The Antitrust Aspects of Bank Mergers - Panel Discussion I: Development of Bank Merger Law
and the Fordham Journal of Corporate & Financial Law for sponsoring
Fordham Journal of Corporate & Financial Law
Dr. Anne Gron
z Copyright c 2008 by the authors. Fordham Journal of Corporate & Financial Law is produced by The Berkeley Electronic Press (bepress). http://ir.lawnet.fordham.edu/jcfl
THE ANTITRUST ASPECTS OF BANK
PANEL DISCUSSION I:
DEVELOPMENT OF BANK MERGER LAW
Fordham University School of Law
President, American Antitrust Institute
Dr. Anne Gron4
NERA Economic Consulting
† The panel discussion herein was held at Fordham University School of Law on
November 13, 2007, for the first annual Fordham Journal of Corporate and Financial
Law Symposium. It has been edited to remove minor cadences of speech that appear
awkward in writing and to provide sources and references to other explanatory material
in respect to certain statements made by the speakers.
1. Carl Felsenfeld is a Professor of Law and Director of the Financial Services
Institute at Fordham University School of Law.
2. Douglas F. Broder is a Partner in the New York office of K&L Gates and has a
primary area of expertise in antitrust litigation.
3. Albert A. Foer is the President of the American Antitrust Institute.
4. Dr. Anne Gron is a Senior Consultant in the Securities and Finance Practice of
NERA Economic Consulting.
PROF. FELSENFELD: I would like to welcome you here this
morning to Fordham Law School.5 My name is Carl Felsenfeld. I am
the Director of the Financial Services Institute, which is one of the
sponsors of today’s program, along with K&L Gates6 and the American
Antitrust Institute7 in Washington.
I would like to mention two major events. One is the United States
Supreme Court case United States v. Philadelphia National Bank in
1963,8 that held, among other things, that banks are subject to the
antitrust laws in their merger activities and that banking is essentially a
local business.9 The Philadelphia National Bank case was strict in their
We jump ahead thirty-one years, when the Riegle-Neal Banking
and Branching Efficiency Act of 1994 was enacted.10 That, for the first
time, enabled banks in the United States generally to go interstate.11
That led, of course, to the mammoth mergers that we have seen, which
brings antitrust laws to mind.
What I have seen since 1994 is that the number one bank in the
country will merge with the number five bank in the country and create
a multi-state institution, with billions of dollars in assets, and if it is
found to violate the antitrust laws, the solution is to knock off half a
dozen branches in the Peoria area or something like that, which makes
me wonder: Do we really have an effective law of antitrust for banks? I
hope the discussion today will consider that.
With that, I just want to thank the Fordham Journal of Corporate &
Financial Law12 for all their help. Doug, why don’t I turn it over to you.
It’s your program.
MR. BRODER: Thank you, Carl.
Good morning. My name is Doug Broder. I’m the co-head of the
Antitrust Group at K&L Gates, resident here in New York. Our panel
this morning is going to discuss the substantive competitive analysis of
bank mergers. We have two very distinguished and knowledgeable
panelists who are going to talk to us about this subject.
Bert Foer is the President of the American Antitrust Institute in
Washington, a nonprofit think tank focused on antitrust issues.13 Bert’s
career has included both private law practice, at Hogan & Hartson and
Jackson & Campbell, and government service, as the Assistant Director
and the Acting Deputy Director of the FTC’s Bureau of Competition.14
While at the FTC, Bert was a commissioner on the National
Commission on Electronic Fund Transfers.15 Bert also served in the
private sector as the CEO of a chain of jewelry stores for twelve years.16
Bert is a graduate of the University of Chicago Law School, has an A.B.
from Brandeis and an M.A. in political science from Washington
University.17 Bert is going to focus on the legal analysis of mergers.
Anne Gron is a senior consultant in the Securities and Finance
Practice of NERA Economic Consulting.18 Anne has conducted
research on, among other things, the effect of regulation in banking and
insurance, and the effect of bank portfolio composition on lending.19
Before joining NERA, Anne was an assistant professor at the Kellogg
School of Management, after beginning her career as an assistant
professor at the University of Chicago’s Graduate School of Business.20
Anne received her Ph.D. in economics from MIT, after graduating from
Williams College with a degree in economics and computer science.21
Anne, logically enough, will focus on the economic side of bank merger
Before I turn things over to Bert and Anne, I would like to briefly
outline the statutory and regulatory context within which bank mergers
are analyzed. Generally, the basic law that governs merger analysis in
the antitrust laws is the Clayton Act, section 7, which outlaws mergers
and acquisitions “the effect of which may be to substantially lessen
competition or create a monopoly.”22 Enforcement of section 7 is
invested primarily in the Department of Justice (“DOJ”), and
specifically in the Antitrust Division of the Department of Justice, and in
the Federal Trade Commission (“FTC”) and its Competition Bureau.23
The way mergers get analyzed now, since 1976, under the
HartScott-Rodino Act, is under a regime of pre-merger notification.24
Anytime two or more parties are planning an acquisition or a transaction
of a certain size—the current threshold is about $60 million—they are
required to notify the DOJ and the FTC of their planned transaction, and
then wait thirty days while the agencies take a preliminary look at the
proposed transaction to see whether there may be a competitive antitrust
problem under section 7.25 If the agencies decide there is not, the parties
can go ahead and close at the end of that thirty-day waiting period.26
Otherwise, the agencies may seek additional information and extend the
waiting period until thirty days after obtaining that information from the
The agencies do not have self-executing power to block a proposed
transaction.28 They have to convince a federal court that the proposed
transaction would violate section 7.29 If they can do that, they can get a
preliminary, and ultimately a permanent, injunction against the
transaction.30 More often than not, when there is an investigation and
the agencies have a problem with a proposed transaction, they can come
up with a settlement with the parties that allows the transaction to go
22. 15 U.S.C. § 18 (1996).
24. Id. § 18a(a).
25. See id. § 18a(a)-(b). On February 28, 2008, the thresholds for all
Hart-ScottRodino transactions increased, the threshold referenced in the text increased to $63.1
million. See FEDERAL TRADE COMMISSION, REVISED JURISDICTIONAL THRESHOLDS FOR
SECTION 7A OF THE CLAYTON ACT 1 (2008).
26. 15 U.S.C. § 18a(b).
28. Id. § 18f.
forward, subject to some kind of remedy in the form of divestiture or,
perhaps, a conduct remedy.31
Bank mergers are subject to similar, but not identical, procedures.
