WORKING PAPER SERIES
Why Do Banks Disappear: The Determinants of
U.S. Bank Failures and Acquisitions
David C. Wheelock
Paul W. Wilson
Working Paper 1995-013B
http://research.stlouisfed.org/wp/1995/95-013.pdf
PUBLISHED: Review of Economics and Statistics,
February 2000, pp. 127-38
FEDERAL RESERVE BANK OF ST. LOUIS
Research Division
411 Locust Street
St. Louis, MO 63102
______________________________________________________________________________________
The views expressed are those of the individual authors and do not necessarily reflect official positions of
the Federal Reserve Bank of St. Louis, the Federal Reserve System, or the Board of Governors.
Federal Reserve Bank of St. Louis Working Papers are preliminary materials circulated to stimulate
discussion and critical comment. References in publications to Federal Reserve Bank of St. Louis Working
Papers (other than an acknowledgment that the writer has had access to unpublished material) should be
cleared with the author or authors.
Photo courtesy of The Gateway Arch, St. Louis, MO. www.gatewayarch.com
WHY
DO
BANKS
DISAPPEAR?
THE
DETERMINANTS
OF
U.S.
BANK
FAILURES
AND
ACQUISITIONS
August
1995
ABSTRACT
This
paper
examines
the
determinants
of
individual
bank
failures
and acquisitions
in
the
United
States
during
1984-1993.
We
use
bank-specific
information
suggested by
examiner CAMEL-
rating
categories
to
estimate
competing-risks hazard
models
with
time-varying
covariates.
We
focus
especially
on
the
role
of
management quality,
as
reflected
in
alternative
measures
of
x
-
efficiency
and
find
the
inefficiency increases
the
risk
of
failure,
while
reducing
the
probability
of
a
bank’s
being
acquired.
Finally, we show
that
the
closer
to
insolvency
a
bank
is,
as
reflected
by
a
low
equity-to-assets ratio, the
more
likely
its
acquisition.
KEYWORDS:
Bank
failures,
bank
acquisitions, managerical
efficiency,
hazard model
estimation
JEL
CLASSWICATION:
021,
034
David
C.
Wheelock
Senior
Economist
Federal
Reserve
Bank
of
St.
Louis
411
Locust
St.
Louis,
MO
63102
Paul
W.
Wilson
Associate
Professor
Department
of
Economics
University
of
Texas
at
Austin
Austin,
TX 78712
We
have benefited from
comments
on
an
earlier
version
by
Alton Gilbert, Shawna
Grosskopf,
Subal
Kumbhakar,
Michel Mouchart,
Larry Wall
and
an
anonymous referee.
Any
remaining
errors
are
our
responsibility.
1.
INTRODUCTION
From
a
post-war
peak
of
over
15,000
banks
in
1984,
the
number
of
commercial banks
in
the
United
States
had
declined
to
10,741
ten
years
later,
and
continues
to
fall.
Banks
disappear
primarily
for
one
of
two
reasons:
either
they
fail,
or
they
are
acquired
by
(or
merge
with)
another
bank.’
Since
1984,
the
number
of
acquisitions
has
been
roughly
four
times
the
number
of
failures,
even
counting
acquisitions
of
insolvent
banks
only
as
failures.
Moreover,
while
the
number
of
failures
has
declined
sharply
since
1990,
acquisitions
con
-
tinue
at
a
record
pace.
The
enactment
of
new
interstate
banking
and
branching
authority
portends
even
more
acquisitions
in
the
future, with
the
possibility
of
a
radical
change
in
the
market
structure
of
the
U.S.
banking industry.
Several
studies
have
sought
to
identify
the
characteristics
that
cause
banks to
fail.
2
Apart
from
excessive
risk-taking,
or
simply
bad
luck,
banks
that
are
poorly managed
are
thought to
be
prone
to
failure.
3
By
contrast,
the
characteristics
that
determine whether
a
bank
will be
a
takeover
target
have
received
comparatively little
attention.
4
One
hy
-
pothesis,
discussed
by
Hannan
and
Rhodes
(1987),
suggests
that
poorly-managed
banks
are
likely
targets
for acquisition
by
bankers
who
think
they
can
enhance
the
target’s
man
-
agement
quality,
and
hence
its profitability
and
value.
Fundamentally,
an
acquirer
will
evaluate whether
the
expected profit
stream
generated
by
an
acquisition,
less
any
costs
of
consummating
the
takeover
and
reorganization,
exceed
the
price
it
must
pay
to
acquire
the
bank.
If
an
acquirer
expects
a
particular
acquisition
to
generate
high
profits,
or
entail
‘Throughout
the
paper
we
use “acquisitions”
and
“mergers”
interchangeably.
Acquisitions
greatly
outnumber
mergers
and
our
data
source,
the
National
Information Center
(NIC)
database,
does
not
distinguish
between acquisitions
and
mergers.
2
See
Thomson
(1991)
for
a
recent
example
and
survey.
3
See,
for example,
Berger
and
Humphrey
(1992a)
and
Barr
and
Siems
(1994).
4
Several
studies
have examined
the
effect
of
mergers
on
bank
performance,
including
Neely
(1987),
Berger
and
Humphrey
(1992b),
Cornett and
Tehranian
(1992),
Linder
and Crane
(1992)
and
Akhavein,
Berger
and
Humphrey
(1996).
We
are aware
of
just
two
studies
examining
the
characteristics
affecting
the
likelihood
that
a
bank
will
be
taken
over.
Hannan
and
Rhodes
(1987)
find no
relationship
between
a
bank’s performance, measured
by
either
rate
of
return
or
rate
of
return
relative
to
market
competitors,
and
the
probability
of
acquisition.
