatedness of banking stability measures", J ournal of
Financial Economic Policy, Vol. 6 Iss 1 pp. 25-45http://dx.doi.org/10.1108/J FEP-01-2013-0002
Faten Ben Bouheni, (2014),"Banking regulation and supervision: can it enhance stability in Europe?",
J ournal of Financial Economic Policy, Vol. 6 Iss 3 pp. 244-269http://dx.doi.org/10.1108/J FEP-11-2013-0059
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Bank branch location and
stability during distress
Jill M. Hendrickson
Department of Economics, University of St. Thomas, St. Paul, Minnesota, USA
Mark W. Nichols
Department of Economics, University of Nevada Reno, Reno, Nevada, USA, and
Daniel R. Fairchild
Department of Economics, University of St. Thomas, St. Paul, Minnesota, USA
Abstract
Purpose – The purpose of this paper is to examine the impact of bank branch location on the likelihood
of bank failure during the most recent fnancial crisis.
Design/methodology/approach – This paper estimates the probit regression to identify the causes
of bank failures and attempts to determine the role of branch location in bank performance.
Findings – Using data from failed and surviving banks in Georgia and Florida, this paper fnds that
diversifying the balance sheet and operating in more competitive markets reduced failure rates, but
branching intensity, measured by number of branches and distance of branches from the home offce
did not signifcantly reduce the probability of failure. This suggests that, at least in today’s market, it is
not important to bank stability to have a branching network a signifcant distance fromthe home offce.
Originality/value – This paper carefully considers the role of branch location in the likelihood of
bank failure during fnancial distress. As such, it contributes to the historical policy debate regarding
regulation prohibiting or minimizing banks’ ability to branch. It also contributes to our understanding
of how banks structure their branching networks in the contemporary banking environment.
Keywords Financial crisis, Branch banking, Bank failures
Paper type Research paper
The USAhas a unique history with branch banking. Since late in the eighteenth century,
during our earliest experience with commercial banking, banks have frequently been
either banned from, or limited in, their ability to branch. From the beginning, there has
existed a policy struggle between those who wanted more branching and those who
wanted to preserve a systemof unit (non-branching) banks. For more than two hundred
years, the policy battle played out until the 1994 passage of the Riegle-Neal Interstate
Banking and Branching Effciency Act (IBBEA). This established the rights to
interstate branching and expanded, in many cases, intrastate branching.
Banks responded to the 1994 legislation by opening increasingly more branches
(Johnson and Rice, 2008). Indeed, the branch system, for many banks, became the model
for growth and stability (Dick, 2006). However, in the wake of the most recent fnancial
crisis, some scholars argued that banks failed to branch far enough fromthe home offce,
which is why they were vulnerable when the fnancial system became fragile
(Aubuchon and Wheelock, 2010). Furthermore, there is a growing group that believes
JEL classifcation – E44, G21, G28
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
Bank branch
location
133
Journal of Financial Economic Policy
Vol. 6 No. 2, 2014
pp. 133-151
©Emerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-07-2013-0026
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the branching system is outdated because of how technology has changed the bank
customers’ interaction with fnancial institutions. Baldwin (2011) argues that
technology, including mobile applications and younger generations of savers already
comfortable with new technology may utilize technology as a substitute for
branching[1]. Last year an article in The Economist (The Economist, 2012b, “Withering
Away”) argued that while branching has enjoyed tremendous growth in the twenty-frst
century, the fnancial crisis and resulting newregulatory environment has signifcantly
compromised bank revenue, making branches unproftable. Furthermore, it is expected
that as bank customers become increasingly comfortable with online services, such as
tax returns, they will embrace electronic banking and so reduce branch use. The article
predicts that there will be far fewer and more effcient branches in the future.
Certainly, many banks are more effcient and diversifed as a result of expanded
branching opportunities. However, some scholars argue that banks have not taken full
advantage of opportunities to geographically expand and so remain relatively
undiversifed. For example, since 2008, close to 400 commercial banks have failed in the
USA. In trying to understand this wave of bank failures, Aubuchon and Wheelock
(2010) argue that banks are not as diversifed as they should be, which led, in part, to
their failure. While Aubuchon and Wheelock (2010) recognize that banks are branching
in record numbers, these authors point out that they are not branching at a signifcant
distance from their home offce. In other words, the aggregate picture may refect that
banks are branching in greater numbers, but are they branching at meaningful
distances? Do the aggregate branching data conceal important trends? The point is, if
the branching is simply within a short radius fromthe home offce, then the full benefts
of branching (diversifcation, in particular) may not be captured.
The purpose of our research is two-fold. First is to determine the extent to which
commercial banks are branching in the post-1994 era by disaggregating the national
data. We carefully consider the branching behavior of a sample of banks by paying
particular attention to the distance and size of the branch from the home offce. Most
banks in the USA that utilize a branch network do not engage in interstate branching.
Indeed, ?90 per cent of all US commercial banks branch exclusively within state or do
not branch at all. Our bank sample contains both failed and surviving banks in Florida
and Georgia. Most of these banks are not engaged in interstate branching. Thus, our
analysis is of the behavior of banks engaged primarily in intrastate branching with
limited interstate-branching activity. Second, empirical methods are used to determine if
the distance and number of branches impact the probability of bank failure. Are the
branching patterns at failed banks different from surviving banks? Are Aubuchon and
Wheelock (2010) correct that failed banks have not branched at signifcant distances
fromthe home offce? To the author’s knowledge, no studies exist which link, at the frm
level, branch distance data with bank failures.
This paper is organized as follows. The next section of the paper briefy reviews recent
branch activity in the USA. The second section of the paper provides a reviewof the salient
literature. The thirdsectionof the paper explains our data, banksample, as well as empirical
methods. Results are found in section four and the fnal section concludes.
1. Contemporary branching activity in the USA
How have banks responded to the opportunities to bank and branch more freely?
Certainly, if there are signifcant benefts to branching, we would expect to see more
JFEP
6,2
134
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
branching per banking institution than in the past. This appears to be the case. As
shown in Dick (2006), both coverage and branch density has increased following the
deregulation of interstate branching. Between 1997 and the end of 2012, there has been
? 37 per cent increase in the aggregate number of branches. The number of branches
per bank increased from5.72 in 1995 to 13.73 branches per bank in 2012 as illustrated in
Figure 1. Furthermore, many of these branches are interstate in nature. As of June 30,
1997, the date that the IBBEA became effective, there were 8,876 interstate branches
throughout the country[2]. By June 30 2012, there were 43,330 interstate branches in the
USA – a 388 per cent increase in the aggregate number of branch locations operating
across state lines (see Figure 2). Thus, the degree to which banks are using branch
offces has steadily increased since the 1994 passage of the IBBEA.
Figure 3 illustrates the growth rate of branches per commercial bank in the USA. It is
interesting that the rate of growth was highest in the years prior to the IBBEA passage
in 1994, and then spiked again in the years prior to the 1997 fnal implementation of the
act. Given the timing of these growth spurts, most of this must have been intrastate
branching. It is possible that incumbent banks anticipated the passing of the IBBEAand
wanted to act prior to policy change. One can imagine two motivating factors for
0
2
4
6
8
10
12
14
16
Note: The number of branches per bank is the ratio of the total number of
branches to the total number of US commercial banks
Source: FDIC at www2.fdic.gov
Figure 1.
Number of branches per
US commercial bank
Source: FDIC at www2.fdic.gov
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
1994 1997 2011 2012
Figure 2.
Total number of interstate
branches
135
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
intrastate branch expansion prior to these dates. First, the existing banks may have
wanted to capture additional market share before the increased competition from
out-of-state branches. Second, and related to the frst, banks may have expanded their
branch network as a strategy of deterring future entry. Recent work by Cohen and
Mazzeo (2010) fnd empirical support for both hypotheses. The empirical work of these
scholars fnds evidence that competition from banks with multiple branches induces
additional branching from other institutions. Furthermore, they fnd that incumbent
banks are able to discourage entry by increasing their own branch network. The
increased growth rate in the 1990s, shown in Figure 3, is plausibly consistent with the
Cohen and Mazzeo (2010) fndings. The increased growth rate in 2006 may refect both
intrastate and interstate branch expansion to capture the growth in residential real
estate.
As mentioned above and illustrated in Figure 2, the total number of interstate
branches has grown signifcantly. However, most commercial banks do not engage in
interstate branching – as of September 30 2012, 566 banks, or 9.17 per cent of all
commercial banks branch across state lines[3]. This means, of course, that almost 91 per
cent do not branch outside of the state in which they are headquartered. Table I
illustrates that even within the small population of banks engaged in interstate
branching the fve largest banks account for almost half of all interstate branching in the
USA. At the other end, some very small banks also engage in interstate branching, but
typically these banks only have one branch[4]. The average size of the 566 banks
currently branching across state lines is just ? $18 billion in assets. As a point of
comparison, the average size of a commercial bank in the USAis just over $13 billion in
assets[5]. Thus, interstate branching banks are larger, on average, than the average of
the entire population of banks. Not surprisingly, then, the few banks that do engage in
extensive interstate branching hold the vast majority of deposits. In 24 of the states,
more than half of the bank deposits are held at a bank from outside of the state. To
Source: FDIC at www2.fdic.gov
Note: The rate of growth is calculated from the ratios in Figure 1
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
Figure 3.
Rate of growth in the
number of branches per
US commercial bank
JFEP
6,2
136
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
illustrate, consider, for example, that in Maryland, 79.2 per cent of bank deposits are
interstate deposits[6]. In only 11 states are interstate deposits ?25 per cent of the total.
Thus, while most banks do not branch across state lines, the few that do so extensively
have captured, in many cases, the majority of deposits within the state.
The post-IBBEA branching landscape is one with many more branch locations
despite corresponding to a time in which bank customers increasingly utilize
technological and electronic banking. However, most banks continue to operate their
branches within their headquartered state. Indeed, almost 91 per cent of US commercial
banks do not engage in interstate branching. Of the population of banks that do engage
in interstate branching, a small portion of them constitute the majority of interstate
branches.
Thus, interstate branching in the USA is dominated by a relatively small number of
commercial banks. This may be interpreted to illustrate that the IBBEA provision to
allow for interstate branching was not widely embraced across a signifcant number of
banks.
2. Existing literature
There are two felds of literature that inform this research. First is the work that
considers the ongoing debate regarding the benefts of branching and the effect of
branching restrictions on bank stability. Second, there is great interest among
regulators and policy makers on the impact that branching has on the nature and
defnition of banking markets. Certainly, this is of great interest because of antitrust and
other fnancial policy issues. Both of these facets of the literature are briefy reviewed
here.
2.1 The stabilizing impact of branching
During the long period of US commercial banking in which both interstate banking and
branching were restricted, there were many debates about the merits of such a
regulatory policy[7]. Fromnearly the beginning there were scholars, bankers and policy
makers who argued that a branching systemwould be more stable. Branching is said to
bring about fnancial stability through two channels: frst, through increased
Table I.
Interstate branching at
commercial banks by
bank size: September
30, 2012
Bank name
Total assets
($000)
Number of states
(including DC)
Number of
interstate offces
Percentage of
total interstate
offces in the USA
(per cent)
Largest banks
JPMorgan Chase Bank 1,850,218,000 26 5,366 12.38
Bank of America 1,448,273,067 36 5,434 12.54
Citibank 1,365,026,000 17 1,058 2.44
Wells Fargo 1,218,796,000 41 6,290 14.51
US Bank 342,627,272 28 2,830 6.53
Total 20,978 48.40
Notes: All data are as of September 30, 2012, unless otherwise noted
Source: Individual bank data are from the Institutional Directory and total branching data are from
both the Institutional Directory and Historical Statistics on Banking all from www2.fdic.gov
137
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
diversifcation, and second, through increased competition. Scholars and bankers long
argued against these restrictions, claiming that they kept banks from diversifying their
balance sheets which, in turn, made themmore vulnerable to local economic downturns.
Another prevalent argument was that the prohibition on interstate banking and
branching created local monopolies in banking and allowed ineffcient banks to remain
in business due to reduced competition. Most scholarship more than ten years old uses
aggregate data and typically does not distinguish between the diversifcation and
competition channel. Indeed, the vast majority of the established scholarship focuses
primarily on how branching may be stabilizing through diversifcation. More recent
scholarship often relies on frm level data and attempts to distinguish between the two
channels through which branching may bring about stability.
Many scholars have studied bank failures during the Great Depression to understand
the role branching played. Some of this work relied on aggregate level data and typically
fnds that branching reduces failure rates. For example, Mitchener (2005) uses
county-level data to test whether regulatory and supervisory differences impacted bank
stability during the Great Depression. Of interest to this study is his fnding that bank
failure rates were higher in counties located in states that prohibited branching.
Mitchener (2007) studies bank failures during the Great Depression and fnds that
branching was stabilizing because it introduced competition and forced ineffcient
banks out of the market either through closure or merger or because it enabled banks to
reduce risk via diversifcation. Calomiris (1990) studies the antebellumand 1920s eras of
commercial banking. During these historical periods, states experimented with state
deposit insurance as well as state-level branching laws. His empirical fndings suggest
that bank-asset growth was signifcantly higher in states that permitted branching. At
the same time, banks in branching states failed at a lower rate than banks in unit
banking states.
The scholarship on the Great Depression that relies on bank level data, perhaps
somewhat surprisingly, tends to be at odds with the aggregate data studies. Rather than
fnd branching to enhance stability, these studies fnd branching may lead banks to take
on additional risk. For example, Carlson (2004) uses bank-level data to determine
whether banks that branched were more likely to survive during the Great Depression.
His fnding, which is surprising because it goes against conventional wisdom, is that
branch banks are less likely to survive than unit banks. He argues that this does not
mean that branching is destabilizing but, rather, that branch bankers made decisions to
increase returns rather than reduce risk. Calomiris and Mason (2003) also fnd that,
during the Great Depression, banks that were members of the Federal Reserve and
engaged in branching failed sooner than unit banks. Much like Carlson, these authors
offer that branching banks may have been more diversifed which permitted them to
take on greater risks. Furthermore, Calomiris and Mason argue that while branching
can provide some diversifcation benefts when there are sector-specifc downturns, it
cannot offer protection when the downturn is across many sectors. In contrast, Ramirez
(2003) utilizes bank-level data and fnds that branching is stabilizing through increased
diversifcation in the years just prior to the Great Depression.
A second body of literature focuses on the competitive channel through which
branching may impact stability. One perspective argues that branching would enhance
competition and stability because weaker banks would be driven out of the market
leaving behind a more stable banking system (see, for example, Carlson and Mitchener,
JFEP
6,2
138
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
2006, Jayaratne and Strahan, 1998, Stiroh and Strahan, 2003). This literature suggests
that while competition reduces the profts of surviving banks, the weaker and ineffcient
banks are driven out of the market, leaving a stronger set of banks. Related, Carlson and
Mitchener (2007) fnd that increasing competition drives incumbent banks to improve
effciency and proftability. In contrast, others (see Collins (1926) or Sprague (1903)) have
argued that branching would create monopoly banking markets. The argument is that
branching banks would drive the smaller banks out of market creating “monopolies” of
one or two larger banks. This perspective was more prevalent historically and has
garnered less empirical support. Consequently, the current state of the literature tends to
favor the view that branching increases competition which, in turn, enhances bank
stability.
Only more recent scholarship on branching attempts to identify the specifc channel
through which branching may bring about stability. Carlson and Mitchener (2006 and
2007) have been the most deliberate in trying to tease out of empirical analysis, whether
it is through diversifcation or through increased competition that branch systems are
more stable. Carlson and Mitchener (2006) fnd that the increase in competition was
more important in bringing about stability during the Great Depression than was
diversifcation. Using frm-level data for banks in California, Carlson and Mitchener
(2007) fnd evidence that branch banking during the Great Depression forced existing
unit banks to be more effcient as a result of competition from branching banks. In this
way, branching is said to produce an externality: improved performance at unit banks to
enhance overall bank stability.