Bank mergers are analyzed under section 7, but with a difference. The
relevant statutes are the Bank Merger Acts of 1960 and 1966 and the
Bank Holding Company Act.32 For the most part, bank mergers do not
require notification under the Hart-Scott-Rodino Act, but that’s because
they require approval from one of several regulatory agencies—the
Comptroller of the Currency, the Board of Governors of the Federal
Reserve, the FDIC, or the Office of Thrift Supervision—depending on
what category the merging banks fall into.33 The banks file notice for
approval with those agencies.34 The agencies notify the DOJ, which
weighs in with its thoughts on the competitive analysis of the proposed
On the substantive side, though, there is a key distinction applied to
the analysis of bank mergers under the Bank Merger Act. The Act
requires the regulator who is making the decision to take into account
the Clayton section 7 standard that says the transaction should be
disapproved if it would violate section 7, unless—and here is the
important caveat—”the anticompetitive effects of the proposed
transaction are clearly outweighed in the public interest by the probable
effect of the transaction in meeting the convenience and needs of the
community to be served.”36
With that, I will turn it over to Bert.
MR. FOER: Thank you, Doug. Thank you all for coming out
How many know what the Herfindahl37 is? I will be focusing
primarily on the market definition and concentration issues that are
It was about thirty years ago, in October of 1977, that the National
Commission on Electronic Fund Transfers issued its report.38 I was the
FTC’s representation, one of twenty-six commissioners. We were
dealing then with the emerging technology of automated teller machines
and debit cards.39 It was a very forward-looking report. It set forth a
program intended to maximize both consumer protection and
competition for this unfolding industry.40
Congress quickly, and maybe surprisingly, enacted most of what we
had proposed.41 In retrospect, the EFT Commission served the country
unusually well for a statutory study commission. We saw then that there
were thousands of banks in the country, mostly small, but that local
banking tended to be rather concentrated. We believed, as a
commission, that consumers would benefit from fewer and larger banks
more directly in competition with one another.42
How do banks compete? Let’s start there.
Bank pricing includes a combination of interest rates and fees.
There is non-price competition also, and that focuses on location,
customer service, alternative delivery channels, the set of products being
offered, brand recognition, and relationship competition.43 To make a
couple of quick generalizations: (1) Large banks often face each other in
multiple markets,44 which might temper their aggressiveness; (2) The
presence or threat of a large bank may inhibit the pricing of a local
dominant bank;45 (3) Small banks are less profitable than large ones, but
beyond something like $50 billion in assets, profitability declines,
suggesting that the largest mergers may not gain much in the way of
economies of scale.46
Because of the risks to individuals and to the economy, banking has
been among the most regulated of American industries.47 However, the
amount of regulation has decreased since the 1970s. After the EFT
report—but surely not because of it—many restrictions on banking, such
as branching and interstate banking, were lifted.
48 By 1995
, all states
allowed interstate banking, and by 1994, only two states prohibited
Extensive consolidation was occurring in the 1980s and 1990s.
There were some very large deals, like NationsBank joining with Bank
of America.50 These were common. Today it is estimated that 630
financial holding companies control over 970 banks and 84% of all
banking assets in the U.S.51 Most of the rest operate as members of one
of almost 5000 bank holding companies that account for 13% of the
The large increase in concentration at the national level—aggregate
concentration—has not been matched by increased concentration at the
local level. In fact, it appears that at the local level concentration, on
average, has not changed very much. This is both because the newer,
larger banks were able to enter local markets through market-extension
mergers and because entry conditions are such that many new banks
were able to get started, even as old banks were failing or merging out of
existence. In other words, neither exit nor entry barriers seemed to be
particularly high. It has generally been the case that retail customers and
small businesses utilize local commercial banks.53 Corporate banking is
regional, national, and sometimes international.
What are the effects of concentration in banking?
Empirical analysis tends to indicate that bank efficiency is
negatively correlated to concentration.54 Banks operating in more
concentrated markets earn higher profits than banks operating in less
concentrated markets.55 One can speculate that consolidation, in
combination with technological advancement, has resulted in major
efficiencies. We can compare the automated teller machine with live
tellers. This combination of technology and consolidation has probably
provided consumers with more choices than they had thirty years ago.
Nonetheless, concentration in specific markets seems to have
negative consequences. This implies, first, that bank mergers will
continue to be a concern for those who weigh the effects on competition.
A second implication is that the starting point in any given competition
analysis has to be finding what the relevant market is, since one cannot
be concerned with concentration until there is a market in which
participants can be counted and their shares can be evaluated.
Aggregate concentration is interesting politically, but in specific markets
that is where the effects of competition have to be analyzed.
Bank regulating agencies have the jurisdiction to approve or deny a
bank merger.56 By statute, probable competitive effects must be taken
into account.57 If the merger is approved, the DOJ has thirty days to
block it on antitrust grounds.58 Generally, there is staff cooperation
between the DOJ and the regulatory agency during a simultaneous
review.59 The DOJ applies the basic antitrust law and analysis used in
other contexts, other industries, although there are a few banking
The key statute is the Clayton Act, specifically section 7, which
makes illegal those mergers whose effect may be substantially to lessen
competition, in any line of commerce, in any section of the country.61
The key Supreme Court case is United States v. Philadelphia National
Bank.62 This 1963 landmark case involved a merger between the
second- and third-largest of forty-two commercial banks in the
Philadelphia metropolitan area.63 The lower court had held that section
7 was not applicable to banks and that, in any event, the merger would
not lessen competition.64 In its decision reversing this, the Supreme
Court emphasized that there was a trend toward concentration in
banking in Philadelphia.65 The Court held—and the author was Justice
Brennan66—that bank mergers are covered by section 7 of the Clayton
Act.67 Second, it held that the Bank Merger Act of 1960, which directed
the banking agencies to consider competitive factors before approving
mergers, did not immunize approved mergers from challenge under the
federal antitrust laws.68 In other words, the DOJ could block a merger,
even though the Comptroller approved it.
Next, the Court turned to the lawfulness of the proposed merger
under section 7.69 The first issue was to determine the line of
commerce.70 That is a statutory line; we call it the relevant market,
relevant product, or relevant geographic market.71 The section of the
country is the relevant geographic market.72 These two aspects of a
market definition interact.
