The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
For release on delivery
10:00 a.m. EDT
September 24, 1991
Testimony by
John P. LaWare
Member, Board of Governors of the Federal Reserve System
before the
Committee on Banking, Finance and Urban Affairs
U.S. House of Representatives
September 24, 1991
I am pleased to appear before this Committee on behalf of the
Federal Reserve Board to discuss issues related to mergers among U.S.
banking organizations.
The last ten years have seen considerable
consolidation of our banking system, a process that probably will
continue for some time.
And while banking consolidation is in many
ways a natural response to the evolution of the overall banking
environment, the significant changes we have observed do raise a
number of public policy questions and concerns.
In the Board's view,
the primary objectives of public policy in this area should be to help
manage the evolution of the banking industry in ways that preserve the
benefits of competition for the consumers of banking services, and to
ensure a safe, sound, and profitable banking system.
My statement
today will focus on how, within the context of existing law, the
Federal Reserve is pursuing these goals, and will review the potential
economic effects of bank mergers.
Merger Trends in the 1980s
It is useful to begin a discussion of the public policy and
other implications of bank mergers with a brief description of recent
bank consolidation trends.
The statistical tables in Appendix A of my
statement provide some detail that may be of interest to the
Committee.
From a variety of perspectives the pace of bank mergers has
accelerated over the last decade.
For example, excluding acquisitions
of failed or failing banks by healthy banks, in 1980 there were 188
bank mergers involving some nine billion dollars in acquired assets;
by 1987 the annual number and dollar value of mergers peaked for the
decade at 710 mergers and $131 billion of acquired assets.
In 1989,
the number of mergers dropped back to an estimated 550 involving an
estimated $60 billion of bank assets acquired.
The number of mergers
involving large bank holding companies also increased.
In 1980 there
were no mergers or acquisitions of commercial banking organizations
where both parties had over one billion dollars in total deposits.
The years 1985 through 1990 averaged 13 such transactions per year.
Another perspective is provided by the fact that the total number of
U.S. banking organizations declined steadily throughout the 1980s.
In
1980 there were 12,679 banking organizations (including 14,737 banks),
by 1985 11,377, and in 1990 some 9,688 (including 12,526 banks), a 24
percent decline in organizations and a 15 percent decline in numbers
of banks from 1980.
These trends have been accompanied by an increase
in the share of total banking assets controlled by the largest banking
organizations.
For example, the proportion of domestic banking assets
accounted for by the 100 largest banking organizations went from 48
percent in 1980, to 55 percent in 1985, to 62 percent at year-end
1990.
The trends I have just described must be placed in proper
perspective, because taken by themselves they hide some of the key
dynamics of the banking industry.
For example, while a major reason
for the decline in the number of banking organizations over the 1980s
was the fact that almost 1,100 banks failed, the decline in the total
number of banks was offset considerably by the fact that over that
decade some 2,700 new banks were formed.
Similarly, while during the
1980s over 6,600 bank branches were closed, the same period saw the
opening of well over 16,000 new branches.
Perhaps even more
significant, the total number of banking offices increased sharply,
from about 48,500 in 1980 to almost 60,000 in 1990, a 23 percent rise.
Data on the nationwide concentration of U.S. banking assets
must also be viewed in perspective.
None of the increase in such
concentration among the 100 largest banking organizations has occurred
among the very largest--the ten largest--banks. Rather, the large
regional banks have accounted for all of the increase in the
concentration ratio.
Of course, if the recently announced mergers of
some of our largest banks are implemented, concentration among the top
ten will increase.
Given the Board’ statutory responsibility to ensure
s
competitive banking markets, it is critical to understand that these
nationwide concentration statistics are not the important concept for
assessing competitive effects.
Virtually all observers agree that the
relevant issue is competition in local banking markets.
And the facts
are that, over the last decade, the average proportion of bank
deposits accounted for by the three largest firms in urban markets has
increased by only one percentage point, and has remained virtually
unchanged in rural markets.
These ratios have actually declined in
both types of markets since the mid-1970s.
The apparent contradiction
between increased concentration ratios nationally and virtually
unchanged ratios locally can be explained by several factors.
While
my statement will provide more detail, key considerations include the
fact that most mergers are between noncompeting banks, and those
between entities in the same market have faced new entrants, antitrust
constraints, and have found that smaller bank competitors effectively
limit their ability to increase market share.
Overall, then, the picture that emerges is that of a dynamic
U.S. banking structure with the number of banking offices increasing
sharply and their location extremely sensitive to the demands of
consumers.
In such an environment it is potentially very misleading
to make broad generalizations without looking more deeply into what
lies below the surface.
In part for the same reasons that make
generalizations difficult, the Federal Reserve devotes considerable
care and substantial resources to analyzing individual merger
applications.
Federal Reserve Methodology for Analyzing Proposed Bank Mergers
The Federal Reserve Board is required by the Bank Holding
Company Act (1956) and the Bank Merger Act (1960) to assess the
effects when (1) a holding company acquires a bank or merges with
another holding company or (2) the bank resulting from a merger is a
state chartered member bank.
The Board must evaluate the likely
effects of such mergers on competition, the financial and managerial
resources and future prospects of the firms involved, the convenience
and needs of the communities to be served, and Community Reinvestment
Act requirements.
This section of my statement briefly discusses the
methodology the Board uses in assessing a proposed merger.
In light
of the Committee's specific questions, emphasis is placed on
competitive factors.
In addition, more detailed discussion of the
legal and economic bases for the Board’s assessment of competition is
found in Appendix B.
Competitive Criteria
In considering the competitive effects of a proposed bank
acquisition, the Board is required to apply the same competitive
standards contained in the Sherman and Clayton Antitrust Acts.
The
Bank Holding Company (BHC) Act and the Bank Merger Act do contain a
special provision, applicable primarily in troubled bank cases, that
permits the Board to balance public benefits from proposed mergers
against potential adverse competitive effects.
-
5
-
The Board’s analysis of competition begins with defining the
geographic areas that are likely to be affected by a merger.
Under
procedures established by the Board, these areas are defined by staff
at the local Reserve Bank in whose District the merger would occur,
with oversight by staff in Washington.
To ensure that market
definition criteria remain current, and in an effort to better
understand the dynamics of the banking industry, the Board has
recently sponsored several surveys, including the 1988 National Survey
of Small Business Finances, the national Survey of Consumer Finances,
and telephone surveys in specific merger cases, to assist it in
defining geographic markets in banking.
These surveys and other
evidence continue to suggest that small businesses and consumers tend
to obtain their financial services in their local area.
This local
geographic market definition would, of course, be less important for
the financial services obtained by large businesses.
With this basic local market orientation of consumers and
small businesses in mind, the staff constructs a local market
Herfindahl-Hirschman index (HHI), which is widely accepted as a
sensitive measure of market concentration, in order to conduct a
preliminary screen of a proposed merger.
The merger would not be
regarded as anticompetitive if the HHI and the change in that index do
not exceed the criteria in the Justice Department's merger guidelines
for banking.
However, while the HHI is an important indicator of
competition, it is not a comprehensive one.
In addition to statistics
on bank concentration, economic theory and evidence suggest that other
factors, such as local market services available from nonbank
providers of financial services and potential competition, may have
important influences on bank behavior.
These other factors have
become increasingly important as a result of many recent
procompetitive changes in the financial sector.
Thus, if the level
and change in the HHI are within Justice Department guidelines, there
is a presumption that the merger is acceptable, but if they are not, a
more thorough economic analysis is required.
Because the importance of the other factors that may
influence competition often varies from case to case and market to
market, an in-depth economic analysis of competition is required in
each of those merger proposals where the Justice Department HHI
guidelines are exceeded.
To conduct such an analysis of competition,
the Board uses information from its own major national surveys noted
above, from telephone surveys of consumers and small businesses in the
market being studied, on-site investigations by staff, as well as from
various standard databases with data on market income, population,
deposits, and other variables.
These data, along with results of
general empirical research by Federal Reserve System staff, academics,
and others, are used to assess the importance of various factors that
may affect competition. To provide the Committee with an indication of
the range of "mitigating" factors the Board may consider in evaluating
competition in local markets, I shall briefly outline these
considerations.
Potential competition, or the possibility that other firms
may enter the market as a result of the merger, may be regarded as a
significant procompetitive factor.
It is most relevant in markets
that are attractive for entry and where barriers to entry, legal or
otherwise, are low.
Thus, for example, potential competition is of
relatively little importance in markets where entry via intra- or
interstate branching is severely restricted, or in markets where
branching is restricted and it may be difficult for investors to raise
the minimum capital needed to start a bank.
For potential competition
-
7-
to apply, it will generally be necessary for there to be potential
acquisition targets as well as meaningful potential entrants.
This
factor is most likely to be relevant in urban markets.
Thrift institution deposits are now typically accorded
50 percent weight in calculating statistical measures of the impact of
a merger on market structure for the Board's analysis of competition.
In some instances, however, a higher percentage may be included if
thrifts in the relevant market look very much like banks, as indicated
by the substantial exercise of their transactions account, commercial
lending, and consumer lending powers.
Competition from other depository and nonbank financial
institutions may also be given weight if such entities clearly provide
substitutes for the basic banking services used by most consumers and
small businesses.
In this context, credit unions and finance
companies may be particularly important.
The competitive significance of the target firm can be a
factor in some cases.
For example, if the bank being acquired is not
a reasonably active competitor in a market, its market share might be
given a smaller weight in the analysis of competition than otherwise.
Adverse structural effects may be offset somewhat if the firm
to be acquired is located in a declining market.
This factor would
apply where a weak or declining market is clearly a fundamental and
long-term trend, and there are indications that exit by merger would
be appropriate because exit by closing offices is not desirable and
shrinkage would lead to diseconomies of scale.
This factor is most
likely to be relevant in rural markets.
Competitive issues may be reduced in importance if the bank
to be acquired has failed or is about to fail.
In such a case, it may
be desirable to allow some adverse competitive effects if this means
that banking services will continue to be made available to local
customers rather than be severely restricted or perhaps eliminated.
A very high level of the HHI could raise questions about the
competitive effects of a merger even if the change in the HHI is less
than the Justice Department criteria.
This factor would be given
additional weight if there has been a clear trend toward increasing
concentration in the market.
Finally, factors unique to a market or firm would be
considered if they are relevant to the analysis of competition.
These
factors might include evidence on the nature and degree of competition
in a market, information on pricing behavior, and the quality of
services provided.
Some merger applications are approved only after the
applicant proposes, or agrees to, the divestiture of offices in local
markets that would otherwise violate Justice Department guidelines,
and where the merger cannot be justified using any of the criteria I
have just discussed.
We believe that these divestiture actions have
deterred many banking organizations from applying for mergers that
would be acceptable to the Board only with divestitures that the
applicant is not willing to make.
Safety and Soundness Criteria
In acting upon merger applications, the Board is required to
consider financial and managerial considerations.
In doing so, the
Board’i goal is to promote and protect the safety and soundness of the
s
banking system, and to encourage prudent acquisition behavior by
applicant banking organizations.
The Board expects that holding company parents will be a
source of strength to their bank subsidiaries.
In doing so, the Board
generally requires that the holding company applicant and its
subsidiaries be in at least overall satisfactory condition, and that
any weaknesses be addressed prior to Board action on a proposal.
The
holding company applicant must be able to demonstrate the ability to
make the proposed acquisition without unduly diverting financial and
managerial resources from the needs of its existing subsidiary banks.
The Board has long stressed the importance of capital in
reviewing applications to expand.
It is the Board’s policy that
acquisitions or mergers should not result in a diminution of the
overall capital strength of the combined organizations.
For this
reason, the Board has generally expected that significant acquisitions
or mergers be funded in whole or in part by the issuance of additional
capital.
In this connection, the Board has held that banking
organizations undertaking significant growth, either internally or
through acquisitions or mergers, should operate with capital ratios
well in excess of the supervisory minima, without significant reliance
on intangible assets.