2.2 Banking market literature
Scholars have become increasingly interested in how technology and policy changes,
such as the IBBEA, impact the structure of banking. Such structural issues have
important antitrust implications since regulators and antitrust evaluations require
accurate defnitions of “a banking market”. Prior to the IBBEA, the standards used by
regulators and policy makers to defne a banking market were largely consistent with
statistical areas. Amel and Starr-McCluer (2002) and Mishkin and Strahan (1999)
indicate a “local banking market” is defned for both research and policy purposes as the
metropolitan statistical area (MSA) or non-MSA counties particularly for bank merger
purposes. Amel and Starr are more specifc with regards to rural areas and explain that
the market is typically assumed to be the size of one or two counties. Thus, the existing
standard is to defne a bank market to be the MSA or rural counties.
With the passage of the IBBEA, scholars are concerned that existing methods of
defning bank markets may no longer be appropriate. Some empirical research suggests
that banking markets remain local for both depositors and small business (see Heitfeld
and Prager, 2004). Using data from the 1998 Survey of Consumer Finances (SCF), Amel
and Starr-McCluer (2002) fnd that the average consumer banks within a two-mile radius
from their home. Most recently, Amel et al. (2008) analyzed data from the 2008 SCF and
fnd that consumers bank within a four-mile radius from the bank which suggests that
banking markets are still local. They defne a local banking market to be a 30-mile
radius, but admit this is somewhat of an arbitrary defnition.
Also using the SCF data, Robbins (2006) hypothesizes that as consumers
increasingly embrace electronic banking technology, the location of the bank and its
branches may be less important. Indeed, in 2007, 49.4 per cent of families reported
139
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
using electronic banking (Bucks et al. (2009)). However, 2007 survey data indicate
that consumers are increasingly utilizing electronic banking products, but still
value location more than any other factor in choosing a bank[8]. In fact, 45.9 per cent
of survey respondents identifed bank location as the most important reason for choosing a
bank. The secondmost important reasonwas identifedbyonly16.2per cent of respondents.
Finally, location appears to be more important now, as more respondents chose
location over past survey results (Bucks et al., 2009). Despite the benefts of
electronic banking services, it appears depositors still want the convenience of
banks that are located relatively close. Like the work of Amel and Starr-McCluer
(2002), Amel et al. (2008) and Heitfeld and Prager (2004), this suggests that bank
markets remain local, at least for depositors, despite the increased use of electronic
banking services.
In contrast to those who fnd that banking markets remain local, Petersen and Ragan
(2002) fnd that the distance between the bank and the small business borrower is
increasing. Using data from the National Survey of Small Business Finance (NSSBF),
the authors fnd that the distance between the bank and the borrower has been
increasing since the 1970s and through 1993. This is caused by technology that has
allowed bankers to use information about the borrower to both make lending decisions
and to monitor loans, reducing the need for the borrower and lender to meet each other
in person. Brevoort et al. (2010) revisit the work of Petersen and Ragan by including
NSSBF data from 1998 and 2003. They fnd distance increased at a faster rate between
1993 and 1997, but halted and even decreased between 1998 and 2003. They interpret
this to mean that the market for small business lending remains local.
2.3 Contributions of this work
It appears that bankers were eager to expand their branch networks in the post-IBBEA
era as illustrated in Figures 1 and 2. The question is was the branching suffcient enough
to add stability to banking? Are banks that branch in further distances from the home
offce less likely to fail? Did banks diversify their balance sheets via branching and, in
the process, reduce the likelihood of failure? While there is existing literature that
considers the relationship between branching and the probability of failure, there is no
literature, to the authors’ knowledge, that consider the intensity of branching and the
probability of failure. Here, branching intensity is measured as both the distance of the
branch from the home offce as well as the number of branches in each bank
organization.
The hypothesis of Aubuchon and Wheelock (2010) is that banks failed to branch in
signifcant ways so as to stabilize banking in recent years. That is, one possible
explanation for the large increase in the number of bank failures following the 2007-2008
fnancial crisis is that US commercial banks had an opportunity to engage in meaningful
interstate, or even intrastate, branching following the 1994 IBBEA, but that they chose
not to. This decision cost the banks because they remained less diversifed by not
expanding and/or their unwillingness to branch allowed ineffcient banks to remain
viable longer than had they faced competition from branch expansion. This empirical
question remains unresolved and the purpose of this paper is to try and fnd satisfactory
answers to this line of inquiry.
JFEP
6,2
140
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Data and empirical analysis
To test whether banks are branching in meaningful distances and whether branch
behavior impacts failure, this work relies on bank-level, branch-level and MSA-level
data. Specifcally, we rely on a randomsample of failed and non-failed banks in Georgia
and Florida. While examining two states is a limiting factor, of the 337 commercial bank
failures between January 2008 and October 2011, Georgia (69 failures) and Florida (47
failures) account for 34 per cent of all failures in the country. These two states,
nonetheless, represent different experiences in the twenty-frst century. Florida
experienced signifcant home infation, followed by signifcant defation surrounding
the fnancial crisis. In contrast, Georgia did not see the home-price swings that Florida
experienced, but suffered even more bank failures than Florida following the crisis.
Thus, to include these two states is to capture both post-crisis experiences in the housing
market.
For each bank in the sample, we use balance sheet and income data for each quarter
between 2000 and 2011. Furthermore, we have annual branch activity data for each
bank, including how many branches the bank operated, the number of miles each
branch is from the home offce, the location of the branch, the type of services offered at
each branch and the size, measured by total deposits, of each branch. Bank-level data
come from the Federal Deposit Insurance Corporation and is available online. In
addition, MSA-level data on total population, housing prices, income and
unemployment are collected to account for local market conditions. All MSA-level
data are public information; population data are from the Census Bureau; home
indices are from the Federal Housing Finance Agency; real per capita personal
income is from the Bureau of Economic Analysis; and unemployment data come
from the Bureau of Labor Statistics.
The number of banks in our sample varies from year to year due to mergers,
openings, and, of course, failures. In 2007, the last year prior to the increase in failures
taking place during 2008-2011, we have 169 commercial banks in our sample, 96 of
which are in Georgia and 73 in Florida. Of those, 82 would go on to fail sometime
between 2008 and 2011, with 49 failures occurring in Georgia and 33 in Florida. Table II
describes and provides summary statistics for our key variables over the years
2000-2011.
3.1 Descriptive statistics for entire sample
Table II demonstrates that, of those banks with branches (approximately 77 per cent of the
sample), most have very fewbranches, and those branches are close to the home offce. The
average distance fromthe home offce is 17.47 miles. However, the median is 9.75 miles and
the ninetiethpercentile is only38 miles. Indeed, 82 per cent of branches are within30 miles of
the home offce, and only 0.43 per cent of branches are interstate. The average number of
branches is just over 3; the median is 2 and the ninetieth percentile 7.
3.2 Failed versus surviving sample of banks
Table III provides a comparison of means between failed banks and branch banks,
respectively. Columns two and three of Table III reveal several interesting differences
between failed and surviving banks. In terms of branching, for those banks that had
branches, failed banks had, on average, approximately one more branch (3.81 vs 2.97)
and branched at a greater distance fromthe home offce (21.7 miles vs 14.1). Also of note,
141
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
however, failed banks had a signifcantly greater proportion of unit banks (31.4 per cent)
than surviving banks (18.10 per cent).
Table III also demonstrates differences in risk taking between failed and surviving
banks. Specifcally, failed banks, while larger in terms of total assets, had a lower return
on assets and a lower capital to asset ratio. They also had a larger proportion of assets
in loans and lower proportion in securities.
Failed banks had a signifcantly higher proportion of real estate loans, but no
difference in past due residential mortgages. Failed banks had a lower proportion of
commercial and industrial loans and were also less diversifed[9]. The failed sample was
also much less experienced than the surviving banks, using age as a proxy for
experience (Carlson (2004)).
Table II.
Variable description and
descriptive statistics, all
banks, 2000-2011
Variable Defnition
Mean (standard
deviation)
Branch variables
Average miles Average distance of bank branches from main offce 17.470 (28.97)
Number of branches Average number of branches bank maintains 3.347 (5.80)
Average branch deposits Average deposits at braches (thousands of dollars) 19318.82 (21582.19)
Percentage of branches out of state Percentage of a bank’s branches that are out of state 0.436 (4.83)
Percentage of branches within 30
miles
Percentage of a bank’s branches that are within 30
miles of the main bank
82.033 (29.35)
Percentage of banks with no
branches
Percentage of banks that have no branches 23.168 (42.20)
Bank variables
Failure Dummy equal to 1 if the bank failed any time during
2000-2011
0.441 (0.496)
Deposit market share The bank’s share of total deposits of all FDIC-
insured banks within the bank’s headquarter city as
of June 30 for each year
24.78 (29.61)
Total assets Total assets of the bank (thousands of dollars) 258369.5 (370021)
Return on assets Net operating income after taxes and extraordinary
items divided by total assets
0.075 (2.09)
Loan-to-asset ratio Total loans divided by total assets 0.689 (0.135)
Security-to-asset ratio Total securities divided by total assets 0.154 (0.113)
Capital-to-asset ratio Total capital divided by total assets 0.109 (0.074)
Non-interest income-to-asset ratio Non-interest income divided by total assets 0.0040 (0.0047)
Real estate loans to total loans Real estate loans divided by total loans 0.807 (0.137)
Commercial and industrial loans to
total loans
Commercial and industrial loans divided by total
loans
0.125 (0.087)
Age Age of the bank, years 34.14 (34.91)
Derivatives to assets Total derivatives divided by total assets 0.0053 (0.038)
Core deposits to total deposits Core deposits divided by total deposits 0.768 (0.109)
Past due mortgages Past due residential mortgages divided by total
residential mortgages
0.017 (0.020)
Diversity Sum of squared ratios of business loans, real estate
loans, consumer loans and securities to total assets
0.392 (0.126)
MSA variables
Change in home price index Annual fourth quarter percentage change in house
price index in the bank’s MSA
3.59 (9.82)
Unemployment rate Unemployment rate in the bank’s MSA 5.66 (2.36)
Population Population of the city where bank located 97803.56 (169532)
Real per capita income Per capita income in the bank’s MSA, 2007 dollars 37352.95 (5591.96)
Georgia Dummy equal to 1 if bank is in Georgia 0.56 (0.496)
JFEP
6,2
142
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table III.
Comparison of means by
bank failure and branch
status, 2000-2011
V
a
r
i
a
b
l
e
F
a
i
l
e
d
b
a
n
k
s
(
n
?
7
2
6
)
S
u
r
v
i
v
i
n
g
b
a
n
k
s
(
n
?
8
9
5
)
D
i
f
f
e
r
e
n
c
e
:
f
a
i
l
e
d
-
s
u
r
v
i
v
i
n
g
(
t
-
s
t
a
t
i
s
t
i
c
)
N
o
b
r
a
n
c
h
b
a
n
k
s
(
n
?
3
8
9
)
B
r
a
n
c
h
b
a
n
k
s
(
n
?
1
2
3
3
)
D
i
f
f
e
r
e
n
c
e
:
n
o
b
r
a
n
c
h
-
b
r
a
n
c
h
(
t
-
s
t
a
t
i
s
t
i
c
)
B
r
a
n
c
h
v
a
r
i
a
b
l
e
s
A
v
e
r
a
g
e
m
i
l
e
s
2
1
.
7
3
1
4
.
1
0
7
.
6
3
*
*
*
(
4
.
9
4
)
N
u
m
b
e
r
o
f
B
r
a
n
c
h
e
s
3
.
8
1
2
.
9
7
0
.
8
4
*
*
*
(
2
.
7
6
)
A
v
e
r
a
g
e
b
r
a
n
c
h
D
e
p
o
s
i
t
s
1
9
3
0
0
.
9
3
1
9
3
3
2
.
9
4
?
3
2
.
0
1
(
?
0
.
0
2
)
P
e
r
c
e
n
t
o
f
b
r
a
n
c
h
e
s
o
u
t
o
f
s
t
a
t
e
0
.
4
5
8
0
.
4
4
7
0
.
0
1
0
8
(
0
.
0
4
)
P
e
r
c
e
n
t
a
g
e
o
f
b
r
a
n
c
h
e
s
w
i
t
h
i
n
3
0
m
i
l
e
s
7
5
.
2
8
8
6
.
5
7
?
1
1
.
2
9
*
*
(
?
6
.
4
2
)
P
e
r
c
e
n
t
a
g
e
o
f
b
a
n
k
s
w
i
t
h
n
o
b
r
a
n
c
h
e
s
3
1
.
4
6
1
8
.
1
0
1
3
.
3
6
*
*
*
(
6
.
1
9
)
B
a
n
k
v
a
r
i
a
b
l
e
s
F
a
i
l
u
r
e
0
.
5
7
8
0
.
3
9
7
0
.
1
8
1
*
*
*
(
6
.
2
9
)
D
e
p
o
s
i
t
m
a
r
k
e
t
s
h
a
r
e
2
1
.
1
3
2
8
.
1
6
?
7
.
0
3
*
*
*
(
4
.
8
6
)
1
9
.
7
3
2
6
.
2
8
?
6
.
5
5
*
*
*
(
3
.
8
9
)
T
o
t
a
l
a
s
s
e
t
s
3
1
6
0
5
4
.
5
0
2
1
9
4
5
7
.
1
0
9
6
5
9
7
.
4
7
*
*
*
(
4
.
8
6
)
1
4
0
2
4
4
.
4
0
2
9
3
9
9
0
.
2
0
?
1
5
3
7
4
5
.
8
0
*
*
*
(
8
.
7
7
)
R
e
t
u
r
n
o
n
a
s
s
e
t
s
?
0
.
2
9
2
0
.
4
7
5
?
0
.
7
6
7
*
*
*
(
?
7
.
1
7
)
?
0
.
4
1
4
0
.
2
2
3
?
0
.
6
3
7
*
*
*
(
4
.
2
7
)
L
o
a
n
t
o
a
s
s
e
t
r
a
t
i
o
0
.
7
3
0
0
.
6
6
4
0
.
0
6
6
*
*
*
(
1
0
.
9
8
)
0
.
6
5
8
0
.
6
9
9
?
0
.
0
4
1
*
*
*
(
4
.
5
8
)
S
e
c
u
r
i
t
y
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
1
1
8
0
.
1
8
4
?
0
.
0
6
5
*
*
*
(
?
1
2
.
8
2
)
0
.
1
6
2
0
.
1
5
2
0
.
0
1
0
(
1
.
2
9
)
C
a
p
i
t
a
l
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
1
0
0
0
.
1
0
9
?
0
.
0
0
9
*
*
*
(
?
3
.
3
6
)
0
.
1
4
0
.
1
0
0
.
0
4
*
*
*
(
8
.
7
8
)
N
o
n
-
i
n
t
e
r
e
s
t
i
n
c
o
m
e
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
0
0
4
0
0
.
0
0
4
1
?
0
.
0
0
0
1
(
?
0
.
5
2
)
0
.
0
0
3
5
0
.
0
0
4
1
?
0
.
0
0
0
6
*
(
1
.
7
8
)
R
e
a
l
e
s
t
a
t
e
l
o
a
n
s
t
o
t
o
t
a
l
l
o
a
n
s
0
.
8
3
9
0
.
7
7
9
0
.
0
5
9
*
*
*
(
9
.
0
3
)
0
.
7
6
9
0
.
8
1
8
?
0
.
0
4
9
*
*
*
(
5
.
5
0
)
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
t
o
t
o
t
a
l
l
o
a
n
s
0
.
1
1
6
0
.
1
3
2
?
0
.
0
1
6
*
*
*
(
3
.
5
8
)
0
.
1
4
7
0
.
1
1
8
0
.
0
2
9
*
*
*
(
5
.
0
0
)
A
g
e
2
6
.
4
3
4
1
.
8
6
?
1
5
.
4
3
*
*
*
(
9
.
0
9
)
2
1
.
4
2
3
7
.
9
7
?
1
6
.