The Court agreed with the lower court’s finding that a cluster of
products and services denoted by the term “commercial banking”
constituted a distinct line of commerce.73 It disagreed, however, on the
appropriate section of the country.74 In banking, the Court said,
individuals and corporations typically confer the bulk of their patronage
on banks in their local community.75 The four-county area in which the
banks’ offices are located would seem to be the relevant geographic
market in which the companies operate and the customers can
practically turn for service.76
Finally, the Court turned to the prediction of a merger’s effect on
competitive conditions.77 This is a prediction about the future; it has to
be. The purpose of section 7 is to arrest anticompetitive tendencies in
their incipiency.78 In a key interpretation, Justice Brennan stated,
Specifically, we think 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 merger is not likely to have such anticompetitive effects.79
This structure-based test, the Court said, “is fully consonant with
economic theory”; namely, that competition is more likely “to be
greatest when there are many sellers, none of which has any significant
market share.”80 The Court went on to convert these observations into
some presumptions and presented a burden-shifting approach, a model,
for how merger analysis should proceed.81
First, if the merger results in a bank controlling at least 30% of the
market, there is a presumption of danger.82 Further, if there is an
increase of more than 33% in concentration, this is significant.83 These
structural indicia raise an inference of illegality, which can then be
rebutted by the merging parties.84
In the case at hand, the Court dismissed the various rebuttal
arguments that were put forward by the banks.85 Philadelphia National
Bank’s emphasis on incipiency meant that merger analysis would focus
on predicted future effects.86 Its emphasis on market structure meant
that measurements of concentration, which are highly dependent on
market definition, would play the essential role.87
The focus on future effects was heightened by passage of the
HartScott-Rodino Act in 1976.88 The Hart-Scott-Rodino Act has had a
largely unanticipated effect of moving the merger law development and
enforcement policy process from a regime of post hoc adjudication to
pre hoc administration/negotiation.89 Today, relatively few mergers are
actually litigated in court.90 A new body of administrative law has
consequently evolved outside of the judiciary’s sight. To a large extent,
the antitrust treatment of mergers is driven by a step-by-step application
of the principles in the horizontal merger guidelines.91
Philadelphia National Bank recognized that merging parties could,
in theory, rebut the presumptions created by concentration, but it did not
provide much hope that rebuttals would be found persuasive.92 This is
what has changed since Philadelphia National Bank. In one Supreme
Court and several appellate cases, ease of entry and efficiencies have
been allowed to rebut the structural presumption of harm.93 To an extent
that probably will not be clear until the next Supreme Court rendezvous
with a merger, it appears that the structural presumption has definitely
been weakened, but nevertheless stands as controlling doctrine.
It has been said that the modern era of merger policy began with the
1982 Merger Guidelines.94 The current guidelines were issued jointly
by the DOJ and the FTC in 1992 and revised a little bit in 1997.95 They
are limited to horizontal mergers—not vertical mergers, not
conglomerate mergers.96 Reflecting the influence of the Chicago
School,97 they focus on market power as the source of concern, and they
don’t talk about trends of concentration, as Justice Brennan did, or
protection of small business, which was one of the objectives of some of
the earlier merger decisions.98
presumption raised by concentration, one of which included new entry).
94. See Jonathan B. Baker, Mavericks, Mergers, and Exclusion: Proving
Coordinated Competitive Effects Under the Antitrust Laws, 77 N.Y.U. L. REV. 135, 148
(stating that the same problem that William Baxter confronted in drafting the
1982 Merger Guidelines—”harmonizing the older structural-era decisions with
contemporary economic thinking—remains central to horizontal merger analysis
95. See Robert M. Vernail, One Step Forward, One Step Back: How the Pass-on
Requirement for Efficiencies Benefits in FTC v. Staples Undermines the Revisions to the
Horizontal Merger Guidelines Efficiencies Section, 7 GEO. MASON L. REV. 133, 134
(1998) (stating that “[i]n April 1997, the FTC and DOJ issued revisions to the
Horizontal Merger Guidelines’ Efficiencies Section”).
96. See Joseph P. Bauer, Symposium, The 1982 Merger Guidelines: Government
Enforcement Policy of Section 7 of the Clayton Act: Carte Blanche for Conglomerate
Mergers?, 71 CAL. L. REV. 348, 350 (1983)
(stating that in 1982 the “Department of
Justice promulgated new Merger Guidelines, indicating that it will not challenge
nonhorizontal mergers, unless the transactions are likely to have an adverse impact on
actual or potential competition”)
97. The “Chicago School” refers to “the approach of the members of the
Department of Economics at the University of Chicago over the past century . . .
associated with a particular brand of economics which adheres strictly to Neoclassical
price theory in its economic analysis, “free market” libertarianism in much of its policy
work and a methodology which is relatively averse to too much mathematical
formalism and willing to forego careful general equilibrium reasoning in favor of more
results-oriented partial equilibrium analysis.” History of Economic Thought, The
Chicago School, http://cepa.newschool.edu/het/schools/chicago.htm (last visited May 1,
98. See Baker, supra note 94, at 147 (stating that a Chicago-oriented legal scholar,
William Baxter, appointed by President Reagan to head the Antitrust Division of the
Department of Justice, issued the 1982 Merger Guidelines where “he confronted the
problem of harmonizing the existing horizontal merger precedent with the economic
The guidelines proceed in five steps to summarize its defining
markets and examine market competition: adverse competitive effects,
entry, efficiencies, and the possibility that one of the firms is failing.99
Merger law is concerned with predicting the competitive effects of a
merger. It is presumed that when there is market power, there will be
negative competitive effects.100 Going back one step further, it is
presumed that when there is a high market share, there will be market
power.101 A crucial link in the inquiry is whether the merging parties
will have a high market share.102
This is not always an easy question to answer. In fact, it’s usually
the nub of a merger antitrust analysis. For instance, there is a pending
merger between XM and Sirius.103 It will go from a duopoly to a
monopoly if the market is defined as satellite radio. The parties argue
that is not the relevant market, because there is also AM and FM, and
there are other types of communications with which they say they
compete. That is going to be the issue that the Justice Department has to
determine: What is the relevant market?
Whole Foods wants to acquire Wild Oats.104 These are the two
primary competitors in the organic/natural foods market. The merger
would create a near-monopoly in that market. But is that the relevant
market? They were able to convince a district court that the market is
actually much broader. It includes Safeway and Kroger’s and all the
approach of the Chicago School”).
99. Mr. Foer failed to name the fifth step, which is “significantly increase
concentration and result in a concentrated market.” See 1992 Horizontal Merger
Guidelines Notice, 57 Fed. Reg. 41552-01 (Apr. 2, 1992) [hereinafter Notice], available
at http://www.ftc.gov/bc/docs/horizmer.htm (last visited Feb. 1
100. See FTC v. Procter & Gamble Co., 386 U.S. 568, 577 (1967) (holding that
“[s]ection 7 of the Clayton Act was intended to arrest the anticompetitive effects of
market power in their incipiency”).
101. See Notice, supra note 99 (stating that when the post-merger
HerfindahlHirschman Index exceeds 1800 points, “it will be presumed that mergers producing an
increase in the HHI of more than 100 points are likely to create or enhance market
power . . . .”).
102. See id. (stating that a federal agency will first assess “whether the merger
would significantly increase concentration and result in a concentrated market”).