The Board has indicated that this cushion
should be at least 100 to 200 basis points above the minimum ratios;
still larger margins could be called for, depending on the actual
financial condition of the organization and the risks being
undertaken.
This emphasis on capital underlies the Board’s strong
preference that expansionary applications be substantially financed
from the proceeds of equity.
Applications from organizations that do not meet these
capital standards would not be approved unless the organization has
underway a capital augmentation program and can demonstrate the
ability to raise additional tier I (essentially equity) capital
contemporaneously with the acquisition.
As noted, additional capital
-
10-
may also be required to correct any weaknesses in the bank or company
to be acquired.
This public policy serves to protect the existing
satisfactory financial strength of the organization and to prevent an
undesirable decline in capital adequacy caused by the acquisition of
significant additional assets.
It also can serve to moderate the rate
of expansion and enable the organization to absorb the additional
risks.
These general principles apply regardless of the type of
acquisition--banking or nonbanking.
The financial and managerial
analysis of the applicant includes an evaluation of the existing bank,
nonbank subsidiaries, the parent company, the consolidated
organization, and the entity to be acquired.
Community Reinvestment Act Criteria
The Community Reinvestment Act (CRA) performance of banking
organizations that seek the Board's approval to acquire a bank or
thrift is a major component of the "convenience and needs” criteria
that must be considered by the Board.
In making its judgments, the
Board pays particular attention to CRA examination findings.
In
addition, any comments received from the public regarding an
applicant's CRA performance become part of the official record, and
such comments are reviewed carefully.
Indeed, the Board has just
announced its intention to hold public meetings in various locations
on the CRA record of the banks involved in a major merger application.
Banks supervised by the Federal Reserve System--regardless of
the size or the geographic scope of a bank's operations --are examined
for CRA purposes at least every 18 to 24 months.
Banking
organizations with identified weaknesses in their consumer compliance
are examined even more frequently.
Our practice is to review the
-
11-
performance of banks with large intrastate branching systems by
examining a sample of branches, which consists of all major branches
plus one-tenth of all small branches selected on a rotating basis.
This type of system probably could be used for large, interstate
branch systems as well, if the Congress agrees to permit interstate
branching.
Some adjustments may be necessary, though, to ensure that
the CRA examination process continues to work well for banking
organizations that span several states.
The Board expects that banking organizations will have
policies and procedures in place and working well to address and
implement their CRA responsibilities prior to Board consideration of
bank expansion proposals.
These efforts must include methods for
ascertaining the credit needs of the entire service area, including
low and moderate income neighborhoods; credit products designed to
meet those identified needs; outreach and marketing efforts throughout
this service area; involvement by senior management and the
institution’ board of directors in establishing and supervising the
s
implementation of those efforts; and a record of performance in
helping to meet the community’s credit needs through products that are
consistent with the institution’s overall business orientation.
The Board generally does not accept promises for future
action in this area as a substitute for a demonstrated record of
performance.
Instead, the Board has accepted commitments for future
action as a means of addressing areas of weakness in an otherwise
satisfactory record.
Where commitments have been accepted, the Board
monitors progress in implementing the proposed actions, both through
reports and through the application process.
-
12-
Protection of the Deposit Insurance Fund
In recent years, many bank merger and acquisition cases have
involved failed or failing banks.
By far the most common resolution
method used by the FDIC has been the so-called "purchase and
assumption" procedure.
Under this procedure, a healthy banking
organization assumes all or a part of the assets and liabilities of a
failed or failing bank.
The Federal Reserve favors continuing to give
the FDIC some flexibility in how it resolves such banks.
The need for flexibility derives from our concern about the
possibility of systemic risk associated with a failing bank.
Systemic
risk refers to the chance that financial difficulties at one bank, or
possibly a small number of banks, may spill over to many more banks
and perhaps the entire financial system.
In principle, systemic risk
could develop if a number of smaller or regional banks were to fail.
However, in practice systemic risk is more likely to be associated
with failures of large institutions.
In any event, in some individual
cases the prevention of systemic risk can be an important factor in
assessing a proposed merger or acquisition.
That systemic risk is most likely in cases of financial
distress at large institutions raises a public policy concern with
mergers that create large banking organizations.
Clearly, it would be
unwise to approve mergers that significantly increase systemic risk.
For this reason the Board places great weight on the capital ratio,
and other indicators of financial strength that I have already
discussed, of the combined firm in any merger application that comes
before it.
However, there is an additional point that should be
stressed.
The logical connection between bank merger policy and the
potential for systemic risk emphasizes the interdependence between our
-
13-
discussion today and the need for comprehensive reform of our system
of banking and financial regulation.
If the United States is to have
a safe, sound, competitive and profitable banking system, then the
Board strongly urges that the Congress pass a broad reform package
along the lines of that proposed by the Treasury and supported by the
Board.
Such legislation would call for strong capital, prompt
corrective action policies to deal with financially distressed
depositories, frequent on-site examinations, increased opportunities
for geographic diversification of risk and reduced costs through full
interstate branching, and a broader range of permissible activities
for financial services holding companies with well-capitalized bank
subsidiaries.
By increasing the safety and soundness of our banking
system, these reforms would lessen the likelihood of a major systemic
threat, and would allow our banking system to adjust to evolving
market and technological realities.
But even with these reforms, the
Board believes it would be a mistake to eliminate entirely the ability
of the authorities to act to protect the economy by assisting in the
acquisition of a large failing bank in such a way as to protect all
depositors.
We agree that this approach has been overused in the
past, and requires some constraints.
We urge, however, that the
authorities’ hands not be tied as they would be under H.R. 6.
Potential Implications of Bank Mergers
The increased rate of bank mergers has raised a number of
concerns regarding the potential effects of banking consolidation on
consumers whose demands for banking services are primarily local in
nature, on the performance of the merged banks (including prices paid
by consumers at those banks), and on the overall structure of the U.S.
banking industry.
-
14-
Bffects of Mergers on Locally Limited Customers
The current merger wave in the banking industry is likely to
have only modest effects on the availability of services to consumers
and small businesses that rely primarily on local providers for their
financial services.
There are two reasons for this:
(1) to date,
most mergers have not been between banks operating in the same local
banking markets; and (2) the effects of intramarket mergers can be,
and thus far have been, limited by antitrust constraints on such
mergers.
Even in those places where in-market mergers have occurred,
the effect on competition has on average not been substantial.
This, of course, does not mean that no consumers have ever been harmed
by mergers.
No policy can guarantee that result.
But it does suggest
that increases in local market concentration have been limited by the
Board’s application of antitrust standards to within-market merger
applications.
In addition, the Board’s policies have almost certainly
discouraged some potential bank mergers before an application was ever
filed.
Moreover, considerable intramarket consolidation could occur
without significant anticompetitive effects.
Many urban markets could
see a relatively large number of in-market mergers before antitrust
guidelines would be violated.
Recent legislative changes have made
thrift institutions more important competitors for banking services,
and this has helped to reduce concerns about anticompetitive effects
from intramarket bank mergers.
Although, as I shall be discussing shortly, small banks
remain viable competitors in markets after larger bank mergers, some
research suggests that large banks may adopt new banking technologies
--such as automated teller machines and bank credit cards--more
rapidly than small banks.
Thus, bank mergers may enhance consumer
-
convenience.
15-
On the other hand, in-market bank mergers often lead to
some branch closings, raising concerns that consumer convenience may
be harmed.
Indeed, one of the factors reviewed in a CRA examination
is the bank’s record of opening and closing offices.
However, as I
pointed out earlier, there has been a substantial increase in the
number of bank offices in the U.S. in recent years.
More important,
there is no reason to suspect that the market factors that have led to
this increase in the number of offices have changed.
Indeed, the
abolition of constraints on interstate branches would greatly
facilitate this process.
That is, if merging banks should close
branches, the opening of branches by existing competitors or by new
entrants to the market is, based on past experience, likely to occur,
and would become even more so with full interstate branching.
If
consumers demand locational convenience, banks of all sizes will need
to be responsive if they expect to remain viables
Effects of Mergers on Bank Performance
Federal Reserve System staff have conducted several studies
over many years on the effects of bank mergers and acquisitions.
Some
of these studies have focused on the effect of mergers on bank profits
and prices, while others have looked at the potential for cost savings
and efficiencies derived from mergers.
At the Committee’s request, a
detailed review of the studies appears in Appendix C.
Of those studies concerned with profits and prices. Some have
looked at the effects of specific mergers, while a majority have
approached this issue more indirectly by examining how bank profits
and prices differ across banking markets.
Each type of study is
relevant to an assessment of the impact of bank mergers on
performance.
-
16-
Studies of differences in bank profitability across markets
with varying degrees of concentration represent the oldest type of
study relevant to the issue.
Typically, such studies have found that
banks operating in more concentrated markets exhibit somewhat higher
profits than do banks in less concentrated markets.
These higher
profits may reflect the lesser degree of competition in more
concentrated markets.
Many have argued, however, that they are simply
an indication of the greater efficiency and lower costs of the largest
firms in such markets.
Because of this fundamental disagreement,
there is no consensus concerning the meaning of this type of study for
merger policy.
Other studies have looked across banking markets for
differences in the prices that banks charge their loan and deposit
customers.
For the most part, such studies have found that banks
located in relatively concentrated markets tend to charge higher rates
for certain types of loans, particularly small business loans, and
tend to offer lower interest rates on certain types of deposits,
particularly transactions accounts, than do banks in less concentrated
markets.
These studies tend to be clearer in terms of their
implications for merger policy, because they suggest that mergers
resulting in relatively high levels of local banking market
concentration can adversely affect local bank customers.
That is,
these studies support the need to maintain antitrust constraints if
locally limited bank customers are to continue to receive
competitively priced banking services.
Whether or,not specific past mergers have resulted in higher
loan rates, lower deposit rates, or in other ways disadvantaged
banking customers is very much a different question.
Studies of the
competitive impact of individual bank mergers in essence focus on the
-
17-
issue of whether regulatory authorities have been successful in
applying antitrust constraints.
In general, such studies have been rare, making
generalizations hazardous.
Of those studies that have been conducted,
however, no evidence of significant anticompetitive effects
attributable to past mergers has been found.
One such effort examined
the impact of the merger of two large in-state banks on two types of
deposit rates, and found no adverse effects on bank customers.
Other
studies using different approaches have also failed to find
anticompetitive effects.
Thus, it appears that while significant
mergers, particularly intramarket mergers that directly affect market
concentration, can in principle adversely affect banking customers,
there is no direct evidence to date that those mergers passing
regulatory scrutiny have in fact done so.
A related issue relevant to the effect of mergers concerns
the prospect that, through merger, greater bank efficiency can be
achieved, thus yielding a healthier, more competitive banking firm.
As in the case of the bank pricing studies, studies of the effect of
mergers on bank efficiency may be divided into those that do and those
that do not look at the effects of specific mergers.
A large number of studies have sought to determine whether
larger banking organizations exhibit lower average costs than do
smaller organizations.
In general, these studies of "scale economies"
find that cost advantages of large firms either do not exist or are
quite small, and most do not find scale economies to exist beyond the
range of a small- to medium-sized bank.
Another strand of research has attempted to discover whether
there are important differences in the efficiency with which banks use
inputs to produce a given level of services.
These studies, which
-
18-
essentially focus on management skills, suggest that some banks, both
large and small, are just a lot better than others at using their
inputs, such as labor and capital, in a productive way.
Indeed,
estimates of these so-called cost inefficiencies suggest that
management skills dominate any benefits from economies of scale.
In
addition, there is some evidence that these differences in management
efficiencies play a role in the incidence of bank failure.
An
estimated over 50 percent of the bank failures in the 1980s came from
the highest (noninterest) cost quartile of banks, while fewer than
10 percent are estimated to have occurred in the lowest cost quartile.