5
5
*
*
*
(
8
.
6
4
)
D
e
r
i
v
a
t
i
v
e
s
t
o
a
s
s
e
t
s
0
.
0
1
1
0
.
0
0
1
2
0
.
0
0
9
*
*
*
(
4
.
3
4
)
0
.
0
1
1
0
.
0
0
3
0
.
0
0
7
*
*
(
2
.
0
4
)
C
o
r
e
d
e
p
o
s
i
t
s
t
o
t
o
t
a
l
d
e
p
o
s
i
t
s
0
.
7
4
2
0
.
7
8
6
?
0
.
0
4
4
*
*
*
(
?
8
.
2
4
)
0
.
7
5
6
0
.
7
7
1
?
0
.
0
1
5
*
*
(
2
.
3
2
)
P
a
s
t
d
u
e
m
o
r
t
g
a
g
e
s
0
.
0
1
7
7
0
.
0
1
7
6
0
.
0
0
0
1
(
0
.
0
9
)
0
.
0
1
5
0
.
0
1
8
?
0
.
0
0
3
*
*
*
(
2
.
9
3
)
D
i
v
e
r
s
i
t
y
0
.
4
2
8
0
.
3
6
7
0
.
0
6
1
*
*
*
(
1
0
.
0
8
)
0
.
3
6
2
0
.
4
0
0
?
0
.
0
3
8
*
*
*
(
4
.
9
2
)
M
S
A
v
a
r
i
a
b
l
e
s
C
h
a
n
g
e
i
n
h
o
m
e
p
r
i
c
e
i
n
d
e
x
3
.
9
4
3
.
3
3
0
.
6
1
(
1
.
2
2
)
4
.
6
1
3
.
2
9
1
.
3
2
*
*
(
2
.
5
6
)
U
n
e
m
p
l
o
y
m
e
n
t
r
a
t
e
5
.
4
1
5
.
8
4
?
0
.
4
3
*
*
*
(
?
3
.
7
5
)
5
.
3
1
5
.
7
6
?
0
.
4
5
*
*
*
(
3
.
7
7
)
P
o
p
u
l
a
t
i
o
n
8
9
5
5
0
.
9
2
1
0
6
5
5
4
.
3
0
?
1
7
0
0
3
.
3
8
*
*
(
2
.
0
0
)
1
1
2
9
6
5
.
9
7
9
3
2
0
4
.
9
3
1
9
7
6
1
.
0
4
*
(
1
.
8
1
)
R
e
a
l
p
e
r
c
a
p
i
t
a
i
n
c
o
m
e
3
8
3
1
7
.
1
7
3
5
7
5
4
.
4
7
2
5
6
2
.
7
0
*
*
*
(
7
.
8
9
)
3
7
5
5
5
.
9
8
3
6
7
0
1
.
1
4
8
5
4
.
8
4
*
*
(
2
.
4
7
)
N
o
t
e
s
:
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
.
A
*
,
*
*
a
n
d
*
*
*
r
e
p
r
e
s
e
n
t
s
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
f
c
a
n
c
e
a
t
t
h
e
1
0
,
5
a
n
d
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
.
B
e
c
a
u
s
e
n
o
t
a
l
l
v
a
r
i
a
b
l
e
s
a
r
e
n
o
r
m
a
l
l
y
d
i
s
t
r
i
b
u
t
e
d
,
t
e
s
t
s
f
o
r
s
i
g
n
i
f
c
a
n
t
d
i
f
f
e
r
e
n
c
e
s
w
e
r
e
a
l
s
o
e
s
t
i
m
a
t
e
d
u
s
i
n
g
t
h
e
n
o
n
-
p
a
r
a
m
e
t
r
i
c
W
i
l
c
o
x
o
n
R
a
n
k
S
u
m
t
e
s
t
.
R
e
s
u
l
t
s
w
e
r
e
q
u
a
l
i
t
a
t
i
v
e
l
y
i
d
e
n
t
i
c
a
l
143
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In summary, Table III indicates that failed banks were more likely to be unit banks, but
those that did branch had more branches at a greater distance. Failed banks had lower
returns on assets, were less experienced and took on greater risk. Failed banks also had
a greater deposit market share, suggesting they operated in less competitive
environments.
3.3 Branching versus unit sample of banks
Given the greater proportion of unit banks that failed, further analysis of the distinction
between unit and branch banks is provided in the last three columns of Table III. Unit
banks have a greater proportion of failures (57.8 per cent) than branch banks (39.7 per
cent). Unit banks also had a signifcantly lower (and negative) return on assets.
Consistent with Carlson (2004), branch banks took on more risk, have lower capital to
asset and security to asset ratios, a higher loan to asset ratio, a greater percentage of
loans in real estate, a greater percentage of past due mortgages and less diversifcation.
However, unit banks had a greater proportion of derivatives to assets, a lower
proportion of core deposits, more commercial and industrial loans and less non-interest
income. Unit banks were also less experienced (younger) and had a lower share of total
deposits in the city where they operated.
The descriptive statistics reveal an interesting pattern of failure and branching.
Failed banks had more branches and those branches were, on average, at a greater
distance from the home offce. Branch banks, however, took on more risk based on
measures such as capital to asset ratio, proportion of total loans in real estate, past due
mortgages, and loan diversifcation. In this regard, the descriptive statistics are
consistent with the branching results found in the Great Depression by Carlson (2004).
At the same time, however, unit banks accounted for a greater percentage of failures,
were younger and more exposed to commercial and industrial loans. Both geographic
and balance sheet differences therefore appear to be correlated with failure.
3.4 Probit analysis
To better distinguish the importance of branching and geographic versus balance sheet
factors a probit regression of bank failure is estimated. We focus on the large number of
bank failures that took place between 2008 and 2011. Consequently, we examine
whether the branching and balance sheet characteristics of the bank in 2007 infuenced
the probability the bank failed sometime during 2008-2011[10].
Our probit model is as follows:
FAIL
i
? ?
0
? ?
1
NOBRANCH
i
? ?
2
AVGMILES
i
? ?
3
BRANCHNUM
i
? ?
4
LN(ASSET
i
) ? ?
5
DEPMKTSHARE
i
? B
6
DIVERSITY
i
? ?
7
RESTATETOLOAN
i
? ?
8
DERIVTOASSET
i
? ?
9
COREDEPTODEPOSIT
i
? ?
10
ROA
i
? ?
11
AGE
i
? ?
12
CHNGHOMEPINDEX
i
? ?
13
CITYPOP
i
? ?
14
UNEM
i
? ?
15
PCINC
i
??
16
GEORGIA
i
? ?
i
(1)
In equation (1), FAIL
i
is a dummy variable equal to one if bank i fails any time between
2008 and 2011. The independent variables are measured in 2007. NOBRANCH
i
is a
dummy equal to one if bank i had no branches in 2007. AVGMILES
i
is the average
JFEP
6,2
144
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
distance branches are fromthe home offce in 2007, and BRANCHNUM
i
is the number of
branches that bank i operates in 2007. The above capture the banks’ branching activity
and geographic diversifcation[11]. DEPMKTSHARE
i
is the bank’s share of total
deposits in 2007 of all Federal Deposit Insurance Corporation (FDIC)-insured banks
where the bank is headquartered. This is designed to capture the competitive
environment in which the bank operates, with higher deposit shares representing more
concentrated and hence less competitive markets[12]. LN (ASSET
i
) is the natural log of
the bank’s assets in 2007 and captures bank size. Other variables account for
balance-sheet differences between banks. DIVERSITY
i
, as defned in the study by
Kolari et al. (2002), is the sum of squared ratios of business loans, real estate loans,
consumer loans and securities to total assets. Higher values represent less
diversifcation. RESTATETOLOAN
i
is the ratio of real estate loans to total loans. ROA
i
is the return on assets (net income to asset ratio). DERIVTOASSET
i
is the dollar value
of the banks derivatives to total assets and COREDEPTODEPOSIT
i
is the ratio of core
to total deposits. A banker’s engagement with derivatives is included to capture the
trend in commercial banking of off-balance sheet activity, while the core deposits
represent the cheapest source of funds to the bank. AGE
i
is the age of the bank in 2007
and proxies for the experience of the bank. The remaining variables capture local
market conditions. CHNGHOMEPINDEX
i
is the change in the fourth quarter home price
index for the MSAwhere the home bank is locate, CITYPOP
i
is the population of the city
where the home bank is located; UNEM
i
is the unemployment rate of the MSAwhere the
bank is located; PCINC
i
is real per capita income; and GEORGIA
i
is a dummy if the bank
was in Georgia and is included to account for any state-specifc effects.
Probit coeffcient estimates are provided in Column two of Table IV. Most relevantly
for this study, when controlling for balance sheet, risk-taking, and local market
differences between banks, unit banks are signifcantly more likely to fail than their
branching counterparts. The coeffcient on NOBRANCH is statistically signifcant at
the 5 per cent level. Its marginal effect is 0.20, indicating banks without any branches are
20 per cent more likely to fail. Interestingly, while not branching signifcantly increases
the likelihood of failure, branching intensity, measured as both average distance from
the home offce and total number of branches is not a signifcant determinant of failure
controlling for risk taking and other factors[13].
The coeffcient on DEPMKTSHARE is positive and statistically signifcant, suggesting
that banks withgreater shares of a city’s total FDICinsureddeposits were more likelyto fail.
To the extent deposit market share refects a bank’s competitive environment; this result
suggests, consistent with Carlson and Mitchener (2006), that increased bank competition
reduces the probability of failure by weeding out weaker banks.
Consistent with Lu and Whidbee (2013), DIVERSITYis also statistically signifcant.
Because higher values represent less diversifcation, the results suggest that banks with
less diversifed balance sheets were more likely to fail. Banks with a greater proportion
of their assets in derivatives were also more likely to fail, but a greater proportion of
loans in real estate was not found to be signifcant. Thus, perhaps surprisingly, it was
not real estate loans that increased the probability of failure but, rather, lack of
diversifcation and derivative activity which explain failures during this period.
Furthermore, banks with a greater proportion of core deposits, a higher net income to
asset ratio, andmoreexperienced(older) bankswerelesslikelytofail. LuandWhidbee(2013)
fnd that de novo banks are more likely to fail during this same time period, which is
145
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table IV.
Probit regression results
for bank failure (banks
failing during 2008-2011)
V
a
r
i
a
b
l
e
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
m
e
a
s
u
r
e
d
i
n
2
0
0
7
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
l
a
g
g
e
d
o
n
e
y
e
a
r
B
r
a
n
c
h
v
a
r
i
a
b
l
e
s
N
o
b
r
a
n
c
h
e
s
(
N
O
B
R
A
N
C
H
)
0
.
7
6
1
*
*
(
0
.
3
6
0
)
0
.
5
3
8
*
(
0
.
3
1
6
)
A
v
e
r
a
g
e
m
i
l
e
s
(
A
V
G
M
I
L
E
S
)
?
0
.
0
0
2
4
(
0
.
0
0
6
5
)
?
0
.
0
0
2
8
(
0
.
0
0
5
7
)
N
u
m
b
e
r
o
f
b
r
a
n
c
h
e
s
(
B
R
A
N
C
H
N
U
M
)
0
.
0
0
4
9
(
0
.
0
2
4
0
)
0
.
0
0
1
2
(
0
.
0
2
4
)
B
a
n
k
b
a
l
a
n
c
e
s
h
e
e
t
v
a
r
i
a
b
l
e
s
D
e
p
o
s
i
t
m
a
r
k
e
t
s
h
a
r
e
(
D
E
P
M
K
T
S
H
A
R
E
)
0
.
0
1
4
8
*
*
(
0
.
0
0
6
4
)
0
.
0
1
5
4
*
*
(
0
.
0
0
6
)
D
i
v
e
r
s
i
t
y
(
D
I
V
E
R
S
I
T
Y
)
4
.
8
5
5
*
*
*
(
1
.
5
7
9
)
3
.
7
7
4
*
*
*
(
1
.
2
8
6
)
L
o
g
o
f
t
o
t
a
l
a
s
s
e
t
s
(
L
N
A
S
S
E
T
)
0
.
4
3
7
*
(
0
.
2
3
2
)
0
.
4
7
5
*
*
(
0
.
2
0
2
)
R
e
a
l
e
s
t
a
t
e
l
o
a
n
s
t
o
t
o
t
a
l
l
o
a
n
s
(
R
E
S
T
A
T
E
T
O
L
O
A
N
)
0
.
9
2
2
(
1
.
5
2
4
)
2
.
3
4
4
*
(
1
.
4
0
4
)
R
e
t
u
r
n
o
n
a
s
s
e
t
s
(
R
O
A
)
?
0
.
3
1
9
*
*
(
0
.
1
3
0
)
?
0
.
3
1
7
*
*
*
(
0
.
0
8
2
)
D
e
r
i
v
a
t
i
v
e
s
t
o
a
s
s
e
t
s
(
D
E
R
I
V
T
O
A
S
S
E
T
)
1
7
.
8
7
7
*
*
*
(
6
.
0
9
5
)
1
3
.
4
9
0
*
*
*
(
5
.
7
3
6
)
C
o
r
e
d
e
p
o
s
i
t
s
t
o
t
o
t
a
l
d
e
p
o
s
i
t
s
(
C
O
R
E
D
E
P
T
O
D
E
P
O
S
I
T
)
?
3
.
2
3
0
*
*
(
1
.
4
0
4
)
?
2
.
1
3
2
*
*
(
1
.
0
2
3
)
A
g
e
(
A
G
E
)
?
0
.
0
1
1
*
*
(
0
.
0
0
4
5
)
?
0
.
0
1
1
*
*
(
0
.
0
0
4
8
)
M
S
A
v
a
r
i
a
b
l
e
s
C
h
a
n
g
e
i
n
c
o
m
e
p
r
i
c
e
i
n
d
e
x
(
C
H
N
G
H
O
M
E
P
I
N
D
E
X
)
?
0
.
1
0
4
*
*
(
0
.
0
5
0
)
?
0
.
0
2
3
(
0
.
0
1
6
)
P
o
p
u
l
a
t
i
o
n
,
t
h
o
u
s
a
n
d
s
(
C
I
T
Y
P
O
P
)
?
0
.
0
0
0
1
(
0
.
0
0
0
0
8
)
?
0
.
0
0
0
1
(
0
.
0
0
0
0
7
)
U
n
e
m
p
l
o
y
m
e
n
t
(
U
N
E
M
)
?
0
.
6
9
5
*
*
(
0
.
3
1
6
)
?
0
.
2
8
4
*
*
*
(
0
.
0
5
4
)
R
e
a
l
p
e
r
c
a
p
i
t
a
i
n
c
o
m
e
,
t
h
o
u
s
a
n
d
s
(
P
C
I
N
C
)
0
.
0
0
0
5
(
0
.
0
0
0
0
2
)
0
.
0
0
0
5
(
0
.
0
0
0
0
2
)
G
e
o
r
g
i
a
(
G
A
)
0
.
9
6
8
*
*
(
0
.
4
5
8
)
0
.
2
2
3
(
0
.
3
2
0
)
P
s
e
u
d
o
R
2
0
.
3
7
7
0
.
3
1
4
L
o
g
l
i
k
e
l
i
h
o
o
d
?
7
2
.
9
5
?
1
9
4
.
2
1
N
u
m
b
e
r
o
f
b
a
n
k
s
1
6
9
1
6
8
N
o
t
e
s
:
R
e
p
o
r
t
e
d
p
r
o
b
i
t
c
o
e
f
f
c
i
e
n
t
s
.
R
o
b
u
s
t
s
t
a
n
d
a
r
d
e
r
r
o
r
s
,
c
l
u
s
t
e
r
e
d
o
n
b
a
n
k
s
,
i
n
p
a
r
e
n
t
h
e
s
e
s
.