103. See Frank McGuire, Year-End Review of Markets & Finance 2007, WALL ST.
J., Jan. 2, 2008, at R12. The DOJ approved the merger on March 24, 2008, leaving only
one more regulatory hurdle—approval from the FCC. Sarah McBride & Amy Schatz ,
XM, Sirius Move Closer To Improbable Merger, WALL ST. J., Mar. 2
104. David Kesmodel, Whole Foods Sets Probe As CEO Apologizes, WALL ST. J.,
July 18, 2007, at A3.
other places where you can buy organic foods, even though they don’t
specialize in them.
We’ll see. That’s on appeal by the FTC.
The question is which close substitutes are close enough so that
they should be considered to be in the same market? Under the
guidelines, we begin the market-definition process by accounting for the
possibility that buyers would defeat an attempt by sellers to profitably
exercise market power because they will purchase close substitutes from
The technique used for this is kind of hypothetical. It is the process
of the hypothetical monopolist. You are trying to define a market as a
collection of products or services that would form a valuable monopoly
within a defined geographic region. Were the hypothetical monopolist
in this market to raise the price by a small but significant and
nontransitory amount, which is usually considered to be 5% for the
foreseeable future105—would enough potential substitutes flow into the
market so that the price would come back down to where it was before
the monopolist raised the price? If the answer is yes, enough would
flow in, then you have defined the market too narrowly. We have to go
back and try a broader definition and do the same test, until we come to
a market that is defined in such a way that a monopolist in that market
could maintain its price increase. Then we have the right definition.
That’s the way this process works.
In Philadelphia National Bank, the market was a cluster of services,
such as loans and other types of credit, deposit accounts, checking
services, trust administration.106 The Court said, “Yes, that cluster is
distinct, and it represents cost advantages and consumer preference. We
think that that is a viable definition of a market.”107 They were not
going through the guidelines test at that time, because it didn’t exist.108
Once you have the market defined satisfactorily, the next step is to
figure out who the participants are in the market. The guidelines count
as firms participating in the market all firms that currently produce or
sell in the market.109 This is not always obvious. For example, in a
commercial bank merger, what do you do about thrifts? Are they in the
market, not in the market, or partially in the market? It depends. In a
retail product market, the thrift and the credit union deposits are given
full weight, because they are alternatives for the customer. But in the
small business lending market, less credit, or maybe no credit, would be
given to thrifts and credit unions.
For the most part, DOJ investigation of a commercial bank merger
is generally similar to its other merger investigations.110 One difference
is that they have an unusually large amount of information because of
regulatory oversight of banks.111 The DOJ staff usually requests
information that it needs through a simple request for information, rather
than a civil investigative demand.112
Unlike other mergers, a bank merger receives antitrust immunity
after the post-approval waiting period expires.113 In other words, if the
DOJ and the regulatory agency both approve it, nobody else can go after
it.114 If the DOJ does file suit, there is an automatic stay, and a federal
district court will conduct a de novo review of the transaction.115
The geographic market determination116 in this area tends to be
more of an issue than the product market determination.117 The starting
point for the DOJ is the Federal Reserve Board-defined banking
market.118 The geographic market definition will probably depend on
the product market at issue. In retail banking, it is generally going to be
defined by the customers—where they live, where they work. For small
businesses, customers generally choose a bank that is within a few miles
of their business location. The geographic market for middle-market
customers is going to be larger than for small businesses; for very large
banks, it could be international. Critically, though, the market definition
is not set in concrete. A new method of doing business might trigger a
change in market definition.
Let’s take Wal-Mart, for instance. Traditionally, Wal-Mart was not
considered to be a supermarket, or in the same market with
supermarkets. So when two supermarkets wanted to merge, Wal-Mart,
operating through a different channel of distribution and with a whole
different style, was excluded. That made it somewhat difficult for large
supermarkets to merge, because, at least in some geographic areas, they
would have a high share of the market. If you include Wal-Mart as a
supermarket, or within that food retailing market, and you say that is the
relevant market and Wal-Mart is bigger than any of the others, you have
a whole different picture.
There is some indication that this is going to happen. It did happen
in a settlement the FTC reached in Puerto Rico when Wal-Mart entered
the market.119 If you shift your paradigm of market definition and you
include Wal-Mart in the market, then the ability of large supermarkets to
merge becomes much greater. The point is that we can have all kinds of
changes that might lead to a different definition of a market. The
changes could involve transportation costs, consumer buying habits,
technological advancements, governmentally-established barriers to
entry, and patterns of trade. All of these are possibilities. There is a
major monopolization case going on right now in which AMD is suing
Intel.120 The complaint describes the relevant geographic market as
nothing less than global.121 As the world changes, market definitions
can expand or, in some cases, shrink.
If the accepted geographic market for commercial banks were to be
enlarged by merger analysis from local to regional, national, or global,
the levels of concentration would in many situations fall below the
thresholds that give today’s antitrust enforcers concern. Similarly, if the
line between commercial banks and non-banking financial institutions is
weakened, the product market definition could change. Banks that are
very large ducks in today’s banking pond could appear much smaller to
the expanded financial lake. In either case, the adoption of a new and
enlarged market definition would open the way to new and potentially
much larger institutions. Needless to say, the implications would be
both enormous and enormously controversial.
There is ample justification for careful and ongoing empirical
research to test the accuracy of both prevalent and alternative
assumptions. Consolidation that has created national banking and
ongoing changes in banking services do assure us that antitrust analysts
will have to continue to visit and revisit the assumptions relating to
market definition, to participants in the market, and to the role of
concentration in commercial markets.
Given the importance of empirical research and what economics
has to tell us about these matters, it’s appropriate to turn the table over to
DR. GRON: Thank you. I am going to be speaking more about the
economic side of the market. This discussion is really going to focus on
retail banking. In the current implementation of the 1963 Supreme
Court ruling, and subsequent judgments upon the matter, bank mergers
at the retail banking services level are primarily analyzed in terms of
their geography and product line.122 The initial analysis is typically that
the geographic market is, in fact, local and that the product line is this
cluster of retail branch banking services that include demand deposits
Implementation, as a first pass, is a pretty standardized process for
bank mergers. However, there are slight variations in implementation at
a screening level—when they initially propose a merger and how
regulators may look at them and evaluate them.
What we know from economic research, in terms of what we can
discern about the size of relevant markets, depends, in some ways, on
the product. The particular product may suggest a narrower or broader
market. The nice thing about the research is that it’s not just an “it
122. See United States v. Phila. Nat’l Bank, 374 U.S. 321 (1963) (defining the
geographic markets where the effect of the merger on competition will be direct and
immediate); see also United States v. Conn. Nat’l Bank, 418 U.S. 656, 666 (1974)
(stating that geographic markets are described as the area in which the acquired bank
was in direct competition with other banks).
depends” answer; it’s actually an answer that provides some insight into
the types of products and the size of the markets.