In the past couple of years, a number of researchers have
sought to determine whether individual past mergers have resulted in
cost savings.
Typically, such studies examine the changes in
noninterest expenses observed before and after the merger and, in some
cases, compare them to the same changes observed concurrently in banks
that did not participate in mergers.
With one or two exceptions,
these studies generally have not found evidence of substantial cost
savings beyond those associated with shrinkage of the firms in
question after merger.
However, the previously noted evidence indicating substantial
differences in the relative efficiency of banks suggests that
substantial cost savings are theoretically possible for many banks.
For example, a study recently completed at the Board has estimated
that annual cost savings on the order of $17 billion would result if
the lowest cost banks in the country were to acquire the highest cost
banks, and if the costs of the acquired banking organizations were
subsequently reduced to the level of the acquiring banks.
While some
of these cost differences may simply reflect differences in the level
of services offered to the public, such results are nevertheless
-
19-
suggestive of potential gains from acquisitions of inefficient firms
by efficient ones.
Indeed, they indicate that one possible future
scenario, as banking becomes even more competitive, is that it may
become increasingly common for relatively efficient banks to take over
relatively inefficient ones and convert the more poorly performing
institutions into viable, low-cost competitors.
Surely consumers of
financial services could only be better off if such a future were to
occur, and competitive markets maintained.
Once again, however, I would point out and emphasize the
connections between our discussion here today and the need for
fundamental reform of our banking and financial regulatory system.
Achievement of the scenario I have just described depends heavily upon
creating an environment not only in which banks can compete more
effectively, but also one where the likelihood that the deposit
insurance funds will suffer losses is greatly reduced, such as would
occur with higher capital, more frequent examinations, and prompt
corrective action.
Such reforms would put even more pressure on
inefficient banks to achieve cost economies.
emphasize one more key point.
In this regard. I would
Care should be taken to ensure that the
bank reform package does not impose costly new regulations on banks
that would substantially offset the cost savings that result from
other reform actions.
A competitive, safe, and sound banking system
must also be one in which banks can make a profit.
The Effects of Mergers on Banking Structure
Ultimately, the effects of bank mergers on consumer welfare
depend to a substantial extent on the resulting degree of
concentration in local banking markets.
As I have already indicated,
one of the tasks of public policy is to apply the antitrust standards
-20-
in such a way as to maintain competitive banking markets.
Because it
appears that anticompetitive concerns are normally most serious in
local banking markets, this section provides somewhat more detail on
the implications of bank mergers for local market concentration.
In
addition, since the Committee’s letter of invitation asked for some
ideas on what the U.S. banking industry might look like by the 21st
century, I shall briefly address this inherently highly speculative
issue.
Metropolitan Statistical Areas (MSAs) and non-MSA counties
are often used as proxies for urban and rural banking markets.
The
average three-firm concentration ratio for urban markets so measured
increased by only one percentage point between 1980 and 1990 (see
table A-6 in Appendix A). Average concentration in rural counties was
virtually unchanged.
Thus, despite the fact that there were over
5,000 bank mergers during the 1980s, local banking market
concentration has remained about the same.
Why haven't all of these mergers increased concentration by a
greater amount?
There are a number of reasons.
First, as I have
already indicated, many mergers are between firms operating in
different local banking markets.
While these mergers may increase
national or state concentration, they do not increase concentration in
any local banking market.
Second, as I have also already pointed out, there is new
entry into banking markets.
In most markets new banks can be formed
fairly easily, and some key regulatory barriers, such as restrictions
on interstate banking, are much lower than they used to be.
Anecdotal
evidence suggests that new independent local banks have been formed in
many of the banking markets that are dominated by the large multistate
banks.
-21-
Third, the Committee may be surprised to discover that the
evidence overwhelmingly indicates that banks from outside a market
usually cannot increase their market share after entering a new market
by acquisition.
An oft-mentioned example here is the inability of the
New York City banks to gain significant market share in upstate New
York.
More general studies indicate that, when a local bank is
acquired by a large out-of-market bank, there is normally some loss of
market share.
The new owners are not able to retain all of the
customers of the acquired bank.
Fourth, it is important to emphasize that small banks have
been and continue to be able to retain their market share and
profitability in competition with larger banks.
Our staff has done
repeated studies of small banks; all these studies indicate that small
banks continue to perform as well as, or better than, their large
counterparts, even in the banking markets dominated by the major
banks.
Finally, administration of the antitrust laws has almost
surely played a role.
At a minimum, banking organizations have been
deterred from proposing seriously anticompetitive mergers.
And in
some cases, to obtain merger approval, banks have agreed to divest
banking assets and deposits in certain local markets where the merger
would have otherwise resulted in substantially adverse effects.
Future Banking Structure
Where will all of these mergers and changes in banking lead
us?
What will the future structure of the banking industry look like?
To the extent that such forecasts can reasonably be made, it seems
quite likely that the future will contain thousands of small banks,
some regionals, some super-regionals, and a small number of large
-22-
nationwide banks.
There is no reason to believe that small banks will
not continue to remain viable head-to-head competitors in local
markets with their larger rivals.
These rivals will be both regional
banks and a few nationwide banks with offices in hundreds of local
markets coast to coast.
Some of today's large bank mergers are
probably the early stage of the formation of nationwide banks.
I hesitate to make a prediction as to the number of banking
organizations in the future.
There is simply no way to know or
forecast that number with any high degree of certainty.
However, a
recent study by Board staff attempted to make some ballpark
projections in this matter.
Relying primarily on trends observed in
the 1970s and 1980s, and on the assumption that interstate banking
would be allowed through holding companies rather than through
branches, this study projected that the total number of U.S. banking
organizations could be about 5,500 by the year 2010.
This number of
holding companies probably implies between 6,000 and 7,000 banks.
These 5,500 banking organizations include a large number of local
community banks, in addition to regional banks and large, nationally
active banking organizations.
I would guess that full interstate
banking via branching would reduce the number of banking organizations
only Somewhat further, because the staff study had already assumed
interstate operations through the more expensive option of using
multi-bank holding companies.
Conclusion
The increased pace of bank mergers since the early 1980s has
greatly reduced the number of U.S. banking organizations, and resulted
in a substantially higher nationwide concentration of banking assets
at the 100 largest banks.
However, concentration in local banking
-23-
markets , which is normally considered most important for the analysis
of potential competitive effects, has remained virtually unchanged.
In addition, there have been a large number of new bank entrants and a
sharp increase in the number of banking offices.
This illustrates
that the U.S. banking structure is highly dynamic, and that sweeping
generalizations are extremely difficult to make.
The dynamic nature of U.S. banking means that analysis of the
potential competitive and other effects of individual bank mergers
must be done on a case by case, market by market, basis.
Reserve devotes considerable resources to this end.
The Federal
Key factors,
including actual competition from bank and nonbank sources, potential
competition, the general economic health of the market, a variety of
factors unique to a given market, and, in the case of mergers
involving failed or failing firms, systemic risk are all considered.
In addition, safety and soundness and CRA concerns are highly
relevant.
In the end, complex judgments are required to ensure the
appropriate balance of benefits and costs in the public interest.
To date, the available evidence suggests that recent mergers
have not resulted in adverse effects on the vast majority of consumers
of banking services.
It is certainly possible that some customers
have been disadvantaged by some mergers.
And, mergers can no doubt be
very disruptive to bank employees as functions are consolidated and
reorganized.
But these disruptions do not appear to differ
substantively from similar disruptions in other industries undergoing
fundamental change.
It is also clear that substantial harm to consumers would
occur if mergers were allowed to decrease competitive pressures
significantly.
Thus, it is crucial that antitrust standards be
enforced by the bank regulatory agencies and the Department of
Justice.
Given the record of success to date, the Board believes that
our current statutory authority in this area is sufficient to meet
existing and foreseeable concerns.
However, if future developments
warrant, the Board would not be reluctant to seek additional authority
in this area.
The evidence to date does not indicate that substantial cost
savings have resulted from bank mergers.
However, our staff work does
suggest the potential for such savings if well-managed entities
acquire and modify the operations of high-cost organizations.
Given
the continuing pressures for cost minimization in banking, it
certainly seems possible that some of the potential will be realized
in the future.
Last, I would emphasize once again the close link between our
discussion here and the need for comprehensive reform of our system of
banking and financial regulation.
All of us want consumers of
financial services to have available competitively priced, high
quality banking services, and we want to ensure that U.S. taxpayers
are not exposed to excessive risk of loss through the deposit
insurance fund.
To achieve these objectives, U.S. banks must have the
ability to compete effectively and profitably both at home and abroad,
and U.S. regulators must be able to act in timely and effective ways.
The Board therefore urges the Congress to pass the reform proposals
that have been advanced by the Treasury and supported by the Board.
APPENDIX A
STATISTICAL TABLES
Table A-l
Bank Mergers and Acquisitions 1978-1989*
Number of Bank
Mergers
Bank Assets
Acquired ( bill.;
$
1978
144
5.5
1979
179
7.5
1980
188
9.3
1981
359
19.5
1982
422
37.1
1983
432
43.0
1984
553
82.7
1985
553
64.7
1986
625
89.1
1987
710
131.4.
1988
592 prel.
107.5 prel.
1989
550 est.
60.0 est.
*Source: Stephen A. Rhoades, "Mergers and Acquisitions by Commercial
Banks, 1960-1983," Staff Studies# No.142 (Federal Reserve Board, January
1985) and updates. Numbers do not include acquisitions of failed banks.
prel.-Figures are preliminary.
est.-Estimate.
Table A-2
Mergers and Acquisitions of Commercial Banking Organizations
where Both Partners were Over $1 billion in Deposits (1980-1990)*
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
Number Large
Acquisitions
0
1
2
6
14
7
20
18
18
6
7
Number Large
_
0
0
0
1
5
3
12
13
12
3
3
* Does not Include acquisitions of thrifts or failing firms.
Commercial banking organizations can either be commercial bank holding
companies or independent commercial banks.
Source: The American J ankes . The Bank Expansion Quarteely and Iks
Banking Volley. Report, The Secura Group, Washington, D.C. Banking.
Policy Reports. Daniélgon Associates Inc., Rockville# MD. Federal
Reserve,Builetin. ftedejeal.
AeservePress Releases and Resetve Banks.
Annual Call Reports, Stephen A. Rhoades, "Mergers and Acquisitions by
Commercial Banks, 1960*1983," Staff Study No. 142 (Federal Reserve
Board, January 1985).
Table A-3
Number of Banks and Banking Organizations, by Year (1970-1990)1
(United States as a Whole)
Year
Banks
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
13,502
13,603
13,722
13,965
14,217
14,372
14,396
14,397
14,374
14,668
14,737
14,717
14,706
14,646
14,636
14,587
14,379
13,870
13,303
12,901
12,526
Banking Organizations
12,644
12,586
12,464
12,404
12,368
12,396
12,403
12,398
12,384
12,727
12,679
12,515
12,261
11,950
11,643
11,377
10,872
10,470
10,183
9,908
9,688
Number of
Banking Offices'
31,20j*
34,042
36,140
38,051
39,957
41,304
44,343
45,736
46,817
48,530
50,162
51,796
52,599
52,883
53,364
54,457
55,894
57,336
54,965
59,842
Source: Bank Call Reports
l. Banking organizations are consolidated in the case of bank holding
companies, and numbers refer to FDIC-insured commercial and savings
banks.
2.
Number does not include branches with deposits of zero.