A
*
,
*
*
a
n
d
*
*
*
r
e
p
r
e
s
e
n
t
s
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
f
c
a
n
c
e
a
t
t
h
e
1
0
,
5
a
n
d
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
JFEP
6,2
146
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
consistent with the fnding that younger banks are more likely to fail. Hendrickson and
Nichols (2011) studybankfailures inGeorgiaandalsofndthat these banks were more likely
to fail if they were young. This post-crisis literature is all consistent with DeYoung (1999),
who found that banks were most vulnerable to failure between the ages of three and fve
years. It is likelythat inbothGeorgia andFlorida manyof the banks hadenteredthe market
while house prices were rising and that these young banks were not yet in a position to be
able to weather the aftermath of the crisis.
Local market conditions also play some role. Rising home values are negatively
correlated with failure, and banks in Georgia were more likely to fail than banks in
Florida. Unemployment, however, is unexpectedly negative and signifcant, although
consistent with the results in Table III which shows a lower unemployment rate for
failed banks relative to surviving banks. Population and per capita income are not
signifcant determinants of bank failure in this study.
As a robustness check we also measure our independent variables lagged one year
rather than their 2007 variables because bank and economic conditions in 2007 may
have a different infuence on banks that failed in 2008 compared to those that failed later
in the sample[14]. These results are reported in column 3 of Table IV. The results and
conclusions are generally consistent across specifcations.
As a fnal robustness check, other control variables, including the loan-to-asset ratio,
security-to-asset ratio, average branch deposits, number of banks in the MSA, number
of non-banks in the MSA, past due mortgages, non-current loans to total loans,
mortgage-backed securities to total assets and non-interest income to assets, were
included as independent variables. These were not statistically signifcant and did not
change the statistical signifcance of the NOBRANCHcoeffcient or the insignifcance of
branch number and distance variables.
In summary, whether examining a simple comparison of means or estimating a
probit regression that controls for balance sheet and geographic differences between
banks, unit banks were more likely to fail during the most recent fnancial crisis than
banks that branched. Diversifying the balance sheet and operating in more competitive
markets reduced failure rates, but branching intensity, measured by number of
branches and distance of branches fromthe home offce, did not signifcantly reduce the
probability of failure.
4. Conclusion
This paper is concerned with the relationship between branch activity and intensity and
bank failure. The results suggest that unit banks are more likely to fail than banks that
have branches. This is important both from the perspective of the banker and from the
perspective of the policy maker. For the individual banker, this fnding suggests that
decisions to branch lessen the probability of failure. Branching, despite technological
advancement, is a strategy that can bring about stability to the banker. From a
policy-making standpoint, if branching enhances stability, policy should remain open to
the creation of branch networks. Furthermore, there is evidence to suggest that bank
failures reduce industrial production and economic growth (Kupiec and Ramirez, 2013)
and other evidence to suggest that branch networks reduce state-level business-cycle
volatility (Strahan, 2003)[15]. If branching can reduce failures, there are important
macroeconomic gains in stability to be had of pursuing a policy that encourages
branching.
147
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
While branching generally is found to be more stabilizing than not branching, branch
intensity does not appear to impact the probability of failure. The aggregate data on
bank branching suggest that ? 10 per cent of all commercial banks have taken
advantage of the policy change in 1994 that allows for interstate branching, but that
overall branching, including intrastate, is growing signifcantly. However, Aubuchon
and Wheelock (2010) were concerned that banks were not branching far enough fromthe
home offce (in miles) to capture geographic diversifcation benefts. The sample
analyzed here of both failed and surviving banks in Florida and Georgia, who are largely
engaged in intrastate branching, suggests that branching intensity, measured by both
number of branches and distance of branches from the home offce, does not
signifcantly impact the probability of failure. In other words, it suggests that even if
banks had branched in signifcant distances and numbers, the probability of failure
would not be affected. Some policy experts have, in the wake of the fnancial crisis,
argued for policy changes that would limit the size of banks. Our results suggest that
such limitations, at least as far as branching is concerned, may not compromise bank
stability.
In the end, we fnd that most banks do not engage in interstate branching and that the
distance of the branch from the home offce does not impact the probability of failure.
However, both balance sheet diversity and operating in more competitive environments
reduce the probability of failure. Like the fndings in Carlson and Mitchener (2006), our
results show increasing bank competition is stabilizing possibly by causing weaker
banks to exit. Our fndings also indicate that less diversifed banks are more likely to fail
consistent with existing literature (Lu and Whidbee, 2013). Thus, to the extent that
branch activity improves bank diversity and increases market competition, a bank
structural strategy that includes a branch network should be stabilizing for the bank.
Notes
1. An article in The Economist makes a similar prediction that branches will become less
important and that there will be far fewer of them as a result of technology (see The
Economist, 2012a “Counter Revolution”).
2. The interstate branching data come from the FDIC found at www2.fdic.gov
3. Branch data come from the FDIC at fdic.gov
4. The smallest fve banks engaged in interstate branching have total assets (in thousands)
ranging from $11,263 to $54,598 and four of the fve had one branch, total, but it was out of
state, and one of these small banks had two branches out of state. These data come fromthe
Institutional Directory at www2.fdic.gov
5. More specifcally, for interstate branching banks, assets average $18,263,658,000, and for the
entire population of commercial banks assets average $13,070,000,000. Data are from
www2.fdic.gov
6. The interstate branching data come from the FDIC found at www2.fdic.gov, and these
fgures are as of June 30, 2012.
7. See Hendrickson (2010) for a complete review of these historical debates. Related literature,
such as Huang (2008), considers whether branching affects local economic growth. This
paper focuses on the relationship between branching and bank stability.
JFEP
6,2
148
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
8. Electronic banking products include direct deposit, ATM cards, debit cards, automatic
payment services and computer banking services (Robbins (2006, p. 4).
9. Diversity, as defnedbyKolari et al. (2002), is the sumof squaredratios of business loans, real
estate loans, consumer loans and securities to total assets. Higher numbers therefore
represent less diversifcation, with a value of 1 indicating none.
10. In this regard, our methodology is similar to Carlson (2004) who examines how branching
and balance sheet differences in 1929 infuence the likelihood of the bank surviving the
Depression. Whereas Carlson examines survival, we examine failure. Carlson notes,
however, that his conclusions hold when analyzing failure.
11. Of course, a bank where NOBRANCH is equal to 1 will also have BRANCHNUM equal to 0
and AVGMILES equal to 0. The correlation between NOBRANCH and BRANCHNUM is
?0.33, and the correlation between NOBRANCH and AVGMILES is ?0.26.
12. Various measures of competition were created. For example, we also found the number of
branches of two large regional banks, Suntrust and Regions, located in the headquarter city
of each parent bank in our sample. The results remained unchanged, and these branches
were not statistically signifcant. Banks with a greater share of deposits are also likely to be
larger. The correlation between deposit markets share and total assets is 0.068.
13. Other measures of branching, such as the percentage of branches within 30 miles or the
percentage of branches out of state, were also not statistically signifcant.
14. We thank an anonymous referee for pointing this out. In addition to lagging the variables
one year, several other specifcations were estimated. Specifcally, we averaged the
independent variables over 2005-2007, 2000-2007 and 1994-2007. We also estimated the
regression lagging the independent variables four and fve years. In all cases, the results
remain consistent, particularly for the NOBRANCH variable.
15. In contrast, Huang (2008) does not fnd evidence of accelerated economic growth following
branching deregulation, so this conclusion is not conclusive.
References
Amel, D.F., Kennickell, A.B. and Moore, K.B. (2008), “Banking market defnition: evidence from
the survey of consumer fnances”, Financial and Economic Discussion Series No. 35, Federal
Reserve Board, Washington, DC.
Amel, D.F. and Starr-McCluer, M. (2002), “Market defnition in banking: recent evidence”, The
Antitrust Bulletin, Vol. 47 No. 1, pp. 63-89.
Aubuchon, C.P. and Wheelock, D.C. (2010), “The geographic distribution and characteristics of US
bank failures”, 2001-2010: do bank failures still refect local economic conditions? Federal
Reserve Bank of St. Louis Review, September/October, pp. 395-415.
Baldwin, W. (2011), “Who needs branches?” Forbes, pp. 48-49.
Brevoort, K.P., Holmes, J.A. and Wolken, J.D. (2010), “Distance Still Matters: The Information
Revolution in Small Business Lending and the Persistent Role of Location, 1993-2003,”
Board of Governors of the Federal Reserve System, Finance and Economics Discussion
Series, 2010-08.
Bucks, B.K., Kennickell, A.B., Mach, T.L. and Moore, K.B. (2009), “Changes in US family fnances
from2004 to 2007: evidence fromthe survey of consumer fnance”, Federal Reserve Bulletin,
pp. A1-A56.
149
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Calomiris, C.W. (1990), “Is deposit insurance necessary? A historical perspective”, Journal of
Economic History, Vol. 50 No. 2, pp. 283-295.
Calomiris, C.W. and Mason, J.R. (2003), “Fundamentals, panics, and bank distress during the
depression”, American Economic Review, Vol. 93 No. 5, pp. 1615-1647.
Carlson, M. (2004), “Are branch banks better survivors? Evidence from the depression era”,
Economic Inquiry, Vol. 42 No. 1, pp. 111-126.
Carlson, M., and Mitchener, K.J. (2006), “Branch banking, bank competition, and fnancial
stability,” Journal of Money, Credit and Banking, Vol. 38 No. 5, pp. 1293-1328.
Carlson, M. and Mitchener, K.J. (2007), “Branch banking as a devise for discipline: competition and
bank survivorship during the great depression”, NBER Working Paper No. 12938,
available at: www.nber.org/papers/w12938 (accessed 2 April 2011).
Cohen, A. and Mazzeo, M.J. (2010), “Investment strategies and market structure: an empirical
analysis of bank branching decisions”, Journal of Financial Services Research, Vol. 38,
pp. 1-21.
Collins, C.W. (1926), The Branch Banking Question, The Macmillan Company, New York, NY.
DeYoung, R. (1999), “Birth, growth, and life or death of newly chartered banks”, Federal Reserve
Bank of Chicago, Economic Perspectives, pp. 18-35.
Dick, A.A. (2006), “Nationwide branching and its impact on market structure, quality, and bank
performance”, The Journal of Business, Vol. 79 No. 2, pp. 567-592.
Heitfeld, E.A. and Prager, R.A. (2004), “The geographic scope of retail banking markets”, Journal
of Financial Services Research, Vol. 25, pp. 37-55.
Hendrickson, J.M. (2010), “The interstate banking debate: a historical perspective,” Academy of
Banking Studies Journal, Vol. 9 No.2, pp. 95-130.
Hendrickson, J.M. and Nichols, M.W. (2011), “The lights went out in georgia: explaining
commercial bank failures in the 21st century”, Academy of Business Journal, Vol. 1,
pp. 37-67.
Huang, R.R. (2008), “Evaluating the real effect of bank branching deregulation: comparing
contiguous counties across US state borders”, Journal of Financial Economics, Vol. 87,
pp. 678-705.
Jayaratne, J. and Strahan, P.E. (1998), “Entry Restrictions, industry evolution, and dynamic
effciency: evidence from commercial banking,” Journal of Law and Economics, Vol. 41,
pp. 239-273.
Johnson, C.A. and Rice, T. (2008), “Assessing a decade of interstate bank branching,” Washington
and Lee Law Review, Vol. 65, pp. 73-127.
Kolari, J., Glennon, D., Shin, H. and Caputo, M. (2002), “Predicting large US commercial bank
failures”, Journal of Economics and Business, Vol. 54 No. 1, pp. 361-387.
Kupiec, P.H. and Ramirez, C.D. (2013), “Bank failures and the cost of systemic risk: evidence from
1900 to 1930”, Journal of Financial Intermediation, Vol. 22, pp. 285-307.
Lu, W. and Whidbee, D.A. (2013), “Bank structure and failure during the fnancial crisis”, Journal
of Financial Economic Policy, Vol. 5 No. 2.
Mishkin, F.S. and Strahan, P.E. (1999), “What will technology do to fnancial structure?”, Working
Paper 6892 at the National Bureau of Economic Research, available at: www.nber.org/
papers/w6892 (accessed 2 March 2011)
Mitchener, K.J. (2005), “Bank supervision, regulation, and instability during the great depression”,
Journal of Economic History, Vol. 65 No. 1, pp. 152-185.
JFEP
6,2
150
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Mitchener, K.J. (2007), “Are prudential supervision and regulation pillars of fnancial stability?
evidence from the great depression”, The Journal of Law & Economics, Vol. 50 No. 2,
pp. 273-302.
Petersen, M.A. and Rajan, R.G. (2002), “Does distance still matter? The information revolution in
small business lending”, The Journal of Finance, Vol. 57 No. 6, pp. 2533-2570.
Ramirez, C.D. (2003), “Did branch banking restrictions increase bank failures? evidence from
Virginia and West Virginia in the late 1920s”, Journal of Economics and Business, Vol. 55
No. 4, pp. 331-352.
Robbins, E. (2006), “Location, location, location: has electronic banking affected the importance of
bank location?”, Financial Industry Perspectives, Federal Reserve Bank of Kansas City,
pp. 1-12.
Sprague, O.M.W. (1903), “Branch banking in the United States”, Quarterly Journal of Economics,
Vol. 17 No. 2, pp. 242-260.
Stiroh, K.J. and Strahan, P.E. (2003), “Competition dynamics of deregulation: evidence from US
banking”, Journal of Money, Credit, and Banking, Vol. 35 No. 5, pp. 801-828.
Strahan, P.E. (2003), “The real effects of US banking deregulation”, Review, Federal Reserve Bank
of St. Louis, July/August, pp. 111-128.
The Economist (2012a), “Counter revolution; consumer banking”, The Economist, 19 May, p. 18.
The Economist (2012b), “Withering away: branches”, The Economist, 19 May, p. 8.
Wheelock, D.C. (2011), “Have acquisitions of failed banks increased the concentration of US
banking markets?”, Review, Federal Reserve Bank of St. Louis, May/June, pp. 155-168.
Further reading
Berger, A.N. (2003), “The economic effects of technological progress: evidence from the banking
industry”, Journal of Money, Credit and Banking, Vol. 35 No. 2, pp. 141-176.
Berger, A.N., Demirguc-Kunt, A., Levine, R. and Haubrich, J.G. (2004), “Bank concentration and
competition: an evolution in the making”, Journal of Money, Credit, and Banking, Vol. 36
No. 3, pp. 433-451.
Calem, P.S. and Nakamura, L.I. (1998), “Branch banking and the geography of bank pricing”, The
Review of Economics and Statistics, Vol. 80 No. 4, pp. 600-610.
Calomiris, C.W. and Mason, J.R. (2000), “Causes of US bank distress during the great depression”,
Working Paper 7919, National Bureau of Economic Research.
Degryse, H. and Ongena, S. (2005), “Distance, lending relationships, and competition”, The Journal
of Finance, Vol. LX No. 1, pp. 231-266.
Harvey, J. and Spong, K. (2007), “The changing banking structure: what expansion strategies are
community banks adopting?”, Financial Industry Perspectives, Federal Reserve Bank of
Kansas City, pp. 1-15.
Kroszner, R.S. and Strahan, P.E. (2007), “Regulation and deregulation of the US banking industry:
causes, consequences, and implications for the future NBERbook in progress”, available at:
www.nber.org/books_in_progress/econ-reg/kroszner-strahan9-26-07.pdf (accessed 25 April).