What is a typical bank merger process? In the beginning of the
process the geographic market, if urban, is typically going to be defined
as the metropolitan area. For rural areas, it’s typically going to be one
county, maybe two, depending on where the population and the bank
branches in those counties tend to be. Initially, the product lines
probably receive the most focus are deposits or loans.123
Merger analysis is typically based on the Department of Justice
Merger Guidelines124 and by the Hirschman-Herfindahl Index,125 which
is really just a measure of market concentration. The typical way to
measure concentration in a market is a concentration ratio.126 For
example, a concentration ratio for the top four banks would tell you
what percent of market share the top four banks have. But if you have a
concentration ratio of 40, you don’t know, for the top four banks,
whether each one has 10% or whether one bank has 35% and the next
three have the remaining 5%.
As an economist judging the performance of this market, you care
about the difference. The Herfindahl Index provides a better index in
those terms, because it is the square of the market shares. As such, a
larger market share, when it gets squared, gets proportionately larger.
That is why we use a Herfindahl Index.
A 200-point increase in the Herfindahl caused by the merger and a
post-merger Herfindahl of 1800 or more is often seen as reason to look
further.127 You look at whether the ease of entry and exit is going to
have an impact on the competition among the banks that result from the
merger. You also look at the types of business in which those particular
banks are engaged. A separate issue is whether the concentration of
deposits that results from this merger gives the merging entity over 35%
of the deposits in the market, which would also result in further scrutiny.
It doesn’t mean that is the end of the process; it just means that there
will be more of the process.
One difference between the different regulatory bodies, to some
extent, is the initial firms that each might include in the market. The
Office of the Comptroller of the Currency and the FDIC typically
include thrifts at full measure.128 If they have $100 in assets, the $100 is
included in the market and included in their shares. The Federal
Reserve Board, in contrast, often includes a thrift deposit at 50%
weight.129 If they have $100, it’s going to be in at $50. The Department
of Justice might include the thrifts in the deposit.130 It depends on the
basic size of the thrift competition in that market.
The research is best grouped into three areas. The first type of
research basically looks at markets defined by “use.” It looks at surveys
of households and small businesses, and asks, “Where do you get your
financial transactions?” If you get your financial transactions from the
set of banks within the local geographic area, as just defined—based on
either a metropolitan area or a county—then the local measure is a pretty
good measure of the market structure. If, however, you are getting a fair
amount of your financial services from outside that market, then clearly
you don’t have all the firms within the market. The size of the market is
correctly defined when you have a market such that a monopolist
operating in the market could raise and hold prices higher by 5%,
because if you have a lot of customers doing business with firms outside
http://www.usdoj.gov/atr/public/testimony/hhi.htm (last visited Feb. 4, 2008) (stating
that “[t]ransactions that increase the HHI by more than 100 points in concentrated
markets presumptively raise antitrust concerns . . . .”).
128. See KEITH R. FISHER, MERGERS AND ACQUISITIONS OF BANKS AND SAVINGS
INSTITUTIONS § 3.12, at 3:91-92 (1993).
129. Robert B. Avery and Katherine A. Samolyk, Bank Consolidation and Small
Business Lending: The Role of Community Banks 19 n.16
(FDIC Working Paper No.
, available at http://www.fdic.gov/bank/analytical/working/wp2003_05/
130. See generally Robert E. Litan, Deputy Assistant Attorney Gen., Dep’t of
Justice, Address before Antitrust Section of ABA on the Antitrust Assessment of Bank
Mergers (Apr. 6, 1994), available at http://www.usdoj.gov/atr/public/speeches/litan.txt.
of your local market structure, you clearly don’t have all the competitive
The second set of research looks at the determinants of prices—that
is, interest rates, not the fees per se. It looks at price variation over a
larger geographic area, and asks, “Are prices seemingly most influenced
by local market conditions—the structure, the profitability of businesses,
and so forth—or are they also influenced by larger economic conditions
over a larger geographic area, like statewide or, in some cases,
There are a couple of studies that are closely related to this
question. The market structure analysis really asks how well different
types of financial institutions compete with each other or, more
importantly from an antitrust standpoint, how important is the
competitive effect of one type of financial institution on another. A
Herfindahl Index or a concentration ratio weighs all firms similarly. If
you decided to weigh the thrifts at 50%, everybody is either full-weight
or 50%. In reality, firms are differentiated. This market structure
analysis tries to look at in what way financial institutions are
differentiated and learn something about their competitive influences on
Finally, some relatively new research looks at the prices of bank
branch sales and asks whether these are influenced by local market
conditions or whether they are influenced more by state market
conditions. The survey data addresses whether households or small
businesses do most of their financial transactions with geographically
local banks. The answer has been that it depends on the type of product.
For consumers, checking and NOW accounts131 remain primarily local.
Most consumers and households get their transaction accounts at banks
within a small local geographic market. That said, the percentage of
households getting all their financial services from banks has fallen
significantly over time.132 The research I saw tended to stop as of the
late 1990s, but I think we can strongly expect the trends to continue. For
example, by 1998, less than 45% of the households in the surveys
borrowed locally.133 That is likely to continue.
This isn’t that surprising. Borrowing is something that, because of
the significant cost to consumers, they are going to search for pretty far.
Meanwhile, transaction accounts are something that you might want
locally, because you are coming in and out of the bank and have a lot of
interactions with them.
For small businesses, the research gives a little more of a mixed
message. Research from the mid-1990s suggested that distances
between borrowers and lenders may have increased and, in fact, are
currently likely to be larger than a local geographic market.134
Specifically, the study looked at a survey of small business lending, and
it found that distances between borrowers and their lenders in the 1970s
were relatively close to a local market.135 They were about seventy
miles apart, on average, whereas by the early 1990s, research indicated
that they were closer to 160 miles away, which is larger than most
geographically local markets.136 The challenge, however, is that we
don’t have a lot of data from the 1970s. We only observed agreements
made in the 1970s in the 1990s, so you have some bias.
Like good academics, people ran off and asked the question again.
The next study basically came out somewhat contradictory. While the
distance between small business borrowers and their lenders varied
according to the type of product, small business borrowers much like
households get their transaction accounts locally, too. Regarding lines
of credit, another measure of lending investigated, small businesses
typically transact with financial institutions within about thirty miles or
so, and have not increased in distance throughout the 1990s.137 This
study concluded that, on average, small businesses get a lot of their
133. See id.
134. See FED. RESERVE BD., 1993 NATIONAL SURVEY OF SMALL BUSINESS FINANCES
(1993), available at http://www.federalreserve.gov/Pubs/Oss/oss3/bibliography/index.h
tml (providing various reports and memoranda summarizing the 1993 survey results and
135. See id.
136. See id.
137. See FED. RESERVE BD., 1998 NATIONAL SURVEY OF SMALL BUSINESS FINANCES
(1998), available at http://www.federalreserve.gov/Pubs/Oss/Oss3/ssbf98/ssbf98hom
e.html (providing various reports and memoranda summarizing the 1998 survey results
and their impact).
financial services from local markets, and that this has not changed over
A second set of research asked, “Do we have all the banks in the
market?” Using some information from the Community Reinvestment
Act,139 researchers looked at how many banks were providing credit for
small businesses within a local geographic area. They found that, while
a lot of the relationships were with local banks, a number of banks were
excluded. In particular, they tended to be large, multimarket bank types.