3. Number missing due to lack of data in 1971.
Table A-4
The Asset Size Distribution of Banking Organizations, by Year (1970-1990)1
(United States as a Whole)
3
Year
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
3
CR50
CR100
41%
40%
40%
41%
42%
41%
40%
40%
41%
38%
38%
38%
40%
41%
42%
43%
45%
46%
48%
49%
50%
50%
50%
50%
51%
52%
51%
50%
50%
51%
47%
48%
48%
51%
52%
53%
55%
57%
59%
60%
61%
62%
Number of Organizations by Size
<100M
10,885
10,673
10,411
10,325
10,388
10,424
10,382
10,255
10,191
10,327
10,407
10,293
10,055
9,722
9,360
9,053
8,526
8,229
7,948
7,630
7,183
100-300M
1,149
1,263
1,354
1,363
1,288
1,293
1,335
1,434
1,479
1,552
1,444
1,429
1,437
1,444
1,510
1,553
1,583
1,494
1,478
1,529
1,685
300M-1B
392
409
438
442
421
415
422
435
430
495
490
462
443
463
450
448
441
441
448
449
507
2
(1989 dollars)
1-10B
>101
197
221
237
249
247
242
238
247
257
323
311
304
296
288
284
276
263
245
250
243
253
21
20
24
25
24
22
26
27
27
30
27
27
30
33
39
47
59
61
59
60
60
Source: Bank Call Reports.
l. Size is measured by consolidated domestic banking assets. Banking
organizations refer to FDIC-insured commercial banking organizations
and savings banks.
2. Figures are adjusted using the consumer price index.
3. CR50 and CR100 denote the proportion of domestic banking assets
accounted for by the largest 50 and 100 banking organizations,
respectively.
Table A-5
Entry and Exit in Banking, 1980-1989
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
Total
____________ Number____________
Failure of
New
FDIC-Insured
Mergers and
Bank Branches
Banks
Acguisitions
Openings Closings
Banks
2,397
287
10
188
206
2,326
364
199
10
359
42
422
316
1,666
443
432
366
48
1,320
567
400
79
553
1,405
889
553
1,480
617
318
120
1,387
763
248
145
625
212
203
710
1,117
960
234
592 prel.
220
1,676
1,082
N.A.
201 prel.
208
1,730 prel. 687 prel
4,434
16,504
6,659
2,700
1 085
,
Sources : From Stephen A. Rhoades, "Commercial Banking: Two Industries and A
Laboratory for Research." New bank data for 1980-1987 are from Dean F. Amel,
"Trends in Banking Structure since the Mid-1970s," Federal Reserve Bulletin
(March 1989), p.124 and for 1988 and 1989 the data are from the Federal Reserve
Board, Annual Statistical Digest, 1988 and preliminary Digest data for 1989.
Failure data are from Annual Reports of the Federal Deposit Insurance Corporation
and statistical releases. Mergers and acquisitions data are from Stephen A.
Rhoades, "Mergers and Acquisitions by Commercial Banks, 1960-1983," Staff
Studies, No.142 (Federal Reserve Board, January 1985) and updates. Branch
openings and closings are from the Federal Reserve Board, Annual Statistical
Digest, relevant years and preliminary data for 1989.
est.-Figures are estimated.
prel.
-Figures are preliminary.
N.A.-Not Available
Table A - 6
Average Three-Firm Concentration Ratios for
Urban and Rural U.S. Banking Markets, 1976-1990*
Year
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
_Average three-firm concentration ratio
Urban markets
Rural-roa
68.45
90.06
67.79
89.97
67.29
89.88
66.78
89.75
66.39
89.65
66.04
89.45
65.83
89.38
65.92
89.41
66.34
89.44
66.71
89.47
67.51
89.47
67.67
89.53
67.78
89.68
67.51
89.70
67.35
89.59
1 . Concentration ratios measured in deposits.
Source: Summary of Deposits.
APPENDIX B
LEGAL AND ECONOMIC FOUNDATIONS FOR THE BOARD'S ANALYSIS
OF COMPETITION IN BANK MERGERS AND ACQUISITIONS
This appendix provides an overview of the legal and economic
foundations for bank merger analysis at the Federal Reserve.
I. Legislative and Judicial Foundations
Prior to 1960, the Board and other federal bank regulators
were required to review applications for bank mergers and acquisitions
but there was no meaningful requirement to assess the competitive
effects.1
The Board’s responsibility for assessing the competitive
effects of mergers stems from the Bank Merger Act passed by Congress
in 1960 and the Bank Holding Company Act (195 6).
As originally
enacted, the Bank Merger Act required that
the appropriate agency shall also take into
consideration the effect of the transaction on
competition (including any tendency toward monopoly)...
The original Bank Merger Act and Bank Holding Company Act,
however, did not specify what standards the banking agencies should apply
in assessing the competitive effects of a bank merger or acquisition and
it was unclear whether the antitrust laws were applicable.
The Supreme
Court clarified this matter in the Philadelphia National Bank case (1963),
wherein the Court clearly held that the antitrust laws, and in particular
section 7 of the Clayton Act (1914), apply to banking.
Specifically, the
Court stated that
1.
An assessment of competitive effects of bank holding company
acquisitions (in contrast to mergers) was required by the Bank Holding
Company Act of 1956. Since, however, prior to 1960 there were very
few significant holding company acquisitions and most bank
consolidations were accomplished by merger, the competitive
requirement of the Bank Holding Company Act was rarely applied and
thus received little attention.
B-2
[Section 7] does require, however, that the forces of
competition be allowed to operate within the broad
framework of governmental regulation of the industry.
The fact that banking is a highly regulated industry
critical to the Nation’s welfare makes the J>lay of
competition not less important but more so.
The Court’s opinion regarding the applicability of section 7 of
the Clayton Act to banking was reaffirmed by the Congress in 1966 when it
amended both the Bank Merger Act (1960) and the Bank Holding Company Act
(1956), and when it passed the Change in Bank Control Act (1978).
The
amendments introduced the language of section 7 into the banking laws’
provisions for the regulatory assessment of competition.
Specifically,
section 7 states:
That no corporation engaged in commerce shall acquire,
directly or indirectly, the whole or any part of the
stock or other share capital and no corporation subject
to the jurisdiction of the Federal Trade Commission
shall acquire the whole or any part of the assets of
another corporation engaged also in commerce, where in
any line n-F rr>mingrpg in any section of the country, the
effect of such acquisition may be substantially to
lessen competition or to tend to create a monopoly.
[emphasis added]
While the standards of section 7 are broad, two points are
explicit:
(1) both product and geographic markets (i.e., "...in any line
of commerce in any section of the country...”) must be defined and (2)
competition must be assessed or measured (i.e., "...the effect of such
acquisition may be substantially to lessen competition...").
These two
requirements of section 7 are not only legal requirements but reflect the
fundamental steps that are required in a good economic analysis of
2.
(1963).
United States v. Philadelphia National Bank, 374 U.S. 321, 372
B-3
competition.
Thus, the law’s requirements do not impose a purely legal
exercise on bank regulators.
Analytical Framework
In addition to clarifying the applicability of the antitrust laws
to banking, the Philadelphia National Bank case is important in three
other respects.
First, the Court clearly accepted the economic theory and
evidence suggesting that market structure, as measured by concentration
and market shares, is a major factor in assessing competition in a market.
The Supreme Court observed that
a merger which produces a firm controlling an undue
percentage share of the relevant market and results in a
significant increase in the concentration of firms in
that market is so inherently likely to lessen
competition substantially that it must be enjoined in
the absence of evidence clearly showing that the mejger
is not likely to have such anticompetitive effects.
Second, the Court accepted the evidence and argument that
important classes of customers are locally limited and therefore the local
geographic area is the relevant geographic market for an analysis of
competition in banking.
According to the Court,
Large borrowers and large depositors...may find it
practical to do a large part of their banking business
outside their home community; very small borrowers and
depositors may, as a practical matter, be confined to
bank offices in their immediate neighborhood...
Individuals and corporations typically confer the bulk
of their patronage on banks in their local community;
they find it impractical to conduct their banking
business at a distance...The factor of convenience
localizes banking competition as effectively as high
transportation cost in other industries.
3. IMd. , p. 363.
4. Ibid.. p. 360.
B-4
Third, the Court found that commercial banking services as a
whole constitute a distinct and relevant product market for analyzing the
competitive effect of a merger.
We agree with the District Court that the cluster of
products (various kinds of credit) and services (such as
checking accounts and trust administration) denoted by the
term "commercial banking" composes a distinct line of
commerce. Some commercial banking products or services are
so distinctive that they are entirely free of effective
competition from products or services of other financial
institutions; the checking account is in this category.
Others enjoy such cost advantages as to be insulated within
a broad range from substitutes furnished by other
institutions. For example, commercial banks compete with
small-loan companies in the personal-loan market; but the
small-loan companies’ rates are invariably much higher than
the banks’, in part, it seems, because the companies’
working capital consists in substantial part of bank loans.
Finally, there are banking facilities which, although in
terms of cost and price they are freely competitive with
the facilities provided by other financial institutions,
nevertheless enjoy a settled consumer preference,
insulating them, to a marked degree, from competition; this
seems to be the case with savings deposits. ’
In short, the Bank Merger Act and the Bank Holding Company Act as
interpreted by the Court and as amended by Congress in 1966, require the
Board to assess the competitive effects of bank merger and acquisition
applications according to the standards of section 7 of the Clayton Act.
The economic evidence and analysis presented to the Supreme Court in
Philadelphia National Bank (1963) led the Court to conclude that in
assessing competition, the relevant product market is commercial banking
5. Ibid., pp. 356-357.
6. As will be noted later, the courts recognize that marketplace
and legislative developments have changed the competitive environment
for banking services since Philadelphia National Bank. However, the
courts continue to hold that fundamental changes in the analytical
approach used in Philadelphia National Bank must be based on
systematic empirical evidence.
B-5
as represented by the cluster of banking products and services, the
relevant geographic market is a local area, and market structure is an
important determinant of competition.
Since Philadelphia National Bank, the courts have indicated a
willingness to consider the argument that, as a result of changes in the
financial sector, commercial banking services as a whole are not a single
product line and that competition may extend beyond local markets.
However, the courts have consistently held that arguments for changing the
basic market definition in banking must be based on persuasive economic
evidence.
For example, in United States v. Connecticut National Bank
(1974) the Supreme Court noted the increasing similarities of thrifts to
commercial banks and argued that,
at some stage in the development of savings banks it
will be unrealistic to distinguish them from commercial
banks...[and]...that point may well be reached when and
if savings banks become significant participants in the
marketing of-bank services to commercial
enterprises.
Similarly, in a recent case (1987) appealed by the Justice
Department, a Court of Appeals held that the District Court did not err
when it,
concluded that the government failed to factually
support its claim that existing circumstances in this
case warranted a departure from the definition of the
relevant product market as the cluster of banking
services traditionally offered in the commercial banking
industry adopted by the Supreme Court in United States
y .. JfaHfldfilghia National.JBank .
7. United States v. Connecticut National Bank. 418 U.S. 656 (1974).
8. United States v. Central State Bank. 817 F.2d 22 (6th Cir.,
1987).
B-6
In another context, the courts highlighted the weight they would
attach to economic evidence in Marine Bancorporation when the government
applied a "linkage" theory of oligopoly.
The Supreme Court noted:
Apart from the fact that the government’s statewide
approach is not supported by the precedents, it is
simply too speculative on this record. The government’s
underlying concern for a linkage or network of statewide
oligopolistic banking markets is, on this record at
least, considerably closer to 'ephemeral possibilities'
than to 'probabilities.' To assume, on the basis of
essentially no evidence, that the challenged merger will
tend to produce a statewide linkage of oligopolies is to
espouse a per se rule against geographic market
extension mergers like the one at issue here. No
section 7 case from^this court has gone that far, and we
do not do so today.
Conclusion
A combination of economic analysis, legislation, and court
decisions provide the rationale and analytical framework for the Board’s
evaluation of the competitive effects of bank mergers and acquisitions.
This is not a strictly fixed framework but may be changed to account for
institutional arrangements and technological advances that evolve.
However, it is evident that the courts expect systematic evidence and
analysis rather than casual empiricism to provide the foundation for
changing the framework.
At present, nonbank thrift institutions are
included in the Board’s competitive analysis on the basis of evidence
regarding their similarity to commercial banks, on a case-by-case basis.