Corresponding author
Jill M. Hendrickson can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
151
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_723624583.pdf
The purpose of this paper is to examine the impact of bank branch location on the likelihood
of bank failure during the most recent financial crisis
Journal of Financial Economic Policy
Bank branch location and stability during distress
J ill M. Hendrickson Mark W. Nichols Daniel R. Fairchild
Article information:
To cite this document:
J ill M. Hendrickson Mark W. Nichols Daniel R. Fairchild , (2014),"Bank branch location and stability during
distress", J ournal of Financial Economic Policy, Vol. 6 Iss 2 pp. 133 - 151
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-07-2013-0026
Downloaded on: 24 January 2016, At: 21:49 (PT)
References: this document contains references to 43 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 201 times since 2014*
Users who downloaded this article also downloaded:
Wenling Lu, David A. Whidbee, (2013),"Bank structure and failure during the financial crisis", J ournal of
Financial Economic Policy, Vol. 5 Iss 3 pp. 281-299http://dx.doi.org/10.1108/J FEP-02-2013-0006
Vighneswara Swamy, (2014),"Testing the interrelatedness of banking stability measures", J ournal of
Financial Economic Policy, Vol. 6 Iss 1 pp. 25-45http://dx.doi.org/10.1108/J FEP-01-2013-0002
Faten Ben Bouheni, (2014),"Banking regulation and supervision: can it enhance stability in Europe?",
J ournal of Financial Economic Policy, Vol. 6 Iss 3 pp. 244-269http://dx.doi.org/10.1108/J FEP-11-2013-0059
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Bank branch location and
stability during distress
Jill M. Hendrickson
Department of Economics, University of St. Thomas, St. Paul, Minnesota, USA
Mark W. Nichols
Department of Economics, University of Nevada Reno, Reno, Nevada, USA, and
Daniel R. Fairchild
Department of Economics, University of St. Thomas, St. Paul, Minnesota, USA
Abstract
Purpose – The purpose of this paper is to examine the impact of bank branch location on the likelihood
of bank failure during the most recent fnancial crisis.
Design/methodology/approach – This paper estimates the probit regression to identify the causes
of bank failures and attempts to determine the role of branch location in bank performance.
Findings – Using data from failed and surviving banks in Georgia and Florida, this paper fnds that
diversifying the balance sheet and operating in more competitive markets reduced failure rates, but
branching intensity, measured by number of branches and distance of branches from the home offce
did not signifcantly reduce the probability of failure. This suggests that, at least in today’s market, it is
not important to bank stability to have a branching network a signifcant distance fromthe home offce.
Originality/value – This paper carefully considers the role of branch location in the likelihood of
bank failure during fnancial distress. As such, it contributes to the historical policy debate regarding
regulation prohibiting or minimizing banks’ ability to branch. It also contributes to our understanding
of how banks structure their branching networks in the contemporary banking environment.
Keywords Financial crisis, Branch banking, Bank failures
Paper type Research paper
The USAhas a unique history with branch banking. Since late in the eighteenth century,
during our earliest experience with commercial banking, banks have frequently been
either banned from, or limited in, their ability to branch. From the beginning, there has
existed a policy struggle between those who wanted more branching and those who
wanted to preserve a systemof unit (non-branching) banks. For more than two hundred
years, the policy battle played out until the 1994 passage of the Riegle-Neal Interstate
Banking and Branching Effciency Act (IBBEA). This established the rights to
interstate branching and expanded, in many cases, intrastate branching.
Banks responded to the 1994 legislation by opening increasingly more branches
(Johnson and Rice, 2008). Indeed, the branch system, for many banks, became the model
for growth and stability (Dick, 2006). However, in the wake of the most recent fnancial
crisis, some scholars argued that banks failed to branch far enough fromthe home offce,
which is why they were vulnerable when the fnancial system became fragile
(Aubuchon and Wheelock, 2010). Furthermore, there is a growing group that believes
JEL classifcation – E44, G21, G28
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
Bank branch
location
133
Journal of Financial Economic Policy
Vol. 6 No. 2, 2014
pp. 133-151
©Emerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-07-2013-0026
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the branching system is outdated because of how technology has changed the bank
customers’ interaction with fnancial institutions. Baldwin (2011) argues that
technology, including mobile applications and younger generations of savers already
comfortable with new technology may utilize technology as a substitute for
branching[1]. Last year an article in The Economist (The Economist, 2012b, “Withering
Away”) argued that while branching has enjoyed tremendous growth in the twenty-frst
century, the fnancial crisis and resulting newregulatory environment has signifcantly
compromised bank revenue, making branches unproftable. Furthermore, it is expected
that as bank customers become increasingly comfortable with online services, such as
tax returns, they will embrace electronic banking and so reduce branch use. The article
predicts that there will be far fewer and more effcient branches in the future.
Certainly, many banks are more effcient and diversifed as a result of expanded
branching opportunities. However, some scholars argue that banks have not taken full
advantage of opportunities to geographically expand and so remain relatively
undiversifed. For example, since 2008, close to 400 commercial banks have failed in the
USA. In trying to understand this wave of bank failures, Aubuchon and Wheelock
(2010) argue that banks are not as diversifed as they should be, which led, in part, to
their failure. While Aubuchon and Wheelock (2010) recognize that banks are branching
in record numbers, these authors point out that they are not branching at a signifcant
distance from their home offce. In other words, the aggregate picture may refect that
banks are branching in greater numbers, but are they branching at meaningful
distances? Do the aggregate branching data conceal important trends? The point is, if
the branching is simply within a short radius fromthe home offce, then the full benefts
of branching (diversifcation, in particular) may not be captured.
The purpose of our research is two-fold. First is to determine the extent to which
commercial banks are branching in the post-1994 era by disaggregating the national
data. We carefully consider the branching behavior of a sample of banks by paying
particular attention to the distance and size of the branch from the home offce. Most
banks in the USA that utilize a branch network do not engage in interstate branching.
Indeed, ?90 per cent of all US commercial banks branch exclusively within state or do
not branch at all. Our bank sample contains both failed and surviving banks in Florida
and Georgia. Most of these banks are not engaged in interstate branching. Thus, our
analysis is of the behavior of banks engaged primarily in intrastate branching with
limited interstate-branching activity. Second, empirical methods are used to determine if
the distance and number of branches impact the probability of bank failure. Are the
branching patterns at failed banks different from surviving banks? Are Aubuchon and
Wheelock (2010) correct that failed banks have not branched at signifcant distances
fromthe home offce? To the author’s knowledge, no studies exist which link, at the frm
level, branch distance data with bank failures.
This paper is organized as follows. The next section of the paper briefy reviews recent
branch activity in the USA. The second section of the paper provides a reviewof the salient
literature. The thirdsectionof the paper explains our data, banksample, as well as empirical
methods. Results are found in section four and the fnal section concludes.
1. Contemporary branching activity in the USA
How have banks responded to the opportunities to bank and branch more freely?
Certainly, if there are signifcant benefts to branching, we would expect to see more
JFEP
6,2
134
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
branching per banking institution than in the past. This appears to be the case. As
shown in Dick (2006), both coverage and branch density has increased following the
deregulation of interstate branching. Between 1997 and the end of 2012, there has been
? 37 per cent increase in the aggregate number of branches. The number of branches
per bank increased from5.72 in 1995 to 13.73 branches per bank in 2012 as illustrated in
Figure 1. Furthermore, many of these branches are interstate in nature. As of June 30,
1997, the date that the IBBEA became effective, there were 8,876 interstate branches
throughout the country[2]. By June 30 2012, there were 43,330 interstate branches in the
USA – a 388 per cent increase in the aggregate number of branch locations operating
across state lines (see Figure 2). Thus, the degree to which banks are using branch
offces has steadily increased since the 1994 passage of the IBBEA.
Figure 3 illustrates the growth rate of branches per commercial bank in the USA. It is
interesting that the rate of growth was highest in the years prior to the IBBEA passage
in 1994, and then spiked again in the years prior to the 1997 fnal implementation of the
act. Given the timing of these growth spurts, most of this must have been intrastate
branching. It is possible that incumbent banks anticipated the passing of the IBBEAand
wanted to act prior to policy change. One can imagine two motivating factors for
0
2
4
6
8
10
12
14
16
Note: The number of branches per bank is the ratio of the total number of
branches to the total number of US commercial banks
Source: FDIC at www2.fdic.gov
Figure 1.
Number of branches per
US commercial bank
Source: FDIC at www2.fdic.gov
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
1994 1997 2011 2012
Figure 2.
Total number of interstate
branches
135
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
intrastate branch expansion prior to these dates. First, the existing banks may have
wanted to capture additional market share before the increased competition from
out-of-state branches. Second, and related to the frst, banks may have expanded their
branch network as a strategy of deterring future entry. Recent work by Cohen and
Mazzeo (2010) fnd empirical support for both hypotheses. The empirical work of these
scholars fnds evidence that competition from banks with multiple branches induces
additional branching from other institutions. Furthermore, they fnd that incumbent
banks are able to discourage entry by increasing their own branch network. The
increased growth rate in the 1990s, shown in Figure 3, is plausibly consistent with the
Cohen and Mazzeo (2010) fndings. The increased growth rate in 2006 may refect both
intrastate and interstate branch expansion to capture the growth in residential real
estate.
As mentioned above and illustrated in Figure 2, the total number of interstate
branches has grown signifcantly. However, most commercial banks do not engage in
interstate branching – as of September 30 2012, 566 banks, or 9.17 per cent of all
commercial banks branch across state lines[3]. This means, of course, that almost 91 per
cent do not branch outside of the state in which they are headquartered. Table I
illustrates that even within the small population of banks engaged in interstate
branching the fve largest banks account for almost half of all interstate branching in the
USA. At the other end, some very small banks also engage in interstate branching, but
typically these banks only have one branch[4]. The average size of the 566 banks
currently branching across state lines is just ? $18 billion in assets. As a point of
comparison, the average size of a commercial bank in the USAis just over $13 billion in
assets[5]. Thus, interstate branching banks are larger, on average, than the average of
the entire population of banks. Not surprisingly, then, the few banks that do engage in
extensive interstate branching hold the vast majority of deposits. In 24 of the states,
more than half of the bank deposits are held at a bank from outside of the state. To
Source: FDIC at www2.fdic.gov
Note: The rate of growth is calculated from the ratios in Figure 1
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
Figure 3.
Rate of growth in the
number of branches per
US commercial bank
JFEP
6,2
136
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
illustrate, consider, for example, that in Maryland, 79.2 per cent of bank deposits are
interstate deposits[6]. In only 11 states are interstate deposits ?25 per cent of the total.
Thus, while most banks do not branch across state lines, the few that do so extensively
have captured, in many cases, the majority of deposits within the state.
The post-IBBEA branching landscape is one with many more branch locations
despite corresponding to a time in which bank customers increasingly utilize
technological and electronic banking. However, most banks continue to operate their
branches within their headquartered state. Indeed, almost 91 per cent of US commercial
banks do not engage in interstate branching. Of the population of banks that do engage
in interstate branching, a small portion of them constitute the majority of interstate
branches.
Thus, interstate branching in the USA is dominated by a relatively small number of
commercial banks. This may be interpreted to illustrate that the IBBEA provision to
allow for interstate branching was not widely embraced across a signifcant number of
banks.
2. Existing literature
There are two felds of literature that inform this research. First is the work that
considers the ongoing debate regarding the benefts of branching and the effect of
branching restrictions on bank stability. Second, there is great interest among
regulators and policy makers on the impact that branching has on the nature and
defnition of banking markets. Certainly, this is of great interest because of antitrust and
other fnancial policy issues. Both of these facets of the literature are briefy reviewed
here.
2.1 The stabilizing impact of branching
During the long period of US commercial banking in which both interstate banking and
branching were restricted, there were many debates about the merits of such a
regulatory policy[7]. Fromnearly the beginning there were scholars, bankers and policy
makers who argued that a branching systemwould be more stable. Branching is said to
bring about fnancial stability through two channels: frst, through increased
Table I.
Interstate branching at
commercial banks by
bank size: September
30, 2012
Bank name
Total assets
($000)
Number of states
(including DC)
Number of
interstate offces
Percentage of
total interstate
offces in the USA
(per cent)
Largest banks
JPMorgan Chase Bank 1,850,218,000 26 5,366 12.38
Bank of America 1,448,273,067 36 5,434 12.54
Citibank 1,365,026,000 17 1,058 2.44
Wells Fargo 1,218,796,000 41 6,290 14.51
US Bank 342,627,272 28 2,830 6.53
Total 20,978 48.40
Notes: All data are as of September 30, 2012, unless otherwise noted
Source: Individual bank data are from the Institutional Directory and total branching data are from
both the Institutional Directory and Historical Statistics on Banking all from www2.fdic.gov
137
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
diversifcation, and second, through increased competition. Scholars and bankers long
argued against these restrictions, claiming that they kept banks from diversifying their
balance sheets which, in turn, made themmore vulnerable to local economic downturns.
Another prevalent argument was that the prohibition on interstate banking and
branching created local monopolies in banking and allowed ineffcient banks to remain
in business due to reduced competition. Most scholarship more than ten years old uses
aggregate data and typically does not distinguish between the diversifcation and
competition channel. Indeed, the vast majority of the established scholarship focuses
primarily on how branching may be stabilizing through diversifcation. More recent
scholarship often relies on frm level data and attempts to distinguish between the two
channels through which branching may bring about stability.
Many scholars have studied bank failures during the Great Depression to understand
the role branching played. Some of this work relied on aggregate level data and typically
fnds that branching reduces failure rates. For example, Mitchener (2005) uses
county-level data to test whether regulatory and supervisory differences impacted bank
stability during the Great Depression. Of interest to this study is his fnding that bank
failure rates were higher in counties located in states that prohibited branching.
Mitchener (2007) studies bank failures during the Great Depression and fnds that
branching was stabilizing because it introduced competition and forced ineffcient
banks out of the market either through closure or merger or because it enabled banks to
reduce risk via diversifcation. Calomiris (1990) studies the antebellumand 1920s eras of
commercial banking. During these historical periods, states experimented with state
deposit insurance as well as state-level branching laws. His empirical fndings suggest
that bank-asset growth was signifcantly higher in states that permitted branching. At
the same time, banks in branching states failed at a lower rate than banks in unit
banking states.
The scholarship on the Great Depression that relies on bank level data, perhaps
somewhat surprisingly, tends to be at odds with the aggregate data studies. Rather than
fnd branching to enhance stability, these studies fnd branching may lead banks to take
on additional risk. For example, Carlson (2004) uses bank-level data to determine
whether banks that branched were more likely to survive during the Great Depression.
His fnding, which is surprising because it goes against conventional wisdom, is that
branch banks are less likely to survive than unit banks. He argues that this does not
mean that branching is destabilizing but, rather, that branch bankers made decisions to
increase returns rather than reduce risk. Calomiris and Mason (2003) also fnd that,
during the Great Depression, banks that were members of the Federal Reserve and
engaged in branching failed sooner than unit banks. Much like Carlson, these authors
offer that branching banks may have been more diversifed which permitted them to
take on greater risks. Furthermore, Calomiris and Mason argue that while branching
can provide some diversifcation benefts when there are sector-specifc downturns, it
cannot offer protection when the downturn is across many sectors. In contrast, Ramirez
(2003) utilizes bank-level data and fnds that branching is stabilizing through increased
diversifcation in the years just prior to the Great Depression.
A second body of literature focuses on the competitive channel through which
branching may impact stability. One perspective argues that branching would enhance
competition and stability because weaker banks would be driven out of the market
leaving behind a more stable banking system (see, for example, Carlson and Mitchener,
JFEP
6,2
138
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
2006, Jayaratne and Strahan, 1998, Stiroh and Strahan, 2003). This literature suggests
that while competition reduces the profts of surviving banks, the weaker and ineffcient
banks are driven out of the market, leaving a stronger set of banks. Related, Carlson and
Mitchener (2007) fnd that increasing competition drives incumbent banks to improve
effciency and proftability. In contrast, others (see Collins (1926) or Sprague (1903)) have
argued that branching would create monopoly banking markets. The argument is that
branching banks would drive the smaller banks out of market creating “monopolies” of
one or two larger banks. This perspective was more prevalent historically and has
garnered less empirical support. Consequently, the current state of the literature tends to
favor the view that branching increases competition which, in turn, enhances bank
stability.