The excluded lending was also typically credit card-related.
If you include these banks, the measure of concentration falls
significantly in the market area. It tends to be that the share of outside
banks in this sort of measure is greater in more concentrated markets. If
you are looking in a market where you might be particularly concerned
that concentration would affect pricing and profitability within the
market, interestingly enough, those are the markets which seem to have
more use of outside credit by small businesses.
There is also a branch of literature that looks at price determination
within a market. This question basically asks, “Are interest rates
reflective of concentration?” The structure-performance paradigm140
really underlies much of the antitrust analysis. If you are looking at an
industry that is very concentrated, the likelihood that competition is very
intense is less than when you have lots of small firms all acting
atomistically in the sort of standard, perfectly competitive world. Under
that scenario, a competitive market would likely have low prices and
low profitability and be very efficient. A very concentrated market is
likely to lead to conduct that is less competitive among firms and likely
to lead to higher profitability. This research looked for both the
connection between prices and profitability, and also the difference
between prices and concentration.
Early research found a very a surprising result. Deposit interest
rates—remember, the higher the deposit interest rate, the better it is for
the consumer—were positively related to concentration measures, which
would suggest that maybe concentrated markets are better, from a
consumer standpoint.141 This turns out to be an artifact of that particular
study, and later researchers have been unable to replicate it. In fact, they
have overturned the results.142
Typically, you look at interest rates on local and state or national
measures and ask, “Is a variation in prices most related to local
concentration and local structure, or is it also influenced, or more
influenced, by state level?” Again, you get some variation on the local.
State measures actually explain a fair amount of the variation. Similar
to the survey data, prices and interest rates on local transaction accounts
appear to be heavily locally determined. However, if you look at
something like six-month certificates of deposit that are pretty liquid and
fairly easy to price, their rates seem much more nationally determined.
The evidence suggests that most deposit rates are, not positively,
but negatively, correlated with concentration, meaning you get lower
interest rates on your deposits if you are in a more concentrated
market.143 But prices are going to be influenced by state and
superregional effects, not just by the local market.
This research went further and actually looked at different types of
banks. This is particularly important when you think about
concentration measures as treating all banks and all products within
them uniformly. This research has separated two kinds of commercial
banks—banks that participate across multiple markets, which we
generally call multimarket banks, and single-market banks. The bulk of
the assets in the U.S. banking system are in multimarket banks. When
you separate them and look at them individually, it turns out that
multimarket banks don’t set prices at individual market levels.144 They
set policies for interest rates and so forth that go across local markets.
They do not vary that much with local market conditions.
Do the multimarket bank interest rates charged or given on different
kinds of financial instruments just represent a weighted average of the
places these banks are in? The answer is actually no. They more
closely represent the statewide conditions than simply the weighted
average of the local conditions these banks are in. In contrast,
singlemarket banks actually do seem to price to the locality.145 They reflect
J.L. & PUB. POL’Y 277, 286-87 (1986).
142. See id. at 287.
143. See id. at 286-87.
144. See Andrew M. Cohen & Michael J. Mazzeo, Market Structure and
Competition Among Retail Depository Institutions, REV. OF ECON. AND STATISTICS,
Feb. 2007, at 60, 61 n.3.
145. See id.
the local market conditions, not as much the state conditions. In fact,
deposit rates are lower when market concentration is higher.
Interestingly enough, the two types of banks do compete against
each other and have an effect on each other. The relationship between
the deposit rates and local market concentration is weaker when
multimarket banks are very strong in the market, suggesting that they are
going to have a significant disciplinary effect on each other.
One important caveat: this structure-performance paradigm is really
about concentration leading to, eventually, higher profits because of less
intense competition. It is difficult to fully know whether the differences
in prices reflect differences in quality or whether they reflect differences
in both quality and costs that are systematically different among the
different types of banks.
That’s where these other studies are attractive. These other studies
try to work in an environment where they don’t look at prices per se.
They have underlying a model of profitability and they then observe
something else in the world. Based on their underlying model of
profitability, they infer something either about competition or about the
value of a certain thing.
The first type of study I call market-structure studies.146 These
studies use the composition of a market to analyze the competition
within and across financial institutions.147 This is supposed to answer,
“How competitive is the market?,” beyond just, “What’s the
concentration ratio? What’s the Herfindahl Index?” They want to look
at the players, look at how they interact. How important is it that you
have one thrift, two single-market banks, one multimarket bank, versus
two multimarkets and two single-markets? What kind of a difference
might it make?
To do this, you basically need geographically isolated markets.
They have typically been done in rural markets, which is a sacrifice,
because you are not going to be able to transfer these results as naturally
to an urban market. But with that sacrifice comes a benefit, in that you
can then look at the number of firms in the market and the types of firms
in the market. You look at the variation across the markets and the
market size, and if these firms are in the market and no more, then we
know that these guys are profitable. Putting in one more of any of those
types, the marginal guy is going to be unprofitable. What can we infer
about the effect of competition of one type versus another?
This research has identified multimarket banks, single-market
banks, and thrifts. Thrifts are probably the most differentiated—not a
surprise, given that the history of thrifts has been somewhat different
than the history of commercial banks, in terms of the product markets
and the things in which they were involved. They have less of a
competitive effect on single- and multimarket banks, except in the rural
markets that are closer to urban markets. The study defined less rural
and more rural markets, depending on whether you were closer to a
larger city than not.
Multimarket and single-market banks are closer substitutes in the
more rural markets. They definitely have competitive effects on each
other. An additional single-market bank, however, does not have as big
an effect on an existing multimarket bank as an additional multimarket
bank. They are not perfect substitutes in terms of the profitability of
these kinds of institutions.
What does it say about merger analysis? It says that you have to
think about the composition of the market, not just the number of
players in the market. Again, it depends on the product lines.
The last type of study I want to mention looks at branch sales,148
instead of prices, which could be influenced by quality or cost
differences across firms. If somebody is going to buy this branch, they
are going to be willing to pay an amount that should reflect the
profitability that this asset should generate. The question is, “Does that
profitability depend on local market conditions—how concentrated that
local market is—or does that price reflect state or super-regional market
structure or market conditions?”