Additionally, based on recent survey evidence covering consumers and small
businesses, the competitive analysis generally focuses on local geographic
9.
(1974).
United States v, Marine Bancorporation. 418 U.S. 602, 622-2.3
B-7
markets and treats banking as a line of commerce.
In some cases involving
larger banks or nonbank activities, the competitive analysis will focus on
specific product lines and use regional, national, or even global market
definitions.
II.
Market Definition
A major concern of antitrust policy is to assess the competitive
effects of mergers and, in particular, identify those mergers that have
the potential for creating market power.
The definition of the relevant
market over both product and geographic dimensions is crucial for any
economic analysis of the effects of a proposed merger.
Based on
microeconomic theory, defining a market is conceptually straightforward:
a market may be characterized as a group of buyers and sellers that
significantly influences price, quality, and production of specific
products or services, and the geographic market area is the area that
encompasses these buyers and sellers.
In a fundamental sense, markets are
defined in order to better predict the behavior of firms, where the
behavior of firms depends, in part, on competing producers of the same
product or close substitutes.
The Supreme Court has recognized the importance of the economic
concept of the market and instructed repeatedly that:
"Determination of
the relevant product and geographic markets is a 'necessary predicate’ to
deciding whether a merger contravenes the Clayton Act."10
Economic
10.
United States v. Marine Bancorporation. 418 U.S. 602, 618
(1974). quoting United States v. DuPont and Co. 353, U.S. 586, 593
.
(1957).
B-8
theory and evidence provide the foundation for determining the relevant
market.
While the concept of a market is fairly straightforward, it is
often difficult to use in practice as is the concept of money.1
1
Theory of the Market
The classical definition of a market arose from Alfred Marshall
who essentially assumed the existence of a product market and suggested
that geographic markets be delineated simply depending on whether "prices
of the same goods tend to equality, easily and quickly," with due
allowances made for transportation costs. 1 2
In the same vein, Joan
Robinson delineated a product market by all sellers of a commodity "which
may be regarded as homogeneous within itself," and notes that in defining
the relevant industry, there would be cases "where a commodity in the real
world is bounded on all sides by a marked gap between itself and its
closest substitutes." 1 3
Even Chamberlin, with his focus on
differentiated products, assumed the existence of a collection of
producers of fairly close substitutes. 1 4
A classically defined market has been interpreted to be that area
in which prices are linked to one another but independent of prices of
11. An analogy can be made to defining the money supply where the
concept is fairly straightforward, but, in practice, the distinctions
between components of Ml, M2, M3, etc., are not always clear.
12. Alfred Marshall, Principles of Economics. Book V (London:
Macmillan, 1920), p. 324.
13. Joan Robinson, The Economics of Imperfect Competition. 2nd ed.
(New York: St. Martin's Press, 1969), p. 17.
14. Chamberlin, however, later expresses doubts about the existence
of a meaningful real-world counterpart to this collection of
producers. Edward Chamberlin, The Theory of Monopolistic Competition.
5th ed. (Cambridge: Harvard University Press), p. 140.
B-9
goods not in the market, i.e., an area within which partial equilibrium
analysis is valid.
More formally, a market can be defined by those
products with non-trivial degrees of substitutability, as measured by
their cross-elasticities of demand and supply.15
The cross-elasticity
of demand relates the quantity demanded by consumers of one product to the
price of another: the cross-elasticity of supply relates the quantity
produced by a seller of one product to the price of another.
For example,
suppose a firm were to increase the price of its product and quantity
falls.
The loss in quantity would be greater if:
(1) the product has
better substitutes so that consumers would switch to other products; or
(2) other producers would expand output by greater amounts (or more
producers of another product would switch to producing the product). If
the loss in quantity to other producers (both of the same and similar
products) is large enough to offset the price increase so that profits
fall, cross-elasticities are considered to be high.
The concept of cross-elasticity then is fundamental to the
concept of a market because it provides the basis for assessing the degree
of substitutability between products.
The theory of markets based on
cross-elasticities, however, is not so precise as to allow one to draw
15.
Nicholas Kaldor, "Mrs. Robinson’s 'Economics of Imperfect
Competition,'" Economica. 1934, pp. 335-341; Robert Triffin,
Monopolistic Competition and General Equilibrium Theory (Cambridge:
Harvard University Press, 1940).
B-10
definite market boundaries for antitrust purposes.16
Apart from the
difficulties of estimating elasticities, it is not clear at what numerical
value a product becomes a close enough substitute to be included in the
market.
In addition, substitutability will depend on the relative prices
of alternative products where a large enough increase in price will
increase the willingness of consumers and producers to substitute.
Further, the length of time of adjustment will influence the cross elasticities as consumers and alternative producers may not be able to
switch quickiy to substitute products.
With respect specifically to geographic markets, both supply and
demand factors also have to be considered.
From a buyer’s perspective,
the market is that area within which the mobility of consumers will ensure
uniform prices, with allowance for transportation costs.
from a seller’s perspective will usually be larger. 1 7
The market area
Theoretically,
the classic market area would be expanded to cover the larger area
encompassed by sellers if they could enter easily into an area, 1 8 even
if demanders are limited to relatively small areas.
Again, for antitrust
16. See Corwin D. Edwards, "Economic Concepts and Antitrust
Litigation: Evolving Complexities," Antitrust Bulletin 19 (1974), pp.
295-319; Richard A. Posner and Frank H. Easterbrook, Antitrust-Cases.
Economic Notes, and Other Materials, 2nd ed., (St. Paul: West
Publishing Co., 1981); Phillip Areeda and Donald F. Turner, Antitrust
Law: An Analysis of Antitrust Principles and Their Application.
(Boston: Little, Brown & Co., 1980); and George J. Stigler, "The
Economist and the Problem of Monopoly," American Economic Review 72
(1982), pp. 1-11.
17. George J. Stigler, The Theory of Price. 3rd ed., (New York:
Macmillan, 1966) p. 85.
18. Transportation costs can create a wedge between prices in two
areas. Prices cannot differ by more than transportation costs if the
two areas are integrated. If prices differ by more, the two areas are
considered to be separate. See George J. Stigler, Ibid. p. 85.
.
B-ll
purposes, a time framework and a price range have to be imposed on the
economic market. 19
In sum, the cross-elasticities between products is important
because it helps to measure the ability of a sole seller of one of the
products to raise his price without reducing his quantity so much as to
lower profits.
As a consequence, when calculating market share or
concentration to help to assess market power, one should include in the
market all products that have high cross-elasticities with respect to the
product because it is this group of sellers who might have the power to
increase their profits by merging or colluding. 2 0
At the same time, one
cannot include in the market the infinite range of possibilities that in
some aspects might be interchangeable and yet still retain any meaning in
the market concept. 2 1
Consequently, a market delineated for antitrust
purposes must be defined with respect to a given time frame and price
range.
19. In particular, one must make assumptions about a reasonable time
period so as to include those producers that can substitute production
in the short run without significant new investment in plant,
equipment, and labor. See F.M. Scherer, Industrial Market Structure
and Economic Performance, 2nd ed. (Boston: Houghton Mifflin, 1980).
20. Richard A. Posner, Antitrust Law - An Economic Perspective
(Chicago: University of Chicago Press, 1976) p. 126.
21. The Supreme Court has recognized that a market should be
restricted to those producers that might have a "direct and immediate"
effect on competition. See Times-Picayune Publishing Company v.
United States. 345 U.S. 594, 612 (1953) and United States v.
Philadelphia National Bank. 374 U.S. 321, 357 (1963).
B-12
Application of the Market Concept
For public policy purposes, the economic theory of the market
probably has been used most frequently in connection with antitrust
enforcements
The courts and regulatory agencies have attempted to
define the relevant market areas for antitrust cases based on the concept
of cross-elasticity although they have not always been consistent in
market determination. 2 3
As noted, cross-elasticities are difficult to
measure and interpret, and are relevant to defining antitrust markets only
for a given time period and price range. 24
Nevertheless, several cases
are cited here to illustrate the practical difficulties encountered and
the typical factors considered when defining the relevant market in terms
of product and geographic dimensions.
22. The Department of Justice in its Merger Guidelines has suggested
a broad framework for defining markets. Essentially, a market is
defined by a group of products and an associated geographic area in
which the exercise of market power would be feasible, i.e., where a
hypothetical, profit-maximizing firm could impose a "small but
significant and nontransitory" increase in price above prevailing or
likely future levels. What constitutes a "small but significant and
nontransitory" price increase is of course subjective, and the
Department of Justice has proposed that in most contexts it will be
interpreted as a price increase of 5 percent lasting one year. (See
Department of Justice Merger Guidelines, 1984, p. 4. See also Federal
Trade Commission 1982 Statement on Horizontal Mergers.)
23. For a comprehensive review of judicial application of the market
concept, see ABA, Antitrust Section, Monograph No. 12, Horizontal
Mergers: Law and Policy. 1986.
24. For criticisms of cross-elasticity as a basis for market
definition, see e.g., Kenneth Boyer, "Industry Boundaries," in T.
Calvanit and J. Siegfried eds. Economic Analysis and Antitrust Law
(Boston: Little, Brown & Co., 1979); Kenneth Elzinga, "Defining
Geographic Market Boundaries," Antitrust Bulletin (1981), pp. 739-746;
Gregory Werden, "Market Delineation and the Justice Department's
Merger Guidelines," Duke Law Journal (1983), 514.
B-13
1
.
Product Markets
The Supreme Court initially recognized the role of cross
elasticities in the well-known cellophane case, where E.I. duPont de
Nemours was charged with monopolizing interstate commerce in cellophane in
violation of section 2 of the Sherman Act. 2 5
In defining the relevant
product market for determining the control of price and competition, the
Supreme Court considered the cross-elasticity of demand between products:
"If a slight decrease in the price of cellophane causes a considerable
number of customers of other flexible wrappings to switch to cellophane,
it would be an indication that a high cross-elasticity of demand exists
between them; that the products compete in the same market."
The Supreme
Court also addressed the issue of defining the market in merger cases
under section 7 of the Clayton Act in Brown Shoe and stated that the test
of the cellophane case (reasonable interchangeability of use or the crosselasticity of demand) was applicable in merger cases. 2 6
Further, the
court noted that the cross-elasticity of production facilities might be an
important factor in defining a product market.
Based on duPont and Brown Shoe, the courts and regulatory
agencies have determined both narrow and broad product markets. For
example, a product market might consist of a group or cluster of products
where the products are related and are usually produced or sold in a full
line by firms in the market.
In United States v. Grinnell Corp. the
.
Supreme Court decided that the cluster of services--protection of property
25. United States v. E.I. duPont de Nemours & Co. 351 U.S. 377
(1956).
26. Brown Shoe Co. v. United States. 370 U.S. 294 (1962).
B-14
(burglary and fire) through use of a central service station--constituted
a distinct market.
Although there were alternative providers of burglary
protection and fire protection separately, and without a central service
station, these services differed in "utility, efficiency, reliability,
responsiveness, and continuity,"2 7 and thus did not meet the demand
interchangeability test of the Cellophane case.
In Grand Union, the
Federal Trade Commission alleged that the product market was supermarket
sales, excluding convenience store sales. 2 8
They argued that
supermarkets were distinguished from other retail food stores on the basis
of amount of dollar sales and physical size.
The courts also have found
that acute hospital services constitute a cluster of services. 29
Though the courts have recognized and sought to use the economic
concept of a market, they are not always consistent in their product
market definitions.
For example, in United States v. Aluminum Company of
America. the Supreme Court found that the combination of bare and
insulated aluminum conductor products constituted a "line of commerce,"
but that copper conductor, because of its relatively high price, is
separable for the purpose of analyzing the competitive effect of the
merger.30
However, in United States v. Continental Can Co.. the Supreme
Court found that glass containers and metal cans constituted a single
27. 384 U.S. 563, 1966.