Only more recent scholarship on branching attempts to identify the specifc channel
through which branching may bring about stability. Carlson and Mitchener (2006 and
2007) have been the most deliberate in trying to tease out of empirical analysis, whether
it is through diversifcation or through increased competition that branch systems are
more stable. Carlson and Mitchener (2006) fnd that the increase in competition was
more important in bringing about stability during the Great Depression than was
diversifcation. Using frm-level data for banks in California, Carlson and Mitchener
(2007) fnd evidence that branch banking during the Great Depression forced existing
unit banks to be more effcient as a result of competition from branching banks. In this
way, branching is said to produce an externality: improved performance at unit banks to
enhance overall bank stability.
2.2 Banking market literature
Scholars have become increasingly interested in how technology and policy changes,
such as the IBBEA, impact the structure of banking. Such structural issues have
important antitrust implications since regulators and antitrust evaluations require
accurate defnitions of “a banking market”. Prior to the IBBEA, the standards used by
regulators and policy makers to defne a banking market were largely consistent with
statistical areas. Amel and Starr-McCluer (2002) and Mishkin and Strahan (1999)
indicate a “local banking market” is defned for both research and policy purposes as the
metropolitan statistical area (MSA) or non-MSA counties particularly for bank merger
purposes. Amel and Starr are more specifc with regards to rural areas and explain that
the market is typically assumed to be the size of one or two counties. Thus, the existing
standard is to defne a bank market to be the MSA or rural counties.
With the passage of the IBBEA, scholars are concerned that existing methods of
defning bank markets may no longer be appropriate. Some empirical research suggests
that banking markets remain local for both depositors and small business (see Heitfeld
and Prager, 2004). Using data from the 1998 Survey of Consumer Finances (SCF), Amel
and Starr-McCluer (2002) fnd that the average consumer banks within a two-mile radius
from their home. Most recently, Amel et al. (2008) analyzed data from the 2008 SCF and
fnd that consumers bank within a four-mile radius from the bank which suggests that
banking markets are still local. They defne a local banking market to be a 30-mile
radius, but admit this is somewhat of an arbitrary defnition.
Also using the SCF data, Robbins (2006) hypothesizes that as consumers
increasingly embrace electronic banking technology, the location of the bank and its
branches may be less important. Indeed, in 2007, 49.4 per cent of families reported
139
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
using electronic banking (Bucks et al. (2009)). However, 2007 survey data indicate
that consumers are increasingly utilizing electronic banking products, but still
value location more than any other factor in choosing a bank[8]. In fact, 45.9 per cent
of survey respondents identifed bank location as the most important reason for choosing a
bank. The secondmost important reasonwas identifedbyonly16.2per cent of respondents.
Finally, location appears to be more important now, as more respondents chose
location over past survey results (Bucks et al., 2009). Despite the benefts of
electronic banking services, it appears depositors still want the convenience of
banks that are located relatively close. Like the work of Amel and Starr-McCluer
(2002), Amel et al. (2008) and Heitfeld and Prager (2004), this suggests that bank
markets remain local, at least for depositors, despite the increased use of electronic
banking services.
In contrast to those who fnd that banking markets remain local, Petersen and Ragan
(2002) fnd that the distance between the bank and the small business borrower is
increasing. Using data from the National Survey of Small Business Finance (NSSBF),
the authors fnd that the distance between the bank and the borrower has been
increasing since the 1970s and through 1993. This is caused by technology that has
allowed bankers to use information about the borrower to both make lending decisions
and to monitor loans, reducing the need for the borrower and lender to meet each other
in person. Brevoort et al. (2010) revisit the work of Petersen and Ragan by including
NSSBF data from 1998 and 2003. They fnd distance increased at a faster rate between
1993 and 1997, but halted and even decreased between 1998 and 2003. They interpret
this to mean that the market for small business lending remains local.
2.3 Contributions of this work
It appears that bankers were eager to expand their branch networks in the post-IBBEA
era as illustrated in Figures 1 and 2. The question is was the branching suffcient enough
to add stability to banking? Are banks that branch in further distances from the home
offce less likely to fail? Did banks diversify their balance sheets via branching and, in
the process, reduce the likelihood of failure? While there is existing literature that
considers the relationship between branching and the probability of failure, there is no
literature, to the authors’ knowledge, that consider the intensity of branching and the
probability of failure. Here, branching intensity is measured as both the distance of the
branch from the home offce as well as the number of branches in each bank
organization.
The hypothesis of Aubuchon and Wheelock (2010) is that banks failed to branch in
signifcant ways so as to stabilize banking in recent years. That is, one possible
explanation for the large increase in the number of bank failures following the 2007-2008
fnancial crisis is that US commercial banks had an opportunity to engage in meaningful
interstate, or even intrastate, branching following the 1994 IBBEA, but that they chose
not to. This decision cost the banks because they remained less diversifed by not
expanding and/or their unwillingness to branch allowed ineffcient banks to remain
viable longer than had they faced competition from branch expansion. This empirical
question remains unresolved and the purpose of this paper is to try and fnd satisfactory
answers to this line of inquiry.
JFEP
6,2
140
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
3. Data and empirical analysis
To test whether banks are branching in meaningful distances and whether branch
behavior impacts failure, this work relies on bank-level, branch-level and MSA-level
data. Specifcally, we rely on a randomsample of failed and non-failed banks in Georgia
and Florida. While examining two states is a limiting factor, of the 337 commercial bank
failures between January 2008 and October 2011, Georgia (69 failures) and Florida (47
failures) account for 34 per cent of all failures in the country. These two states,
nonetheless, represent different experiences in the twenty-frst century. Florida
experienced signifcant home infation, followed by signifcant defation surrounding
the fnancial crisis. In contrast, Georgia did not see the home-price swings that Florida
experienced, but suffered even more bank failures than Florida following the crisis.
Thus, to include these two states is to capture both post-crisis experiences in the housing
market.
For each bank in the sample, we use balance sheet and income data for each quarter
between 2000 and 2011. Furthermore, we have annual branch activity data for each
bank, including how many branches the bank operated, the number of miles each
branch is from the home offce, the location of the branch, the type of services offered at
each branch and the size, measured by total deposits, of each branch. Bank-level data
come from the Federal Deposit Insurance Corporation and is available online. In
addition, MSA-level data on total population, housing prices, income and
unemployment are collected to account for local market conditions. All MSA-level
data are public information; population data are from the Census Bureau; home
indices are from the Federal Housing Finance Agency; real per capita personal
income is from the Bureau of Economic Analysis; and unemployment data come
from the Bureau of Labor Statistics.
The number of banks in our sample varies from year to year due to mergers,
openings, and, of course, failures. In 2007, the last year prior to the increase in failures
taking place during 2008-2011, we have 169 commercial banks in our sample, 96 of
which are in Georgia and 73 in Florida. Of those, 82 would go on to fail sometime
between 2008 and 2011, with 49 failures occurring in Georgia and 33 in Florida. Table II
describes and provides summary statistics for our key variables over the years
2000-2011.
3.1 Descriptive statistics for entire sample
Table II demonstrates that, of those banks with branches (approximately 77 per cent of the
sample), most have very fewbranches, and those branches are close to the home offce. The
average distance fromthe home offce is 17.47 miles. However, the median is 9.75 miles and
the ninetiethpercentile is only38 miles. Indeed, 82 per cent of branches are within30 miles of
the home offce, and only 0.43 per cent of branches are interstate. The average number of
branches is just over 3; the median is 2 and the ninetieth percentile 7.
3.2 Failed versus surviving sample of banks
Table III provides a comparison of means between failed banks and branch banks,
respectively. Columns two and three of Table III reveal several interesting differences
between failed and surviving banks. In terms of branching, for those banks that had
branches, failed banks had, on average, approximately one more branch (3.81 vs 2.97)
and branched at a greater distance fromthe home offce (21.7 miles vs 14.1). Also of note,
141
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
however, failed banks had a signifcantly greater proportion of unit banks (31.4 per cent)
than surviving banks (18.10 per cent).
Table III also demonstrates differences in risk taking between failed and surviving
banks. Specifcally, failed banks, while larger in terms of total assets, had a lower return
on assets and a lower capital to asset ratio. They also had a larger proportion of assets
in loans and lower proportion in securities.
Failed banks had a signifcantly higher proportion of real estate loans, but no
difference in past due residential mortgages. Failed banks had a lower proportion of
commercial and industrial loans and were also less diversifed[9]. The failed sample was
also much less experienced than the surviving banks, using age as a proxy for
experience (Carlson (2004)).
Table II.
Variable description and
descriptive statistics, all
banks, 2000-2011
Variable Defnition
Mean (standard
deviation)
Branch variables
Average miles Average distance of bank branches from main offce 17.470 (28.97)
Number of branches Average number of branches bank maintains 3.347 (5.80)
Average branch deposits Average deposits at braches (thousands of dollars) 19318.82 (21582.19)
Percentage of branches out of state Percentage of a bank’s branches that are out of state 0.436 (4.83)
Percentage of branches within 30
miles
Percentage of a bank’s branches that are within 30
miles of the main bank
82.033 (29.35)
Percentage of banks with no
branches
Percentage of banks that have no branches 23.168 (42.20)
Bank variables
Failure Dummy equal to 1 if the bank failed any time during
2000-2011
0.441 (0.496)
Deposit market share The bank’s share of total deposits of all FDIC-
insured banks within the bank’s headquarter city as
of June 30 for each year
24.78 (29.61)
Total assets Total assets of the bank (thousands of dollars) 258369.5 (370021)
Return on assets Net operating income after taxes and extraordinary
items divided by total assets
0.075 (2.09)
Loan-to-asset ratio Total loans divided by total assets 0.689 (0.135)
Security-to-asset ratio Total securities divided by total assets 0.154 (0.113)
Capital-to-asset ratio Total capital divided by total assets 0.109 (0.074)
Non-interest income-to-asset ratio Non-interest income divided by total assets 0.0040 (0.0047)
Real estate loans to total loans Real estate loans divided by total loans 0.807 (0.137)
Commercial and industrial loans to
total loans
Commercial and industrial loans divided by total
loans
0.125 (0.087)
Age Age of the bank, years 34.14 (34.91)
Derivatives to assets Total derivatives divided by total assets 0.0053 (0.038)
Core deposits to total deposits Core deposits divided by total deposits 0.768 (0.109)
Past due mortgages Past due residential mortgages divided by total
residential mortgages
0.017 (0.020)
Diversity Sum of squared ratios of business loans, real estate
loans, consumer loans and securities to total assets
0.392 (0.126)
MSA variables
Change in home price index Annual fourth quarter percentage change in house
price index in the bank’s MSA
3.59 (9.82)
Unemployment rate Unemployment rate in the bank’s MSA 5.66 (2.36)
Population Population of the city where bank located 97803.56 (169532)
Real per capita income Per capita income in the bank’s MSA, 2007 dollars 37352.95 (5591.96)
Georgia Dummy equal to 1 if bank is in Georgia 0.56 (0.496)
JFEP
6,2
142
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table III.
Comparison of means by
bank failure and branch
status, 2000-2011
V
a
r
i
a
b
l
e
F
a
i
l
e
d
b
a
n
k
s
(
n
?
7
2
6
)
S
u
r
v
i
v
i
n
g
b
a
n
k
s
(
n
?
8
9
5
)
D
i
f
f
e
r
e
n
c
e
:
f
a
i
l
e
d
-
s
u
r
v
i
v
i
n
g
(
t
-
s
t
a
t
i
s
t
i
c
)
N
o
b
r
a
n
c
h
b
a
n
k
s
(
n
?
3
8
9
)
B
r
a
n
c
h
b
a
n
k
s
(
n
?
1
2
3
3
)
D
i
f
f
e
r
e
n
c
e
:
n
o
b
r
a
n
c
h
-
b
r
a
n
c
h
(
t
-
s
t
a
t
i
s
t
i
c
)
B
r
a
n
c
h
v
a
r
i
a
b
l
e
s
A
v
e
r
a
g
e
m
i
l
e
s
2
1
.
7
3
1
4
.
1
0
7
.
6
3
*
*
*
(
4
.
9
4
)
N
u
m
b
e
r
o
f
B
r
a
n
c
h
e
s
3
.
8
1
2
.
9
7
0
.
8
4
*
*
*
(
2
.
7
6
)
A
v
e
r
a
g
e
b
r
a
n
c
h
D
e
p
o
s
i
t
s
1
9
3
0
0
.
9
3
1
9
3
3
2
.
9
4
?
3
2
.
0
1
(
?
0
.
0
2
)
P
e
r
c
e
n
t
o
f
b
r
a
n
c
h
e
s
o
u
t
o
f
s
t
a
t
e
0
.
4
5
8
0
.
4
4
7
0
.
0
1
0
8
(
0
.
0
4
)
P
e
r
c
e
n
t
a
g
e
o
f
b
r
a
n
c
h
e
s
w
i
t
h
i
n
3
0
m
i
l
e
s
7
5
.
2
8
8
6
.
5
7
?
1
1
.
2
9
*
*
(
?
6
.
4
2
)
P
e
r
c
e
n
t
a
g
e
o
f
b
a
n
k
s
w
i
t
h
n
o
b
r
a
n
c
h
e
s
3
1
.
4
6
1
8
.
1
0
1
3
.
3
6
*
*
*
(
6
.
1
9
)
B
a
n
k
v
a
r
i
a
b
l
e
s
F
a
i
l
u
r
e
0
.
5
7
8
0
.
3
9
7
0
.
1
8
1
*
*
*
(
6
.
2
9
)
D
e
p
o
s
i
t
m
a
r
k
e
t
s
h
a
r
e
2
1
.
1
3
2
8
.
1
6
?
7
.
0
3
*
*
*
(
4
.
8
6
)
1
9
.
7
3
2
6
.
2
8
?
6
.
5
5
*
*
*
(
3
.
8
9
)
T
o
t
a
l
a
s
s
e
t
s
3
1
6
0
5
4
.
5
0
2
1
9
4
5
7
.
1
0
9
6
5
9
7
.
4
7
*
*
*
(
4
.
8
6
)
1
4
0
2
4
4
.
4
0
2
9
3
9
9
0
.
2
0
?
1
5
3
7
4
5
.
8
0
*
*
*
(
8
.
7
7
)
R
e
t
u
r
n
o
n
a
s
s
e
t
s
?
0
.
2
9
2
0
.
4
7
5
?
0
.
7
6
7
*
*
*
(
?
7
.
1
7
)
?
0
.
4
1
4
0
.
2
2
3
?
0
.
6
3
7
*
*
*
(
4
.
2
7
)
L
o
a
n
t
o
a
s
s
e
t
r
a
t
i
o
0
.
7
3
0
0
.
6
6
4
0
.
0
6
6
*
*
*
(
1
0
.
9
8
)
0
.
6
5
8
0
.
6
9
9
?
0
.
0
4
1
*
*
*
(
4
.
5
8
)
S
e
c
u
r
i
t
y
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
1
1
8
0
.
1
8
4
?
0
.
0
6
5
*
*
*
(
?
1
2
.
8
2
)
0
.
1
6
2
0
.
1
5
2
0
.
0
1
0
(
1
.
2
9
)
C
a
p
i
t
a
l
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
1
0
0
0
.
1
0
9
?
0
.
0
0
9
*
*
*
(
?
3
.
3
6
)
0
.
1
4
0
.
1
0
0
.
0
4
*
*
*
(
8
.
7
8
)
N
o
n
-
i
n
t
e
r
e
s
t
i
n
c
o
m
e
-
t
o
-
a
s
s
e
t
r
a
t
i
o
0
.
0
0
4
0
0
.
0
0
4
1
?
0
.
0
0
0
1
(
?
0
.
5
2
)
0
.
0
0
3
5
0
.
0
0
4
1
?
0
.
0
0
0
6
*
(
1
.
7
8
)
R
e
a
l
e
s
t
a
t
e
l
o
a
n
s
t
o
t
o
t
a
l
l
o
a
n
s
0
.
8
3
9
0
.
7
7
9
0
.
0
5
9
*
*
*
(
9
.
0
3
)
0
.
7
6
9
0
.
8
1
8
?
0
.
0
4
9
*
*
*
(
5
.
5
0
)
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
t
o
t
o
t
a
l
l
o
a
n
s
0
.
1
1
6
0
.