Any price is the outcome of a bargaining situation. Prices are going
to be higher if the seller has more bargaining power, if the potential
synergies are very large, and if the buyer is more interested in the
product and there are multiple bidders. When you have competition for
an asset, price goes up. When you have less competition bidding, price
goes down. With that in mind, however, the research suggests that
branch prices are correlated with state, not local, factors.149 The
148. See Martin R. Garay, III, New Developments in Savings and Loan Law, 39
BUS. LAW. 781, 791-91 (1984).
149. See TIMOTHY H. HANNAN & ROBIN A. PRAGER, FED. RESERVE BD.,
inference is that profitability is not a locally based event. It’s actually
something that depends on the market conditions at the state level.
Remember, when you are looking at a branch, you are looking at a
cluster of products, some of which may be more local, some of which
may be serving a broader market. You again come back to this issue
that when you are looking at a specific merger, you want to consider the
products that are most effective at that time.
So what do we learn? First, many markets for banks appear to be
broader than a narrow local geography—although NOW and checking
accounts are not—and they appear to be getting broader. Second, there
are significant variations in competitors and competition, both within
and across markets. There is room for people to discuss—jobs for
lawyers, jobs for economists. Third, the relevant market is likely to shift
as the composition of competitors within markets shifts. That is
particularly important because there has been a lot of consolidation in
the U.S. banking industry, and to the extent multimarket banks are going
to continue to increase their market presence, you are going to get a
change in how pricing occurs. To the extent multimarket banks price
over a larger area than single-market banks, there will continue to be
changes in the breadth of markets considered for competition.
Some trends exist in the marketplace that are likely to influence the
nature of the relevant market as we continue forward. Financial services
are no longer one-stop shops. They are no longer all together. With
securitization,150 you can see how many parts of the mortgage process
have been unbundled.
Banks, up to this point, have not been able to identify the potential
switchers from locked-in consumers. Locked-in consumers are, of
course, the ones for which you can potentially raise price and create a
niche market, where you will price-discriminate and charge them more
than the potential switchers. To the extent that banks or other financial
institutions will be able to identify the switchers from the locked-in
consumers in the future, you will get this kind of differentiation in your
MULTIMARKET BANK PRICING: AN EMPIRICAL INVESTIGATION OF DEPOSIT INTEREST
RATES 1 (2004), available at http://www.federalreserve.gov/pubs/feds/2004/200438/200
438pap.pdf (stating that a study found that for multimarket banks, state level
concentration is more strongly and consistently related to deposit interest rates than
local market concentration).
150. For a definition and discussion of securitization, see Claire A. Hill,
Securitization: A Low-Cost Sweetener for Lemons, 74 WASH. U. L.Q. 10
products—one aimed at the group that is locked in and one aimed at the
group that is more likely to switch.
There have also been huge changes in information availability,
which is particularly pertinent for financial institutions. Obviously, the
information provided by the Internet is going to tend to increase the
market size and to have a significant price effect. The six-month CD is
a great example. To the extent that it’s easy to get the information about
what interest rate is being offered in what location, you are going to put
your money where the highest interest can be earned rather than a local
market. Having that information at low cost is going to just increase the
extent to which that’s true. Internet access to banking came along a little
slower than some people might have thought, but it is still growing. It is
definitely going to have an effect on how you would define a local
market and the competition among players.
Almost all households with a bank account have at least one
electronic fund feature. Many are not things that we would worry about
as having significant effects on the market structure—like direct deposit.
However, in 2003, 32% of households reported using computer
banking,151 up from 4% in 1995.152 The early adopters were
characterized by higher incomes, more financial assets, younger heads
of household.153 They were less than fifty-five years old.154 They had
more formal education, on average.155
Now that we have expanded to about 32%, the surveys suggest that
the users of computer banking are actually much more diverse than one
might have predicted. A typical prediction would be that the people
using computer banking would stay younger and stay within the
higherincome group, because they would have easier access to computers.
Interestingly enough, it’s much more diverse in terms of age.156
Researchers commented that they were older than they thought they
would be. Education of household head has changed a lot, too.157 As
151. See Christoslav E. Anguelov et al., U.S. Consumers and Electronic Banking,
1995-2003, Fed. Res. Bull. (Winter 2004), available at http://www.federalreserve.gov/
152. See id.
153. See id.
154. See id.
155. See id.
156. See id.
157. See id.
we move forward, that is going to have an impact on how you want to
think about the relevant market.
Securitization is what I consider predominantly an effect.
Securitization reduces the cost of rapidly expanding loans and changes
the dynamics of the competition. It is also important in terms of the
market structure. Securitization has been around for a long time. There
has been tons of experience with government-sponsored entities like
Freddie Mac158 and Ginnie Mae159 and Fannie Mae160 for the
conforming mortgages. But that is for your typical mortgage. Of
interest is that there has really been an expansion of the use of
securitization into other forms of credit lending—for example,
subprime161 and jumbos.162 Auto loans are in there; student loans are in
there; credit card securitization is in there. All kinds of lending has been
subject to securitization.
What does that do to the local market structure? First of all, it has
contributed to an unbundling of the lending process. Brokers specialize
in originating loans. Lenders are no longer just banks; they include
nonbank firms, firms organized as REITs,163 firms organized as specialty
financial organizations. Specialized firms service the loans, and so on
and so forth. By having outside investors provide the liquidity, you are
no longer tied to the deposits in your marketplace in the same way—
hence, breaking that link between your concentration of deposits and
what you might be doing in terms of the lending side of the market.
What does that do for your market analysis? The presence of the
secondary loan markets and other developments in structured finance
means a lower cost of entry. Not only are entry and exit already pretty
easy in the banking industry, this makes it potentially much easier to
enter at a much larger scale, quickly. When there is an attractive profit
opportunity—which might come about if a merger has reduced
competition and prices have been raised and profitability is higher—it
should be easier for a bank or another financial institution to expand its
presence in a local market, in a lending market, if they have access to
securitization. Securitization is a force that is likely to reduce the cost of
entry, providing the liquidity for lending and affecting loan rates,
bringing prices back down, to the extent that concentration might have
From a consumer perspective, securitization helps broaden the
market. It helps bring other competitors into the market. It is
particularly relevant for antitrust analysis when a merger may reduce the
number of firms in particular markets. For example, a recent merger
was proposed between two of the largest student lenders in the
business.164 The government is also a big contributor of student
lending.165 There are many ways in which you might want to analyze
this. The fact that securitization is out there—it’s a profitable
business—means that even though they are likely to increase
concentration in the industry, there is going to be a significant offsetting
effect of likely entry.
MR. BRODER: Thank you, Anne. I think we have a few minutes,
if there are any questions for Anne or Bert.