28. The Grand Union Co.. 102 F.T.C. 812 (1983).
29. American Medicorp v. Humana. Inc.. 445 F. Supp. 589 (E.D. Pa.
1977); American Medical International. Inc.. 3 Trade Reg. Rep. (CCH)
Para. 22,170 (FTC July 2, 1984); Hospital Corporation of America. 3
Trade Reg. Rep. (CCH) Para. 22,301 (FTC October 25, 1985).
30. 371 U.S. 271, 1964.
B-15
market because competition between metal and glass containers was
"insistent, continuous, effective and quantity-wise very substantial." 3 1
2.
Geographic Markets
Cross-elasticity, as a legal standard to define geographic
markets, was set out in Brown Shoe. Most cases, however, have relied on
the standard set in Tampa Electric, which defines geographic markets based
on the "market area in which the seller operates and to which buyers can
practicably turn for supplies." 3 2
This standard has focused attention
on shipping and purchasing patterns to help define geographic markets.
Geographic markets have been found to be local, regional, and
national.
In grocery retailing, markets were considered to be Standard
Metropolitan Statistical Areas (SMSAs), cities or towns3 3 because
convenience of location is considered to be essential to effective
service. 34
Similarly, in Brown Shoe, the relevant geographic markets
for shoe retail are the separate cities and surrounding areas in which the
parties sell shoes.
The courts also have found that markets are local in
hospital care. 3 5
31. 378 U.S. 441, 1964.
32. Tampa Electric Co. v. Nashville Coal Co.. 365 U.S. 320, 327
(1961) .
33. An SMSA is determined by the political subdivisions (normally
counties) where at least 15 percent of the labor force commutes to the
central core city or town. SMSAs are defined by the Office of
Statistical Policy, Department of Commerce.
34. United States v. Von’s Grocery Co.. 384 U.S. 270, 461 (1966);
EEC v. Food Town Stores. Inc.. 539 F.2d 1339, 1344-45 (4th Cir. 1976);
and The Grand Union Co.. 102 F.T.C. 812 (1983).
35. See e.g., United States v. Hospital Affiliates International.
Inc. . 1980-1981 Trade Cas. (CCH) Para. 63,721 (E.D. La. 1980);
Hospital Corporation of America . 3 Trade Reg, Rep , (CCH) Para. 22,301
(FTC October 25, 1985).
B-16
On the other hand, the court found in Brown Shoe that shoe
manufacturing is a nationwide market because shoes are distributed
nationwide.
In addition, liquid bleach and other products sold in retail
grocery stores are considered to be national markets, and beer is
generally considered a regional market, depending on the distribution
system. 36
Determination of Banking Markets
In banking, the economic evidence and analysis brought before the
regulatory agencies and the Supreme Court in Philadelphia National and
subsequent cases have led to the adoption of a product market limited to
the cluster of commercial banking services and a local geographic
market. 3 7
The product market had been defined using primarily the
concept of the cross-elasticity between the cluster of commercial banking
products and financial products offered by alternative financial
institutions within a given time period and a plausible range of prices.
The geographic market was defined as local based on the Tampa Electric
standard, which encompasses that "area where the seller operates and to
which buyers can practicably turn for supplies."
This, of course, applies
36. See e.g., United States v. Pabst Brewing Co.. 384 U.S. 546
(1966); F. & M. Schaefer Corp. v, C. Schmidt & Sons. 597 F.2d 814 (2d.
Cir. 1979); FTC v. Procter & Gamble Co.. 386 U.S. 568 (1967).
37. United States v. Philadelphia National Bank. 374 U.S. 321, 359
(1963). See also United States v. Connecticut National Bank. 418 U.S.
656 (1974); United States v. Phillipsburg National Bank & Trust Co..
399 U.S. 350 (1970). For a discussion of the theoretical issues
raised in defining banking markets, and in particular geographic
markets, see John D. Wolken, "Geographic Market Delineation; A Review
of the Literature." Staff Studies, No. 140, Board of Governors of the
Federal Reserve System, 1984.
B-17
to the vast majority of banks and services that are provided to most
consumers and small businesses, rather than to the small number of U.S.
global banks that compete in international markets for the business of
major corporations in competition with the capital and commercial paper
markets.
The courts have indicated recently a willingness to consider
arguments that, as a result of changes in the financial sector, the
cluster of commercial banking services is not a single product line and
banking markets are not local. 3 8
But as shown by opinions in
Connecticut National and more recently, Central. S
. tate, the courts are not
willing to expand the relevant market without empirical evidence. 3 9
Recent empirical evidence, however, indicates that the cluster of
commercial banking services may still be the relevant product market.
Evidence from two recent surveys indicates that small businesses and
38. For a discussion of whether a cluster of services is still the
relevant product line, see e.g., Michael Bleier and Robert Eisenbeis,
"Commercial Banking as the 'Line of Commerce’ and the Role of
Thrifts," 98 Banking Law Journal (1981), 374; Bronsteen, "Product
Market Definition in Commercial Bank Merger Cases," 30 Antitrust
Bulletin (1985), p.677; Alan J. Daskin, "Horizontal Merger Guidelines
and the Line of Commerce in Banking," 30 Antitrust Bulletin (1985),
p.651. For a discussion of geographic markets, see Langenfeld and
McKenzie, "Financial Deregulation and Geographic Market Delineation,"
30 Antitrust Bulletin (1985), p.695.
39.
United States v. Central State Bank, et al.. (1987). In this
case, the Department of Justice proposed to treat transactions
accounts and small business loans as separate product lines. The
District Court found, and the Appeals Court upheld, that the
Department failed to factually support its claim that existing
circumstances in this case warranted a departure from the definition
of banking markets in Philadelphia National.
B-18
households tend to cluster their purchases of financial services. 4 0
Small businesses tend to obtain multiple financial services from the same
financial organization from which they obtain checking services, most of
which are commercial banks.
In contrast, small businesses typically
obtain only one financial service from nondepository and nonlocal
suppliers.
For Japanese households, 69 percent cluster their savings
deposits, cash withdrawal, and remittance at one financial
institution.41
Further empirical evidence on the geographic scope of banking
markets suggests that banking markets are still local for most consumers
and small businesses.
From recent surveys, and in particular, the
national survey of small businesses, the data indicate that the vast
majority of small businesses obtain their financing from local financial
institutions, primarily commercial banks. 4 2
The survey of Japanese
households indicated that 80 percent listed "geographic proximity," as the
first reason for selecting their primary financial institution. 4 3
Other
empirical studies that indicate differences in deposit and loan rates
across MSAs, and non-MSA counties, or cities, also suggest that banking
40. See Gregory Elliehausen and John Wolken, "Banking Markets and
the Use of Financial Services by Small and Medium-Sized Businesses,"
Staff Studies 160 (Federal Reserve Board, September 1990), p. 15.
41. See Akira Kurukawa, "Retail Banking and Consumer Choice in
Japan," mimeo (Institute for Posts and Telecommunications Policy,
March 1990), esp. pp. 31 and 44.
42. Elliehausen and Wolken, op. cit. pp. 16-19. In addition, 10
small surveys conducted by Reserve Banks during 1988-1990 in
connection with BHC bank acquisition applications suggest that banking
matters are local.
43. Furukawa, op. cit.. pp. 31 and 44.
B-19
markets are local. 44
Conclusion
Cross-elasticities of demand and supply provide a theoretical
foundation for market definition.
can be difficult.
In practice, however, defining a market
Market definitions adopted by the courts and regulatory
agencies in antitrust cases have not always been consistent and illustrate
the difficulties in market definition.
Numerous court decisions have held
that the relevant banking market is a cluster of banking services in a
local market area where consumers and small businesses obtain their
financial services.
The finding of a local geographic market for banking
is not unique in that for nonbanking industries, geographic markets have
also been defined as local.
In any event, recent survey evidence
indicates that the relevant banking market remains a cluster of banking
services in a local market area.
This does not apply to the global
competition faced by a few very large U.S. banks who face competition from
foreign banks, commercial paper, and securities markets in their dealings
with large, often international corporations.
44.
See Allen N. Berger and Timothy H. Hannan, "The PriceConcentration Relationship in Banking," Review of Economics and
Statistics (May 1989), pp. 291-99; Timothy H. Hannan, "Bank Commercial
Loan Rates and the Market for Commercial Loans," Journal of Banking
and Finance 15 (February 1991), pp.133-149; Stephen A. Rhoades, "Local
v. National Banking Markets: Evidence from an Analysis of Mortgage
Loan Rates in 20 Cities," mimeo (Federal Reserve Board, 1990); and
Timothy H. Hannan, "The Functional Relationship between Prices and
Concentration: The case of the Banking Industry," mimeo (Federal
Reserve Board, 1991) .
B-20
in.
Assessing Competition:__The . ole of Market Structure
R
The economic rationale for focusing on market structure (whether
local, regional or national in scope) in assessing competition stems from
a lengthy history of theoretical and empirical work.
Theory
Industrial organization, the branch of applied microeconomics
that seeks to explain the behavior of firms in a market, relies
principally on economic models that fall within a broad framework usually
referred to as the "structure-conduct-performance paradigm."
This
framework, in which market structure is the key element, was developed out
of theoretical work by Chamberlin45 and others in the 1930s and was
first advanced by Mason. 4 6
Bain was the first to undertake significant
development and testing of the paradigm. 4
7
Despite its name, the structure-conduct-performance (S-C-P)
paradigm is usually presented as having four components connected by a
causal relationship.48
The basic conditions underlying an industry
45. E. H. Chamberlin, The Theory of Monopolistic Competition
(Cambridge: Harvard University Press, 1933) (8th ed., 1962).
46. Edward S. Mason, "Price and Production Policies of Large-Scale
Enterprise," American Economic Review. Vol. 29, No. 1, Pt.2 Supplement
(March 1939), pp. 61-74.
47. Joe S. Bain, Barriers to New Competition (Cambridge: Harvard
University Press, 1956); and Industrial Organization. 2nd ed. (New
York: John Wiley & Sons, 1968).
48. For textbook explanations of the S-C-P model, see F. M. Scherer,
Industrial Market Structure and Economic Performance. 2nd ed.
(Chicago: Rand McNally), 1980, p. 4; Douglas F. Greer, Business.
Government. and Society (New York: Macmillan, 1983), p. 15; Michael
Waterson, Economic Theory of the Industry (Cambridge: Cambridge
University Press, 1984), p. 3; and William G. Shepherd, The Treatment
of Market Power (New York: Columbia University Press, 1975), p. 12.
B-21
affect market structure.
Market structure affects the conduct of firms in
the market and conduct in turn influences firm performance.
Basic market
conditions include both the level of demand and the shape of the demand
curve ("tastes and preferences") and the shape of the supply curve
("technology" or "costs"). Basic conditions also include any legal
constraints on firm actions.
Market structure refers to the number and
size distribution of firms in the market (the "actual" competitors) and
the ease or difficulty of entry into the market by nonincumbent firms (the
"potential" competitors). Conduct includes the degree to which firms in
the market compete or collude with each other and the choices of
competitive strategies by firms.
Performance is typically measured by the
level of profits or prices, by the rate of firm growth, or by other
variables thought to reflect firm goals.
From the initial work of Mason it has been recognized that the
S-C-P framework presents a simplified view of how markets actually work.
In particular, it has been recognized that the one-way flow of causation
described above does not always hold.
But economists differ on the
importance of the feedback effects of firm performance on conduct and of
both performance and conduct on market structure.
Some economists think
that the feedback effects are sufficiently weak that it is possible to
analyze firm performance while treating market structure as if it were
exogenous.
Other economists think that the feedback effects are strong
enough that any analysis that fails to treat both firm performance and
market structure as endogenous features determined by basic market
B- 22
conditions is likely to yield incorrect conclusions. 4 9
Empirical
evidence on the importance of feedback effects will be discussed below.