Amel
and
Rhodes
(1989),
however,
find
a
negative
relationship
between
performance
and
acquisition.
relatively
low
reorganization
costs,
it
might
willingly
pay
a
considerable
premium
over
book
value
to
acquire
a
controlling
interest
in another
bank.
When
the
expected
profit
stream
is
relatively
modest, or
significant
costs
of
reorganization
are
anticipated,
however,
an
acquirer
might
be
unwilling
to
pay
such
a
high
purchase
price.
5
Because
of
this
tradeoff,
we
have
no
a
priori
expectation
that
poorly-managed
banks
will
have
a
higher
probability
of
being acquired
—
that
is
an
empirical
question,
and one
with
considerable
relevance
for
understanding
the
likely
outcome
of
continued
banking
industry
consolidation.
Management
quality
is
difficult
to
measure
directly
because
it
can
take
several
forms.
A
considerable
literature has
developed
on
the
measurement
of
productive
efficiency in
banking,
however,
which
conceivably
reflects
management
quality.
Researchers
have
found
that
banks in
general
suffer from
considerable
managerial,
or
“x-,” inefficiency,
as
opposed
to
scale
or
scope
inefficiency.
6
There
are,
however,
a
number
of
ways
to
measure
managerial
inefficiency.
In
this
paper,
we
investigate
whether
managerial
inefficiency,
measured
using
two
of
the
most
common techniques,
influences
the
probability of
failure
or
acquisition,
after
controlling for
bank
portfolio
characteristics
and
operating
environments.
Because
banks
may
disappear through
either
failure
or
acquisition,
and
since
occurrence
of
either
event
precludes
the
other,
we
use
a
competing-risks
hazard
model
framework
to
identify
characteristics
leading
to
each
outcome.
Unlike
more
commonly
used
discrete-outcome
models,
hazard
models
make
more
efficient
use
of
the
data
by
explicitly
incorporating
information
about
the
timing
of
alternative
outcomes.
Also,
unlike
most
other
banking
studies,
which
are typically based
on
relatively
small
samples
and
short
periods,
we
use
quarterly
data
for
1984-93 on
the
universe
of
U.S.
banks
in
existence
in
1984
to
examine
5
Note
that
because
the
expected acquisition
costs
and
profits
generated
by
an acquisition
will
differ
among
potential
acquirers,
no two
potential acquirers
will
necessarily
be
willing
to pay
the
same
price
for
controlling
interest
in
a
given
bank.
Just
as
a
handyperson
may
be
willing
to
buy
a
“fixer-upper”
house
at
a
price
that
other
buyers
would
not
pay,
some
bankers
specialize
in
purchasing
inefficient
banks
with
the
aim
of
improving
their
management
and,
hence,
value, while
other
bankers
might prefer
to
acquire
well-managed banks.
6
Berger
and
Humphrey
(1991)
is
perhaps
the
most
important
article
in this
literature.
See
Berger,
Hunter
and
Timme
(1993)
for
a
survey.
—2—
both
failures
and
acquisitions.
Federal
regulators
evaluate
banks
on five
criteria:
capital
adequacy,
asset
quality,
man
-
agement, earnings
and
liquidity (CAMEL).
We
base
our empirical model
on
these
criteria,
and
identify
a
number
of
characteristics
significantly
affecting
the
likelihood
that
a
bank
will
disappear
because
of
failure
or
acquisition.
Not surprisingly,
we
find
that
highly-
leveraged
banks,
banks
with
low
earnings,
low
liquidity,
or
risky asset
portfolios
are
more
likely
to
fail
than
other
banks.
Holding
other
factors
constant,
we
find
that
banks
located
in
states
that
permit
branching
are
less likely
to
fail,
indicating
perhaps
the
benefits
of
ge
-
ographic
diversification.
And,
finally, we
find
strong
evidence
that
managerial
inefficiency
increases
the
likelihood
of
bank
failure.
We also find
that
proximity
to
insolvency
strongly
affects
the
likelihood
that
a
bank
will
be
acquired.
All
else
equal,
the
less-well
capitalized
a
bank
is,
the
greater
the
probability
that it
will
be
acquired, suggesting
the
acquisition
of
some
banks
just
before
they
become
insolvent.
We also
find
that
banks
with
low
earnings,
low
liquidity,
or
relatively
high
non-performing
loan
ratios
are
less
attractive
takeover
targets.
Banks
located
in
states
permitting
branching,
as
well
as
small
banks
in
general,
have
been more
likely
to
be
ac
-
quired.
And,
finally, we find
that
inefficient
banks
are
less
likely
to
be
acquired,
controlling
for
leverage
and
other
balance sheet
and
environmental characteristics. Managerial
ineffi
-
ciency
could
reflect
excessive
use
of,
or
payment
for,
physical
plant
or
labor,
or
excessive
deposit interest
cost.
The
cost
of
reorganizing
an
inefficient
bank
could
thus
be
high~
7
Moreover,
managerial
inefficiency
might
be
taken
as
a
signal
of
potential
problems
that
are
themselves
unobservable
(e.g.,
bad
loans
or
accounting irregularities). Thus,
holding
other
portfolio
and
environmental conditions
constant,
acquirers
on
average
apparently
prefer
not
to
purchase
inefficient
banks.
7
Large
layoffs
of
personnel
or
branch
closings
are
sometimes
necessary
to
improve
a
bank’s
efficiency.
Perhaps
because
such
actions
can
entail
considerable
cost,
both
monetarily
and in terms
of
public
relations,
studies
have
generally
found
few
cost
efficiency
gains
associated
with
acquisitions
and
mergers
of
banks.
Akhavein, Berger
and
Humphrey
(1996),
for example,
find
almost
no
such gains
for
recent
mergers
of
large
bank
holding companies.
—3—