1
3
2
?
0
.
0
1
6
*
*
*
(
3
.
5
8
)
0
.
1
4
7
0
.
1
1
8
0
.
0
2
9
*
*
*
(
5
.
0
0
)
A
g
e
2
6
.
4
3
4
1
.
8
6
?
1
5
.
4
3
*
*
*
(
9
.
0
9
)
2
1
.
4
2
3
7
.
9
7
?
1
6
.
5
5
*
*
*
(
8
.
6
4
)
D
e
r
i
v
a
t
i
v
e
s
t
o
a
s
s
e
t
s
0
.
0
1
1
0
.
0
0
1
2
0
.
0
0
9
*
*
*
(
4
.
3
4
)
0
.
0
1
1
0
.
0
0
3
0
.
0
0
7
*
*
(
2
.
0
4
)
C
o
r
e
d
e
p
o
s
i
t
s
t
o
t
o
t
a
l
d
e
p
o
s
i
t
s
0
.
7
4
2
0
.
7
8
6
?
0
.
0
4
4
*
*
*
(
?
8
.
2
4
)
0
.
7
5
6
0
.
7
7
1
?
0
.
0
1
5
*
*
(
2
.
3
2
)
P
a
s
t
d
u
e
m
o
r
t
g
a
g
e
s
0
.
0
1
7
7
0
.
0
1
7
6
0
.
0
0
0
1
(
0
.
0
9
)
0
.
0
1
5
0
.
0
1
8
?
0
.
0
0
3
*
*
*
(
2
.
9
3
)
D
i
v
e
r
s
i
t
y
0
.
4
2
8
0
.
3
6
7
0
.
0
6
1
*
*
*
(
1
0
.
0
8
)
0
.
3
6
2
0
.
4
0
0
?
0
.
0
3
8
*
*
*
(
4
.
9
2
)
M
S
A
v
a
r
i
a
b
l
e
s
C
h
a
n
g
e
i
n
h
o
m
e
p
r
i
c
e
i
n
d
e
x
3
.
9
4
3
.
3
3
0
.
6
1
(
1
.
2
2
)
4
.
6
1
3
.
2
9
1
.
3
2
*
*
(
2
.
5
6
)
U
n
e
m
p
l
o
y
m
e
n
t
r
a
t
e
5
.
4
1
5
.
8
4
?
0
.
4
3
*
*
*
(
?
3
.
7
5
)
5
.
3
1
5
.
7
6
?
0
.
4
5
*
*
*
(
3
.
7
7
)
P
o
p
u
l
a
t
i
o
n
8
9
5
5
0
.
9
2
1
0
6
5
5
4
.
3
0
?
1
7
0
0
3
.
3
8
*
*
(
2
.
0
0
)
1
1
2
9
6
5
.
9
7
9
3
2
0
4
.
9
3
1
9
7
6
1
.
0
4
*
(
1
.
8
1
)
R
e
a
l
p
e
r
c
a
p
i
t
a
i
n
c
o
m
e
3
8
3
1
7
.
1
7
3
5
7
5
4
.
4
7
2
5
6
2
.
7
0
*
*
*
(
7
.
8
9
)
3
7
5
5
5
.
9
8
3
6
7
0
1
.
1
4
8
5
4
.
8
4
*
*
(
2
.
4
7
)
N
o
t
e
s
:
t
-
s
t
a
t
i
s
t
i
c
i
n
p
a
r
e
n
t
h
e
s
e
s
.
A
*
,
*
*
a
n
d
*
*
*
r
e
p
r
e
s
e
n
t
s
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
f
c
a
n
c
e
a
t
t
h
e
1
0
,
5
a
n
d
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
.
B
e
c
a
u
s
e
n
o
t
a
l
l
v
a
r
i
a
b
l
e
s
a
r
e
n
o
r
m
a
l
l
y
d
i
s
t
r
i
b
u
t
e
d
,
t
e
s
t
s
f
o
r
s
i
g
n
i
f
c
a
n
t
d
i
f
f
e
r
e
n
c
e
s
w
e
r
e
a
l
s
o
e
s
t
i
m
a
t
e
d
u
s
i
n
g
t
h
e
n
o
n
-
p
a
r
a
m
e
t
r
i
c
W
i
l
c
o
x
o
n
R
a
n
k
S
u
m
t
e
s
t
.
R
e
s
u
l
t
s
w
e
r
e
q
u
a
l
i
t
a
t
i
v
e
l
y
i
d
e
n
t
i
c
a
l
143
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In summary, Table III indicates that failed banks were more likely to be unit banks, but
those that did branch had more branches at a greater distance. Failed banks had lower
returns on assets, were less experienced and took on greater risk. Failed banks also had
a greater deposit market share, suggesting they operated in less competitive
environments.
3.3 Branching versus unit sample of banks
Given the greater proportion of unit banks that failed, further analysis of the distinction
between unit and branch banks is provided in the last three columns of Table III. Unit
banks have a greater proportion of failures (57.8 per cent) than branch banks (39.7 per
cent). Unit banks also had a signifcantly lower (and negative) return on assets.
Consistent with Carlson (2004), branch banks took on more risk, have lower capital to
asset and security to asset ratios, a higher loan to asset ratio, a greater percentage of
loans in real estate, a greater percentage of past due mortgages and less diversifcation.
However, unit banks had a greater proportion of derivatives to assets, a lower
proportion of core deposits, more commercial and industrial loans and less non-interest
income. Unit banks were also less experienced (younger) and had a lower share of total
deposits in the city where they operated.
The descriptive statistics reveal an interesting pattern of failure and branching.
Failed banks had more branches and those branches were, on average, at a greater
distance from the home offce. Branch banks, however, took on more risk based on
measures such as capital to asset ratio, proportion of total loans in real estate, past due
mortgages, and loan diversifcation. In this regard, the descriptive statistics are
consistent with the branching results found in the Great Depression by Carlson (2004).
At the same time, however, unit banks accounted for a greater percentage of failures,
were younger and more exposed to commercial and industrial loans. Both geographic
and balance sheet differences therefore appear to be correlated with failure.
3.4 Probit analysis
To better distinguish the importance of branching and geographic versus balance sheet
factors a probit regression of bank failure is estimated. We focus on the large number of
bank failures that took place between 2008 and 2011. Consequently, we examine
whether the branching and balance sheet characteristics of the bank in 2007 infuenced
the probability the bank failed sometime during 2008-2011[10].
Our probit model is as follows:
FAIL
i
? ?
0
? ?
1
NOBRANCH
i
? ?
2
AVGMILES
i
? ?
3
BRANCHNUM
i
? ?
4
LN(ASSET
i
) ? ?
5
DEPMKTSHARE
i
? B
6
DIVERSITY
i
? ?
7
RESTATETOLOAN
i
? ?
8
DERIVTOASSET
i
? ?
9
COREDEPTODEPOSIT
i
? ?
10
ROA
i
? ?
11
AGE
i
? ?
12
CHNGHOMEPINDEX
i
? ?
13
CITYPOP
i
? ?
14
UNEM
i
? ?
15
PCINC
i
??
16
GEORGIA
i
? ?
i
(1)
In equation (1), FAIL
i
is a dummy variable equal to one if bank i fails any time between
2008 and 2011. The independent variables are measured in 2007. NOBRANCH
i
is a
dummy equal to one if bank i had no branches in 2007. AVGMILES
i
is the average
JFEP
6,2
144
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
distance branches are fromthe home offce in 2007, and BRANCHNUM
i
is the number of
branches that bank i operates in 2007. The above capture the banks’ branching activity
and geographic diversifcation[11]. DEPMKTSHARE
i
is the bank’s share of total
deposits in 2007 of all Federal Deposit Insurance Corporation (FDIC)-insured banks
where the bank is headquartered. This is designed to capture the competitive
environment in which the bank operates, with higher deposit shares representing more
concentrated and hence less competitive markets[12]. LN (ASSET
i
) is the natural log of
the bank’s assets in 2007 and captures bank size. Other variables account for
balance-sheet differences between banks. DIVERSITY
i
, as defned in the study by
Kolari et al. (2002), is the sum of squared ratios of business loans, real estate loans,
consumer loans and securities to total assets. Higher values represent less
diversifcation. RESTATETOLOAN
i
is the ratio of real estate loans to total loans. ROA
i
is the return on assets (net income to asset ratio). DERIVTOASSET
i
is the dollar value
of the banks derivatives to total assets and COREDEPTODEPOSIT
i
is the ratio of core
to total deposits. A banker’s engagement with derivatives is included to capture the
trend in commercial banking of off-balance sheet activity, while the core deposits
represent the cheapest source of funds to the bank. AGE
i
is the age of the bank in 2007
and proxies for the experience of the bank. The remaining variables capture local
market conditions. CHNGHOMEPINDEX
i
is the change in the fourth quarter home price
index for the MSAwhere the home bank is locate, CITYPOP
i
is the population of the city
where the home bank is located; UNEM
i
is the unemployment rate of the MSAwhere the
bank is located; PCINC
i
is real per capita income; and GEORGIA
i
is a dummy if the bank
was in Georgia and is included to account for any state-specifc effects.
Probit coeffcient estimates are provided in Column two of Table IV. Most relevantly
for this study, when controlling for balance sheet, risk-taking, and local market
differences between banks, unit banks are signifcantly more likely to fail than their
branching counterparts. The coeffcient on NOBRANCH is statistically signifcant at
the 5 per cent level. Its marginal effect is 0.20, indicating banks without any branches are
20 per cent more likely to fail. Interestingly, while not branching signifcantly increases
the likelihood of failure, branching intensity, measured as both average distance from
the home offce and total number of branches is not a signifcant determinant of failure
controlling for risk taking and other factors[13].
The coeffcient on DEPMKTSHARE is positive and statistically signifcant, suggesting
that banks withgreater shares of a city’s total FDICinsureddeposits were more likelyto fail.
To the extent deposit market share refects a bank’s competitive environment; this result
suggests, consistent with Carlson and Mitchener (2006), that increased bank competition
reduces the probability of failure by weeding out weaker banks.
Consistent with Lu and Whidbee (2013), DIVERSITYis also statistically signifcant.
Because higher values represent less diversifcation, the results suggest that banks with
less diversifed balance sheets were more likely to fail. Banks with a greater proportion
of their assets in derivatives were also more likely to fail, but a greater proportion of
loans in real estate was not found to be signifcant. Thus, perhaps surprisingly, it was
not real estate loans that increased the probability of failure but, rather, lack of
diversifcation and derivative activity which explain failures during this period.
Furthermore, banks with a greater proportion of core deposits, a higher net income to
asset ratio, andmoreexperienced(older) bankswerelesslikelytofail. LuandWhidbee(2013)
fnd that de novo banks are more likely to fail during this same time period, which is
145
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table IV.
Probit regression results
for bank failure (banks
failing during 2008-2011)
V
a
r
i
a
b
l
e
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
m
e
a
s
u
r
e
d
i
n
2
0
0
7
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
l
a
g
g
e
d
o
n
e
y
e
a
r
B
r
a
n
c
h
v
a
r
i
a
b
l
e
s
N
o
b
r
a
n
c
h
e
s
(
N
O
B
R
A
N
C
H
)
0
.
7
6
1
*
*
(
0
.
3
6
0
)
0
.
5
3
8
*
(
0
.
3
1
6
)
A
v
e
r
a
g
e
m
i
l
e
s
(
A
V
G
M
I
L
E
S
)
?
0
.
0
0
2
4
(
0
.
0
0
6
5
)
?
0
.
0
0
2
8
(
0
.
0
0
5
7
)
N
u
m
b
e
r
o
f
b
r
a
n
c
h
e
s
(
B
R
A
N
C
H
N
U
M
)
0
.
0
0
4
9
(
0
.
0
2
4
0
)
0
.
0
0
1
2
(
0
.
0
2
4
)
B
a
n
k
b
a
l
a
n
c
e
s
h
e
e
t
v
a
r
i
a
b
l
e
s
D
e
p
o
s
i
t
m
a
r
k
e
t
s
h
a
r
e
(
D
E
P
M
K
T
S
H
A
R
E
)
0
.
0
1
4
8
*
*
(
0
.
0
0
6
4
)
0
.
0
1
5
4
*
*
(
0
.
0
0
6
)
D
i
v
e
r
s
i
t
y
(
D
I
V
E
R
S
I
T
Y
)
4
.
8
5
5
*
*
*
(
1
.
5
7
9
)
3
.
7
7
4
*
*
*
(
1
.
2
8
6
)
L
o
g
o
f
t
o
t
a
l
a
s
s
e
t
s
(
L
N
A
S
S
E
T
)
0
.
4
3
7
*
(
0
.
2
3
2
)
0
.
4
7
5
*
*
(
0
.
2
0
2
)
R
e
a
l
e
s
t
a
t
e
l
o
a
n
s
t
o
t
o
t
a
l
l
o
a
n
s
(
R
E
S
T
A
T
E
T
O
L
O
A
N
)
0
.
9
2
2
(
1
.
5
2
4
)
2
.
3
4
4
*
(
1
.
4
0
4
)
R
e
t
u
r
n
o
n
a
s
s
e
t
s
(
R
O
A
)
?
0
.
3
1
9
*
*
(
0
.
1
3
0
)
?
0
.
3
1
7
*
*
*
(
0
.
0
8
2
)
D
e
r
i
v
a
t
i
v
e
s
t
o
a
s
s
e
t
s
(
D
E
R
I
V
T
O
A
S
S
E
T
)
1
7
.
8
7
7
*
*
*
(
6
.
0
9
5
)
1
3
.
4
9
0
*
*
*
(
5
.
7
3
6
)
C
o
r
e
d
e
p
o
s
i
t
s
t
o
t
o
t
a
l
d
e
p
o
s
i
t
s
(
C
O
R
E
D
E
P
T
O
D
E
P
O
S
I
T
)
?
3
.
2
3
0
*
*
(
1
.
4
0
4
)
?
2
.
1
3
2
*
*
(
1
.
0
2
3
)
A
g
e
(
A
G
E
)
?
0
.
0
1
1
*
*
(
0
.
0
0
4
5
)
?
0
.
0
1
1
*
*
(
0
.
0
0
4
8
)
M
S
A
v
a
r
i
a
b
l
e
s
C
h
a
n
g
e
i
n
c
o
m
e
p
r
i
c
e
i
n
d
e
x
(
C
H
N
G
H
O
M
E
P
I
N
D
E
X
)
?
0
.
1
0
4
*
*
(
0
.
0
5
0
)
?
0
.
0
2
3
(
0
.
0
1
6
)
P
o
p
u
l
a
t
i
o
n
,
t
h
o
u
s
a
n
d
s
(
C
I
T
Y
P
O
P
)
?
0
.
0
0
0
1
(
0
.
0
0
0
0
8
)
?
0
.
0
0
0
1
(
0
.
0
0
0
0
7
)
U
n
e
m
p
l
o
y
m
e
n
t
(
U
N
E
M
)
?
0
.
6
9
5
*
*
(
0
.
3
1
6
)
?
0
.
2
8
4
*
*
*
(
0
.
0
5
4
)
R
e
a
l
p
e
r
c
a
p
i
t
a
i
n
c
o
m
e
,
t
h
o
u
s
a
n
d
s
(
P
C
I
N
C
)
0
.
0
0
0
5
(
0
.
0
0
0
0
2
)
0
.
0
0
0
5
(
0
.
0
0
0
0
2
)
G
e
o
r
g
i
a
(
G
A
)
0
.
9
6
8
*
*
(
0
.
4
5
8
)
0
.
2
2
3
(
0
.
3
2
0
)
P
s
e
u
d
o
R
2
0
.
3
7
7
0
.
3
1
4
L
o
g
l
i
k
e
l
i
h
o
o
d
?
7
2
.
9
5
?
1
9
4
.
2
1
N
u
m
b
e
r
o
f
b
a
n
k
s
1
6
9
1
6
8
N
o
t
e
s
:
R
e
p
o
r
t
e
d
p
r
o
b
i
t
c
o
e
f
f
c
i
e
n
t
s
.