The use of market structure in analyzing and assessing
competition has a solid theoretical foundation.
There have been
significant developments in the field of economics relevant to such
matters (Industrial Organization) in recent years.
However, to date, much
of this work has been theoretical and has yet to provide much that is
empircally supportable or refutable that may be directly applied in
antitrust policy toward mergers.50
Since theory alone has generally
been viewed by the courts as inadequate for supporting public policy, a
review of relevant empirical evidence is appropriate.
Empirical Evidence
Empirical estimation of the S-C-P model has generally found that
higher levels of market concentration are associated with higher levels of
prices and profits.
However, because of measurement problems, the
49. For a statement of this view, see Almarin Phillips, "A Critique
of Empirical Studies of Relations between Market Structure and
Profitability," Journal of Industrial Economics (June 1976), pp. 241249 and for a theoretical treatment see William J. Baumol, John C.
Panzar and Robert D. Willig, Contestable Markets and the Theory of
Industry Structure (New York: Harcourt Brace Jovanovich, 1982).
50. For a rather complete and technical overview, see Richard
Schmalensee and Robert D. Willig, eds. The Handbook of Industrial
,
Organization, 2 vols. (New York: North Holland, 1989). For very
interesting reviews of this work, see Franklin M. Fisher, "Organizing
Industrial Organization: Reflections on The Handbook of Industrial
Organization," Brookings Papers: Microeconomics 1991 (Washington,
D.C.: Brookings Institution, 1991) pp.201-255; Alvin K. Klevorick,
"Directions and Trends in Industrial Organization: A Review Essay on
The Handbook of Industrial Organization." Ibid.. pp.241-264, and
Robert H. Porter, "A Review Essay on Handbook of Industrial
Organization." Journal of Economic Literature (June 1991), pp.553-572.
B-23
reliability and interpretation of these results have been subject to
question. 51
As noted above, market structure refers to both the actual
competitors in a market and the ease with which potential competitors can
enter the market.
In practice, actual competition is much easier to
measure than the threat of potential entry.
The number and size
distribution of firms (market concentration) is typically measured by
either the Herfindahl-Hirschman Index, which is the sum of the squares of
the market shares of all firms in the market, or by an n-firm
concentration ratio, which sums the market shares of the n largest firms
in the market. 52 Because the number of potential entrants, their
effects on pricing by the incumbent firms in a market and the probability
that they will enter a market are extremely difficult to gauge, potential
competition is normally measured by some indirect method.
For example,
the difficulty of entering a market may be measured by the average size of
the firms in that market since this average gives some idea of the amount
of capital that must be raised to enter the market and the scale of
operations that must be achieved to operate efficiently.
Conduct is even more difficult to measure than is potential
competition.
The degree to which firms compete with each other is usually
51. The inability to accurately measure basic tastes and technology
(e.g., demand elasticities) was a major reason for the development of
the S-C-P model (see Mason), op. cit. While structure, conduct and
performance are easier to quantify than underlying market conditions,
their measurement poses substantial difficulties.
52. The four-firm concentration ratio is the most commonly used
ratio in the industrial sector. In banking, the three-firm
concentration ratio is the standard.
B-24
not directly observable or quantifiable.
For this reason, economists have
tended to focus on the direct linkages between structure and
competition. 3
5
Firm performance can be measured in terms of any goal to which
the firms in a market are presumed to aspire.
Since most economic theory
assumes that firms are profit maximizers, profits is a logical measure of
firm performance.
However, there are many reasons why accounting profits
may be poor measures of the economic profits to which theory refers.
54
Some have argued that performance can be measured more accurately by
breaking profit into its component parts of revenues (price times
quantity) and costs.
Price is then used as the performance measure and
costs are controlled for in the estimation procedure.
The S-C-P model has typically been tested by estimating a simple
regression with the performance measure (e.g., profits) as the dependent
variable and measures of market structure (e.g., the Herfindahl Index) and
conduct as explanatory variables.
Such an estimation method assumes that
53. Measures of product differentiation, advertising and research
and development can give some idea of differences in conduct across
firms. When they are important elements of firm behavior, these
variables are often included along with structural variables in
empirical models.
54. Some go so far as to say that accounting profits are worthless
as measures of economic profits. See George J. Benston, "The Validity
of Profit-Structure Studies with Particular Reference to the FTC's
Line of Business Data," American Economic Review. Vol. 75, No. 1
(March 1985), pp. 37-67 and Franklin M. Fisher and John J. McGowen,
"On the Misuse of Accounting Rates of Return to Infer Monopoly
Profits," American Economic Review. Vol. 73, No. 1 (March 1983), pp.
82-97. For critical comments on these papers from a number of
economists defending the use of accounting data, see American Economic
Review. Vol. 77, No. 1 (March 1987), pp. 205-217 and American Economic
Review. Vol. 74, No. 3 (June 1984), pp. 492-508, respectively.
B-25
market structure and conduct can be treated as exogenous, i.e., that
feedback effects from performance on conduct and market structure are not
important and can be ignored.
Results from these regressions generally
show that higher market concentration is associated with higher profits
and higher prices.
Measures of entry barriers are also positively
correlated with higher prices and profits.55
The results of these studies have been given two explanations.
The "traditional" explanation is that greater market concentration gives
firms greater market power, so that by èither explicit or implicit
collusion they can raise prices and profits above competitive levels.
Thus, higher levels of concentration are thought to cause higher profits.
A "revisionist" explanation, first put forward by Demsetz in 1973, argues
that the observed concentration-profits relationship is not a causal one.
Demsetz argues that more efficient firms tend to grow larger than other
firms, so that in markets where firms differ in efficiency some
(efficient) firms will have large market shares and measures of market
concentration will be high.
The concentration-profits relationship is
therefore the result of efficient firms out-competing inefficient firms
and not the result of market power being used to generate monopoly
profits.
56
55. For reviews of these studies, see Scherer, op. clt.. or Leonard
W. Weiss, "The Concentration-Profits Relationship and Antitrust," in
H. J. Goldschmid, H. M. Mann and J. F. Weston, eds., Industrial
Concentration: The New Learning (Boston: Little, Brown and Co..
1974).
56. Harold Demsetz, "Two Systems of Belief about Monopoly," in
Goldschmid, Mann and Weston, Ibid.
B-26
Two directions in recent empirical work have tried to distinguish
between the traditional and revisionist hypotheses.
The first direction
uses price data to distinguish between the hypotheses, since the
traditional hypothesis predicts a positive correlation between prices and
market concentration, while the revisionist hypothesis predicts a negative
correlation.
Such studies tend to support the traditional view over the
revisionist explanation. 5 7
The second direction attempts to account for the endogeneity of
conduct and structure variables implied by the feedback effects through
the estimation of simultaneous-equations models.
Studies that
simultaneously estimate the determinants of profits and the determinants
of variables measuring firm conduct and market structure have generally
found that results are qualitatively the same as those for the simple
single-equation models described above. 5 8
Estimation of the S-C-P model for the U.S. banking industry has
found that a statistically significant and positive relationship exists
57. See Allen N. Berger and Timothy H. Hannan, "The PriceConcentration Relationship in Banking," Review of Economics and
Statistics. Vol. 71, No. 2 (May 1989), pp. 291-299; Timothy H. Hannan,
"Bank Commercial Loan Rates and the Market for Commercial Loans,"
unpublished paper (1990); and Stephen A. Rhoades, "Local v. National
Banking Markets: Evidence from an Analysis of Mortgage Loan Rates in
20 Cities," mimeo (1990).
58. See Jeffrey A. Clark, "Single-Equation, Multiple-Regression
Methodology: Is It an Appropriate Methodology for the Estimation of
the Structure-Performance Relationship in Banking?" Journal of
Monetary Economics. Vol. 18 (November 1986), pp. 295-312.
between market concentration and profitability. 5 9
These studies focus
on the local banking markets in which most U.S. banks compete rather than
on the global banking markets in which the very large banks compete
against foreign banks, the commercial paper market, securities firms, and
so forth.
Moreover, because virtually all research on economies of scale
in banking has found that such economies exist only for small banks, this
relationship is less likely to be due to efficiency differences than in
other industries.60
However, the magnitude of the effect of
concentration on profits seems to be smaller in banking than in many other
industries.61
Studies of bank prices find a positive and economically
meaningful relationship between prices and market concentration.
59. Most banking S-C-P studies have controlled for differences in
potential competition by taking into account differences in the legal
ability to expand geographically by branching or bank holding
companies across states. A few studies have used legal branching
restrictions to try to measure directly the number of potential
entrants into banking markets. See Timothy Hannan, "Limit Pricing and
the Banking Industry." Journal of Money. Credit and Banking. Vol. 11,
No. 4 (November 1979), pp. 438-446.
60. For a general review of the literature on scale economies in
banking, see R. Alton Gilbert, "Bank Market Structure and Competition:
A Survey," Journal of Money. Credit and Banking. Vol. 16, No. 4, Pt. 2
(November 1984), pp. 617-645. For more recent reviews of more
sophisticated econometric studies, see Allen N. Berger, Gerald A.
Hanweck and David B. Humphrey, "Competitive Viability in Banking:
Scale. Scope, and Product Mix Economies," Journal of Monetary
Economics. Vol. 20, No. 4 (December 1987), pp. 501-520; and Allen N.
Berger and David B. Humphrey, "The Dominance of Inefficiencies over
Scale and Product Mix Economies in Banking," Finance and Economics
Discussion Series No. 107, Board of Governors of the Federal Reserve
System, January 1990.
61. For reviews of S-C-P studies in banking, see Stephen A. Rhoades,
"Structure-Performance Studies in Banking: An Updated Summary and
Evaluation," Staff Study No. 119, Board of Governors of the Federal
Reserve System, August 1982, and R. Alton Gilbert, op. clt.
62. See papers cited in footnote 57.
B-28
FttbXAc. golAcy
Empirical results from estimation of the S-C-P model have
generally been interpreted as indicative of a significant causal
relationship from greater market concentration to higher firm profits and
prices.
Antitrust policy has relied heavily on this interpretation in its
attempts to prevent the monopolization of U.S. markets.
Current
Department of Justice merger guidelines limit increases in market
concentration due to mergers.
While the guidelines recognize that ease of
entry and other factors affect the degree of competition within a market,
the core of the guidelines is a limit on the level and increase in the
Herfindahl Index that is allowed without triggering a Justice Department
examination and possible legal challenge to the merger.
The numerical
limits chosen are admittedly somewhat arbitrary but they serve the purpose
of "reduc[ing] the uncertainty associated with enforcement of antitrust
laws in this area."6 3
The guidelines are not adhered to rigidly; if
entry conditions or other factors indicate that a merger that violates the
numerical limits on concentration would not be anticompetitive, an
exception to the guidelines can be made.
For example, in 1984 the Justice
Department relaxed the numerical rules for banking mergers in recognition
of the increased competition banks are facing from other providers of
financial services.
It is notable, however, that a recent Court of
Appeals decision highlighted the importance of having solid empirical
63. U.S. Department of Justice Merger Guidelines. June 14, 1982, p.
1.
B-29
evidence before altering significantly the analytical framework for
assessing competition in banking. 64
Conclusion
The focus on market structure in assessing competition stems from
a large body of theoretical and empirical research.
Despite a continuing
debate over the causes of the relationship between market structure and
firm performance, antitrust policy has relied heavily on limiting market
concentration in its attempts to prevent the monopolization of U.S.
markets. Research on the banking industry generally supports the view
that more concentrated local banking markets are less competitive than
less concentrated markets.
There are, of course, a few very large global
U.S. banks that face competition on a global scale for very large
corporate customers, to which these research findings for local U.S.
markets do not directly apply.
iv.
Overall Conclusion
From this overview, it is clear that there is an extensive legal
and economic foundation for the Board’s assessment of the likely
competitive effects of bank mergers and acquisitions.
The analysis
conducted in individual merger cases is guided by this foundation in
64.