R
o
b
u
s
t
s
t
a
n
d
a
r
d
e
r
r
o
r
s
,
c
l
u
s
t
e
r
e
d
o
n
b
a
n
k
s
,
i
n
p
a
r
e
n
t
h
e
s
e
s
.
A
*
,
*
*
a
n
d
*
*
*
r
e
p
r
e
s
e
n
t
s
t
a
t
i
s
t
i
c
a
l
s
i
g
n
i
f
c
a
n
c
e
a
t
t
h
e
1
0
,
5
a
n
d
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
JFEP
6,2
146
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
consistent with the fnding that younger banks are more likely to fail. Hendrickson and
Nichols (2011) studybankfailures inGeorgiaandalsofndthat these banks were more likely
to fail if they were young. This post-crisis literature is all consistent with DeYoung (1999),
who found that banks were most vulnerable to failure between the ages of three and fve
years. It is likelythat inbothGeorgia andFlorida manyof the banks hadenteredthe market
while house prices were rising and that these young banks were not yet in a position to be
able to weather the aftermath of the crisis.
Local market conditions also play some role. Rising home values are negatively
correlated with failure, and banks in Georgia were more likely to fail than banks in
Florida. Unemployment, however, is unexpectedly negative and signifcant, although
consistent with the results in Table III which shows a lower unemployment rate for
failed banks relative to surviving banks. Population and per capita income are not
signifcant determinants of bank failure in this study.
As a robustness check we also measure our independent variables lagged one year
rather than their 2007 variables because bank and economic conditions in 2007 may
have a different infuence on banks that failed in 2008 compared to those that failed later
in the sample[14]. These results are reported in column 3 of Table IV. The results and
conclusions are generally consistent across specifcations.
As a fnal robustness check, other control variables, including the loan-to-asset ratio,
security-to-asset ratio, average branch deposits, number of banks in the MSA, number
of non-banks in the MSA, past due mortgages, non-current loans to total loans,
mortgage-backed securities to total assets and non-interest income to assets, were
included as independent variables. These were not statistically signifcant and did not
change the statistical signifcance of the NOBRANCHcoeffcient or the insignifcance of
branch number and distance variables.
In summary, whether examining a simple comparison of means or estimating a
probit regression that controls for balance sheet and geographic differences between
banks, unit banks were more likely to fail during the most recent fnancial crisis than
banks that branched. Diversifying the balance sheet and operating in more competitive
markets reduced failure rates, but branching intensity, measured by number of
branches and distance of branches fromthe home offce, did not signifcantly reduce the
probability of failure.
4. Conclusion
This paper is concerned with the relationship between branch activity and intensity and
bank failure. The results suggest that unit banks are more likely to fail than banks that
have branches. This is important both from the perspective of the banker and from the
perspective of the policy maker. For the individual banker, this fnding suggests that
decisions to branch lessen the probability of failure. Branching, despite technological
advancement, is a strategy that can bring about stability to the banker. From a
policy-making standpoint, if branching enhances stability, policy should remain open to
the creation of branch networks. Furthermore, there is evidence to suggest that bank
failures reduce industrial production and economic growth (Kupiec and Ramirez, 2013)
and other evidence to suggest that branch networks reduce state-level business-cycle
volatility (Strahan, 2003)[15]. If branching can reduce failures, there are important
macroeconomic gains in stability to be had of pursuing a policy that encourages
branching.
147
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
While branching generally is found to be more stabilizing than not branching, branch
intensity does not appear to impact the probability of failure. The aggregate data on
bank branching suggest that ? 10 per cent of all commercial banks have taken
advantage of the policy change in 1994 that allows for interstate branching, but that
overall branching, including intrastate, is growing signifcantly. However, Aubuchon
and Wheelock (2010) were concerned that banks were not branching far enough fromthe
home offce (in miles) to capture geographic diversifcation benefts. The sample
analyzed here of both failed and surviving banks in Florida and Georgia, who are largely
engaged in intrastate branching, suggests that branching intensity, measured by both
number of branches and distance of branches from the home offce, does not
signifcantly impact the probability of failure. In other words, it suggests that even if
banks had branched in signifcant distances and numbers, the probability of failure
would not be affected. Some policy experts have, in the wake of the fnancial crisis,
argued for policy changes that would limit the size of banks. Our results suggest that
such limitations, at least as far as branching is concerned, may not compromise bank
stability.
In the end, we fnd that most banks do not engage in interstate branching and that the
distance of the branch from the home offce does not impact the probability of failure.
However, both balance sheet diversity and operating in more competitive environments
reduce the probability of failure. Like the fndings in Carlson and Mitchener (2006), our
results show increasing bank competition is stabilizing possibly by causing weaker
banks to exit. Our fndings also indicate that less diversifed banks are more likely to fail
consistent with existing literature (Lu and Whidbee, 2013). Thus, to the extent that
branch activity improves bank diversity and increases market competition, a bank
structural strategy that includes a branch network should be stabilizing for the bank.
Notes
1. An article in The Economist makes a similar prediction that branches will become less
important and that there will be far fewer of them as a result of technology (see The
Economist, 2012a “Counter Revolution”).
2. The interstate branching data come from the FDIC found at www2.fdic.gov
3. Branch data come from the FDIC at fdic.gov
4. The smallest fve banks engaged in interstate branching have total assets (in thousands)
ranging from $11,263 to $54,598 and four of the fve had one branch, total, but it was out of
state, and one of these small banks had two branches out of state. These data come fromthe
Institutional Directory at www2.fdic.gov
5. More specifcally, for interstate branching banks, assets average $18,263,658,000, and for the
entire population of commercial banks assets average $13,070,000,000. Data are from
www2.fdic.gov
6. The interstate branching data come from the FDIC found at www2.fdic.gov, and these
fgures are as of June 30, 2012.
7. See Hendrickson (2010) for a complete review of these historical debates. Related literature,
such as Huang (2008), considers whether branching affects local economic growth. This
paper focuses on the relationship between branching and bank stability.
JFEP
6,2
148
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
8. Electronic banking products include direct deposit, ATM cards, debit cards, automatic
payment services and computer banking services (Robbins (2006, p. 4).
9. Diversity, as defnedbyKolari et al. (2002), is the sumof squaredratios of business loans, real
estate loans, consumer loans and securities to total assets. Higher numbers therefore
represent less diversifcation, with a value of 1 indicating none.
10. In this regard, our methodology is similar to Carlson (2004) who examines how branching
and balance sheet differences in 1929 infuence the likelihood of the bank surviving the
Depression. Whereas Carlson examines survival, we examine failure. Carlson notes,
however, that his conclusions hold when analyzing failure.
11. Of course, a bank where NOBRANCH is equal to 1 will also have BRANCHNUM equal to 0
and AVGMILES equal to 0. The correlation between NOBRANCH and BRANCHNUM is
?0.33, and the correlation between NOBRANCH and AVGMILES is ?0.26.
12. Various measures of competition were created. For example, we also found the number of
branches of two large regional banks, Suntrust and Regions, located in the headquarter city
of each parent bank in our sample. The results remained unchanged, and these branches
were not statistically signifcant. Banks with a greater share of deposits are also likely to be
larger. The correlation between deposit markets share and total assets is 0.068.
13. Other measures of branching, such as the percentage of branches within 30 miles or the
percentage of branches out of state, were also not statistically signifcant.
14. We thank an anonymous referee for pointing this out. In addition to lagging the variables
one year, several other specifcations were estimated. Specifcally, we averaged the
independent variables over 2005-2007, 2000-2007 and 1994-2007. We also estimated the
regression lagging the independent variables four and fve years. In all cases, the results
remain consistent, particularly for the NOBRANCH variable.
15. In contrast, Huang (2008) does not fnd evidence of accelerated economic growth following
branching deregulation, so this conclusion is not conclusive.
References
Amel, D.F., Kennickell, A.B. and Moore, K.B. (2008), “Banking market defnition: evidence from
the survey of consumer fnances”, Financial and Economic Discussion Series No. 35, Federal
Reserve Board, Washington, DC.
Amel, D.F. and Starr-McCluer, M. (2002), “Market defnition in banking: recent evidence”, The
Antitrust Bulletin, Vol. 47 No. 1, pp. 63-89.
Aubuchon, C.P. and Wheelock, D.C. (2010), “The geographic distribution and characteristics of US
bank failures”, 2001-2010: do bank failures still refect local economic conditions? Federal
Reserve Bank of St. Louis Review, September/October, pp. 395-415.
Baldwin, W. (2011), “Who needs branches?” Forbes, pp. 48-49.
Brevoort, K.P., Holmes, J.A. and Wolken, J.D. (2010), “Distance Still Matters: The Information
Revolution in Small Business Lending and the Persistent Role of Location, 1993-2003,”
Board of Governors of the Federal Reserve System, Finance and Economics Discussion
Series, 2010-08.
Bucks, B.K., Kennickell, A.B., Mach, T.L. and Moore, K.B. (2009), “Changes in US family fnances
from2004 to 2007: evidence fromthe survey of consumer fnance”, Federal Reserve Bulletin,
pp. A1-A56.
149
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Calomiris, C.W. (1990), “Is deposit insurance necessary? A historical perspective”, Journal of
Economic History, Vol. 50 No. 2, pp. 283-295.
Calomiris, C.W. and Mason, J.R. (2003), “Fundamentals, panics, and bank distress during the
depression”, American Economic Review, Vol. 93 No. 5, pp. 1615-1647.
Carlson, M. (2004), “Are branch banks better survivors? Evidence from the depression era”,
Economic Inquiry, Vol. 42 No. 1, pp. 111-126.
Carlson, M., and Mitchener, K.J. (2006), “Branch banking, bank competition, and fnancial
stability,” Journal of Money, Credit and Banking, Vol. 38 No. 5, pp. 1293-1328.
Carlson, M. and Mitchener, K.J. (2007), “Branch banking as a devise for discipline: competition and
bank survivorship during the great depression”, NBER Working Paper No. 12938,
available at: www.nber.org/papers/w12938 (accessed 2 April 2011).
Cohen, A. and Mazzeo, M.J. (2010), “Investment strategies and market structure: an empirical
analysis of bank branching decisions”, Journal of Financial Services Research, Vol. 38,
pp. 1-21.
Collins, C.W. (1926), The Branch Banking Question, The Macmillan Company, New York, NY.
DeYoung, R. (1999), “Birth, growth, and life or death of newly chartered banks”, Federal Reserve
Bank of Chicago, Economic Perspectives, pp. 18-35.
Dick, A.A. (2006), “Nationwide branching and its impact on market structure, quality, and bank
performance”, The Journal of Business, Vol. 79 No. 2, pp. 567-592.
Heitfeld, E.A. and Prager, R.A. (2004), “The geographic scope of retail banking markets”, Journal
of Financial Services Research, Vol. 25, pp. 37-55.
Hendrickson, J.M. (2010), “The interstate banking debate: a historical perspective,” Academy of
Banking Studies Journal, Vol. 9 No.2, pp. 95-130.
Hendrickson, J.M. and Nichols, M.W. (2011), “The lights went out in georgia: explaining
commercial bank failures in the 21st century”, Academy of Business Journal, Vol. 1,
pp. 37-67.
Huang, R.R. (2008), “Evaluating the real effect of bank branching deregulation: comparing
contiguous counties across US state borders”, Journal of Financial Economics, Vol. 87,
pp. 678-705.
Jayaratne, J. and Strahan, P.E. (1998), “Entry Restrictions, industry evolution, and dynamic
effciency: evidence from commercial banking,” Journal of Law and Economics, Vol. 41,
pp. 239-273.
Johnson, C.A. and Rice, T. (2008), “Assessing a decade of interstate bank branching,” Washington
and Lee Law Review, Vol. 65, pp. 73-127.
Kolari, J., Glennon, D., Shin, H. and Caputo, M. (2002), “Predicting large US commercial bank
failures”, Journal of Economics and Business, Vol. 54 No. 1, pp. 361-387.
Kupiec, P.H. and Ramirez, C.D. (2013), “Bank failures and the cost of systemic risk: evidence from
1900 to 1930”, Journal of Financial Intermediation, Vol. 22, pp. 285-307.
Lu, W. and Whidbee, D.A. (2013), “Bank structure and failure during the fnancial crisis”, Journal
of Financial Economic Policy, Vol. 5 No. 2.
Mishkin, F.S. and Strahan, P.E. (1999), “What will technology do to fnancial structure?”, Working
Paper 6892 at the National Bureau of Economic Research, available at: www.nber.org/
papers/w6892 (accessed 2 March 2011)
Mitchener, K.J. (2005), “Bank supervision, regulation, and instability during the great depression”,
Journal of Economic History, Vol. 65 No. 1, pp. 152-185.
JFEP
6,2
150
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Mitchener, K.J. (2007), “Are prudential supervision and regulation pillars of fnancial stability?
evidence from the great depression”, The Journal of Law & Economics, Vol. 50 No. 2,
pp. 273-302.
Petersen, M.A. and Rajan, R.G. (2002), “Does distance still matter? The information revolution in
small business lending”, The Journal of Finance, Vol. 57 No. 6, pp. 2533-2570.
Ramirez, C.D. (2003), “Did branch banking restrictions increase bank failures? evidence from
Virginia and West Virginia in the late 1920s”, Journal of Economics and Business, Vol. 55
No. 4, pp. 331-352.
Robbins, E. (2006), “Location, location, location: has electronic banking affected the importance of
bank location?”, Financial Industry Perspectives, Federal Reserve Bank of Kansas City,
pp. 1-12.
Sprague, O.M.W. (1903), “Branch banking in the United States”, Quarterly Journal of Economics,
Vol. 17 No. 2, pp. 242-260.
Stiroh, K.J. and Strahan, P.E. (2003), “Competition dynamics of deregulation: evidence from US
banking”, Journal of Money, Credit, and Banking, Vol. 35 No. 5, pp. 801-828.
Strahan, P.E. (2003), “The real effects of US banking deregulation”, Review, Federal Reserve Bank
of St. Louis, July/August, pp. 111-128.
The Economist (2012a), “Counter revolution; consumer banking”, The Economist, 19 May, p. 18.
The Economist (2012b), “Withering away: branches”, The Economist, 19 May, p. 8.
Wheelock, D.C. (2011), “Have acquisitions of failed banks increased the concentration of US
banking markets?”, Review, Federal Reserve Bank of St. Louis, May/June, pp. 155-168.
Further reading
Berger, A.N. (2003), “The economic effects of technological progress: evidence from the banking
industry”, Journal of Money, Credit and Banking, Vol. 35 No. 2, pp. 141-176.
Berger, A.N., Demirguc-Kunt, A., Levine, R. and Haubrich, J.G. (2004), “Bank concentration and
competition: an evolution in the making”, Journal of Money, Credit, and Banking, Vol. 36
No. 3, pp. 433-451.
Calem, P.S. and Nakamura, L.I. (1998), “Branch banking and the geography of bank pricing”, The
Review of Economics and Statistics, Vol. 80 No. 4, pp. 600-610.
Calomiris, C.W. and Mason, J.R. (2000), “Causes of US bank distress during the great depression”,
Working Paper 7919, National Bureau of Economic Research.
Degryse, H. and Ongena, S. (2005), “Distance, lending relationships, and competition”, The Journal
of Finance, Vol. LX No. 1, pp. 231-266.
Harvey, J. and Spong, K. (2007), “The changing banking structure: what expansion strategies are
community banks adopting?”, Financial Industry Perspectives, Federal Reserve Bank of
Kansas City, pp. 1-15.
Kroszner, R.S. and Strahan, P.E. (2007), “Regulation and deregulation of the US banking industry:
causes, consequences, and implications for the future NBERbook in progress”, available at:
www.nber.org/books_in_progress/econ-reg/kroszner-strahan9-26-07.pdf (accessed 25 April).
Corresponding author
Jill M. Hendrickson can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
151
Bank branch
location
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
:
4
9
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_723624583.pdf