Justice
banking
of this
In United States v. Central State Bank (1987), the Department of
was defeated in its attempt to change the definition of
product markets. The Supreme Court declined to hear an appeal
case.
B-30
developing relevant data from a variety of sources including surveys,
interviews, on-site investigations, and various databases.
APPENDIX C
STUDIES RELEVANT TO THE IMPACT OF BANK MERGERS
I.
Bank Prices and Profits
Studies of the impact of mergers on the prices charged b y
banks and the profits earned by them may be usefully divided into two
types: those that do not look specifically at the actual effects of
specific mergers and those that do.
Of all the studies conducted
within the Federal Reserve System that are relevant to this issue, the
vast majority have been designed to draw inferences about the possible
effects of mergers in a way that does not look specifically at the
effects of individual past mergers.
A few studies, however, have
looked at the change in bank prices and profits resulting from
specific, individual mergers.
Each of these two types of studies are
reviewed in turn.
A.
Studies that do not Focus on Individual Bank Mergers
Studies of the relationship between bank profitability and
banking market concentration represent the oldest type of study having
potential relevance to the impact of bank mergers.
Typically, such
studies define local banking markets as Metropolitan Statistical
Areas, counties, or some other geographic entity chosen to represent
the area in which banks are presumed to compete for customers of the
more locally oriented bank products, such as small business loans and
certain types of deposits.
Measures of concentration are calculated
for each market thus defined.
Commonly used concentration measures
are the market deposit share of the largest three banks in the market
and the Herfindahl index, defined as the sum of the squared value of
each market share in the market.
Statistical tests are then performed
C-2
to determine whether banks operating in areas that register higher
levels of concentration also exhibit higher levels of profitability,
controlling for other things that may also influence firm
profitability.
One review of this very large literature was done by Stephen
Rhoades of the Board's staff.1 Mr. Rhoades concludes that most of
these studies find that banks operating in more concentrated markets
enjoy on average higher levels of profitability.
These findings are
consistent with the argument that banks in highly concentrated markets
charge higher loan rates and offer lower deposit rates, suggesting in
turn that mergers resulting in high levels of market concentration can
adversely affect bank customers.
Many have argued, however, that
these findings simply reflect the greater efficiency and lower costs
of the largest firms in concentrated markets rather than
noncompetitive behavior.
Because of this fundamental disagreement,
there is little consensus concerning the meaning of this type of study
for merger policy.
Another type of study with relevance to merger policy
examines the relationship between market concentration and bank
prices, including various types of loan rates charged by banks and
different deposit rates offered by them.
Typically, these studies
define local banking markets and calculate measures of market
concentration in much the same way as do the studies that focus on
bank profitability.
Their primary difference is that instead of
determining whether banks in more concentrated markets exhibit greater
profitability, they seek to determine whether such banks charge loan
rates and offer deposit rates that are less attractive to bank
1.
Stephen A. Rhoades, "Structure-Performance Studies in Banking:
An Updated Summary and Evaluation," Staff Study #119, Board of
Governors of the Federal Reserve System, 1982.
C-3
customers.
Such studies have been made possible in part by Board
surveys that provide information on bank loan and deposit rates.
A
good review of these "price-concentration" studies in banking, as well
as similar studies applying to other industries, has been presented by
Leonard Weiss.
A number of studies of this type have been
conducted recently by members of the Board's staff.
3
All find some
evidence suggesting that if banking markets become highly
concentrated, bank customers tend to pay higher loan rates and/or
receive lower deposit rates than do bank customers in less
concentrated markets.
These price studies tend to be clearer in terms
of their implications for merger policy than are the profit studies,
since they suggest with less ambiguity that mergers resulting in
relatively high levels of banking market concentration can adversely
affect bank customers.
B.
Studies that Focus on Individual Bank Mergers
Whether or not specific past mergers have resulted in higher
loan rates, lower deposit rates, or in other ways disadvantaged
banking customers is a different question.
Banking organizations must
pass regulatory scrutiny before they are allowed to merge, and mergers
are typically not approved if they are judged likely to have a
2. Leonard W. Weiss, ed.. Concentration and Price (MIT Press:
Cambridge), 1989.
3. See, for example, Allen N. Berger and Timothy H. Hannan, "The
Price-Concentration Relationship in Banking," Review of Economics and
Statistics 71 (May 1989), pp. 291-99; Timothy H. Hannan, "Bank
Commercial Loan Markets and the Role of Market Structure: Evidence
from Surveys of Commercial Lending," Journal of Banking and Finance 15
(February 1991), pp. 133-49; Allen N. Berger and Timothy H. Hannan,
"The Price-Concentration Relationship in Banking: Reply," Review of
Economics and Statistics (forthcoming); Stephen A. Rhoades, "Local Vs.
National Banking Markets: Evidence from an Analysis of Mortgage Loan
Rates in 20 Cities," (mimeo) Board of Governors of the Federal Reserve
System, 1991; and Timothy H. Hannan, "The Structural Relationship
Between Prices and Market Concentration: The Case of the Banking
Industry," (mimeo). Board of Governors of the Federal Reserve System,
1991.
C-4
significantly adverse effect on competition.
Thus, studies of the
competitive impact of individual bank mergers in essence focus on the
issue of whether regulatory authorities have been correct in their
assessments.
Only a few studies have looked either directly or indirectly
at the pricing practices of banks before and after mergers.
The only
study to look directly at at the pricing effects of a large merger was
reported in 1987 by Frederick Furlong.
This study examined the
rates offered for Money Market Deposit Accounts and six-month time
deposit accounts of less than $100,000 both before and after the
merger between Wells Fargo and Crocker National.
After accounting for
the relationship between these rates and short-term money market rates
over time, the study finds no general tendency on the part of the
merged firm to offer depositors less attractive rates after the merger
than were offered before the merger.
Other studies have looked indirectly at the effect of mergers
on bank pricing by examining the reaction in a bank’s stock price
brought about by the announcement of a merger involving the bank's
competitors.
If observed mergers allow merged entities to raise loan
rates or lower deposit rates, the reasoning goes, then banks competing
in the same market should also benefit from the price changes.
A
finding of positive abnormal returns in the stock of banks competing
with merger participants occurring at the time of the announcement
would represent a finding consistent with this hypothesis.
In
general, those studies that have searched for such abnormal returns
have failed to find them, thus providing no evidence of
4.
Frederick T. Furlong, "Assessing Bank Antitrust Standards,"
Weekly Letter of the Federal Reserve Bank of San Francisco, May 15.
1987.
C-5
anticompetitive pricing brought about by past mergers.5
II. Costs and Efficiency
Another issue relevant to the effect of mergers concerns the
prospect that through merger, greater bank efficiency can be achieved,
thus yielding a healthier and potentially more competitive banking
firm.
As in the case of the effect of mergers on bank prices and
profits, studies that examine potential changes in bank costs or
efficiency may be divided into studies that do and do not look at the
effect of specific past mergers.
A.
Studies that do not Focus on Individual Bank Mergers
A great deal of research both within and outside of the
Federal Reserve System has been devoted to the question of how the
costs incurred by banks vary with bank size.
While not focusing on
individual mergers per se. this line of research can indicate whether
mergers, which make larger banks out of smaller ones, can be expected
to lower costs as a result of "economies of scale" or to raise costs
as a result of "diseconomies of scale."
These studies seek to answer
this question by examining how the costs incurred by banks vary with
bank size, controlling statistically for input prices and other
potential determinants of bank costs.
Some of the more recent studies
of this type conducted within the Federal Reserve System include a
study by Berger, Hanweck, and Humphrey in 1987 and a study by Berger
5.
See Christopher M. James and Peggy Wier, "Returns to Acquirers
and Competition in the Acquisition Market: The Case of Banking,"
Journal of Political Economy 95 (May 1987), pp. 355-70, and Frederick
T. Furlong, op. cit.
C-6
and Humphrey in 1990. 6
The most recent review of this large
literature has appeared in the Economic Review of the Federal Reserve
Bank of Richmond.
In general, these studies have not found
evidence of a significant cost advantage on the part of larger banks.
They find that scale economies, if they exist, are very small, and
most studies do not show such scale economies to exist beyond a small
to medium sized bank.
Thus, this line of research has not provided
strong evidence suggesting that large mergers in general can be
counted on to achieve substantial cost savings.
In a somewhat different approach, results of the 1990 study
by Berger and Humphrey, referred to above, suggest that banks may
differ considerably in their ability to control costs and that such
differences are far more important than those differences that might
be attributed to differences in institution size.
It thus appears
that substantial cost savings are possible for many banking
organizations, regardless of size.
B.
Studies that Focus on Individual Bank Mergers
In the past few years, a number of studies have attempted to
determine whether individual past mergers have resulted in significant
cost savings.
While such studies typically focus on the change in
noninterest expenses before and after the merger, changes in
profitability and market share are also sometimes examined.
In some
cases, before and after changes are compared to the same changes
6. Allen N. Berger, Gerald A. Hanweck, and David B. Humphrey,
"Competitive Viability in Banking: Scale, Scope, and Product Mix
Economies," Journal of Monetary Economics 20 (December 1987), pp. 50120, and Allen N. Berger and David B. Humphrey, "The Dominance of
Inefficiencies Over Scale and Product Mix Economies in Banking,"
Journal of Monetary Economics 28 (August 1991), pp. 117-48.
7. See David B. Humphrey, "Why do Estimates of Bank Scale Economies
Differ?" Economic Review. Federal Reserve Bank of Richmond,
September/October 1990.
C-7
observed concurrently in banking organizations that did not engage in
mergers.
In assessing the results of these studies, it is useful to
make a distinction between cost savings that are achieved simply by
shrinking the size of the resulting firm and cost savings that lower
costs attributable to a firm of a given size.
A number of studies of
recent mergers have found evidence that the combined entities after
merger were reduced in size relative to the size before merger of the
g
two merging organizations.
But with one or two exceptions, most
studies have not found strong evidence of a reduction in costs after
accounting for this size effect.
9
As noted above, however, it has also been found that banking
organizations vary considerably in terms of their ability to control
costs, no matter what their size.
This implies that for most banking
organizations, much can be done to reduce costs and create a more
healthy banking organization.
Estimates made by researchers at
the Federal Reserve Board indicate that if the lowest cost banks in
the country were to acquire the highest cost banks and if costs of the
acquired banking organizations were subsequently reduced to the level
of the acquiring banks, then annual cost savings of $17 billion would
8. See, for example, Ethan M. Heisler, "Savings Resulting from the
Acquisition of Irving Bank Corp. by The Bank of New York Co., Inc."
Federal Reserve Bank of New York, September 18, 1990; Donald T.
Savage, "Mergers, Branch Closings, and Cost savings," (mimeo) Board of
Governors of the Federal Reserve System, May 1991; and Dwight Crane
and Jane C. Linder, "Bank Mergers: Integration and Profitability,"
(mimeo) Harvard Graduate School of Business, 1991. Investigations of
the merger between Crocker National and Wells Fargo also indicate a
reduction in size after the merger.
9. See, for example, Donald Savage, op. cit.; Dwight Crane and
Jane C. Linder, op. cit.; and Stephen A. Rhoades, "Billion Dollar Bank
Acquisitions: A Note on the Performance Effects," (mimeo) Board of
Governors of the Federal Reserve System, 1990. More recent work at
the Board has also failed to find, in the case of a number of large
mergers, a reduction in costs after accounting for the size effect.
The study by Ethan M. Heisler (op. cit.) of the merger between Irving
Bank Corp. and The Bank of New York did find evidence of a reduction
in noninterest expenses beyond that attributable to the shrinkage of
the combined firm after the acquisition.
C-8
result.
However, some of these cost differences may simply reflect
differences in the level of services, and as noted, evidence to date
suggests that past mergers have not in general yielded significant
cost savings beyond that obtained through shrinking the size of the
banking organization.