QUESTION: Anne, as I understand it, deposits are really down in
banks. A deposit is really the other side of a loan. How does it go into
your analysis, that true deposits are way down from where they were
DR. GRON: That’s a very good point. When you talk about local
banking, NOW accounts and checking accounts are still being held
largely locally. Casual empiricism says people you know have a local
164. See Press Release, JPMorgan Chase & Co., Chase Student Loans to Purchase
Collegiate Funding Services, Creating a Comprehensive Education Finance (Student
Lending) Business (Dec. 15, 2005), available at
165. For more information, please visit the Federal Student Aid website,
http://studentaid.ed.gov (last visited Mar. 1, 2008).
bank account, but the money gets parked there for a little bit and then it
goes into a money market account somewhere or other places.
That’s going to feed into the analysis, to the extent that other
financial instruments are important for the local market—line of
business again. The local bank, by having transaction accounts available
for you, is allowing you that ready access. The fact that deposits are
down just means that overall market size has declined over time.
Profitability might be expected to decline. You have to ask where are
banks getting their other profits from? How is the business model
I don’t think you can ask the question in isolation. You have to
look at the banking model as a whole. To the extent that deposits alone
were a big source of profits for banks and now they have declined in
magnitude, clearly banks will be less profitable. Some of the mergers
might be caused by banks that are not doing as well as they used to.
That is, again, part of the whole merger analysis that you can easily
QUESTION: When looking at the projected interest rates that will
be paid on savings as a result of a merger, is there any weight to the
argument that it’s better for a bigger bank, which is usually more highly
concentrated, to pay higher interest rates than smaller banks that are
probably adversely affected by the merger? Larger banks are better
regulated, and there is more competition for deposits that causes smaller
banks to up the price on their deposits.
MR. FOER: I think there is a false assumption in this conversation.
The price that a bank offers for a service is some combination of interest
and a variety of fees that hit from all different directions, making it very
difficult to compare sometimes, but also making it somewhat of a false
reference, if we are going to make judgments just on interest.
It might be that large banks offer higher interest, but lower fees. I
don’t know. I don’t know how to answer the question, but I am worried
about the way we measure this. We may not have any choice, because it
gets too complicated.
DR. GRON: Some of the research looks at a combination of
average fees and interest rates. You can’t measure exactly what you
would like in all of these things. You have to take it with a grain of salt.
But your question is whether it’s better to have a large bank competing
for consumers’ deposits because they can better manage the risk?
DR. GRON: There is the question of efficiency, whether a larger
bank has lower costs, and whether that is offset in the consumers’
interest. If they are that much more efficient, they may be able to offer
consumers a better deal at the same time, by passing some of that on.
But it’s not going to be the only reason. You need a pretty strong
efficiency argument to say that you are going to do better than some of
the smaller banks.
QUESTION: One thing shown in the credit crunch that we are
facing today is that the bigger banks are not better “managers” in terms
of the interest rates they charge. In fact, some may have to be given
back if there is a bailout of the type that took place in the S&L crisis.166
You get 5% today, but if your taxes go up later on because they have to
pay $250 billion or $1 trillion, then that is a phony interest rate. We
have had lots of mergers, and now enormously powerful banking
institutions straddle the planet.167 When we talk about market power,
we talk about a conscious act by a company that may be a monopoly or
that can elbow people away. The New York Times, on Sunday, had an
article that if they are too big to fail, they are too big to manage.168
Citicorp is $2.5 trillion.169
I don’t know whether this is a question or a comment.
It seems that when you talk about market power and about mergers
that lead to gargantuan institutions, the antitrust laws ought to have some
room for the concept of dysfunctional power. The same effect can take
place. It’s not that they are consciously elbowing people away, but they
affect competition because of the damage they cause, for example, in the
MR. FOER: There is a lot of stuff there, and you are hitting some
very basic points about the way we deal with mergers and about the
course of antitrust in the last thirty years. The course of antitrust has
been determined by the Chicago School, where we focus on
microeconomics, and in particular on short-term price effects—not
completely, but largely, to the exclusion of other issues—which means
aggregate concentration, the total amount of wealth in a small number of
banks, is irrelevant. That’s a political issue.
This is the argument: If we wanted to deal with that, we would have
to deal with it politically, through regulation. But the predominant
philosophy, at least for thirty years, has been that we don’t want to go
that route. We want the markets to prevail, for a variety of reasons.
Laissez-faire is triumphant right now.
That leads to this concentration movement that we are seeing in
every industry. New industries are invented and they bring new kinds of
competition. The Internet is going to have, and has already had, an
impact on financial institutions and the way we borrow money. Things
are constantly changing. The one consistency is the increase in
concentration. Where we used to worry about industries with six or ten
major competitors, now we only worry about industries with three or
four. Two or one is a problem still. The threshold of concern in
antitrust enforcement today is more like a merger from four to three than
from eight to seven, or even six to five.
Ask the question, as a general matter: “Would you rather have
industries, including banking, where there are eight or ten major players,
as opposed to three?” It’s not such an easy question, actually. First of
all, every industry is different. Second, you may be giving up some
important efficiencies if you opt for the eight to ten. But there may be
some benefits of the type to which you were alluding, perhaps, that are
It has to do with power, political power. That’s one reason why
commercial banking was separated from investment banking, why we
had branching limitations and unit banking rules and so forth. These
were political decisions that were based not on how you achieve
efficiency, but how you keep the economic power from getting too
We moved away from that, for a variety of reasons. Maybe it’s
time to move back in that direction a little bit. I do think we are paying
a price for having institutions that are too large—unnecessarily large.
As I pointed out, above $50 billion, it does not appear that there are
DR. GRON: I would hesitate to want the government to be the
decider of whether a company can manage itself well or not. I think our
ability to assess that is imperfect at best. As a management challenge,
yes, big institutions are a challenge. But I don’t think that the
government or a set of standards is a better way to analyze that than the
market. Failures of organizations are perhaps the best discipline in
telling people not to undertake certain types of endeavors. Unless their
money is on the line, they are not going to make the same kind of
decision. I don’t see why the government would make a better decision
to start off with.
I also observe that booms and busts have existed before
concentration was high and they will exist whether concentration is high
or not. That is actually an aspect of the marketplace. I definitely agree
that the subprime mortgage situation170 is a mess. I think that is an
uncontroversial statement. I don’t know if government intervention
would help it particularly. It’s going to be very costly to get out of. I
don’t think that concentration alone is going to prove to be the cause of
the effect. There are multiple things happening at the same time. I think
you can point to things that government did that may have contributed to
it, not that the government is solely at fault.
There are many things that happened that have led to that. There
has been an overall expansion of the use of credit. There is something
much larger than just a concentration in a couple of investment banks
which, I agree, is high. But it’s beyond something that we have
discussed here. To the extent that investment banks are very dominant
in certain types of product lines, it’s really a different question when you
look at their types of mergers as well.
QUESTION: In retail banking, when you do your economic
analysis, how much are you looking at the possibility that these
structures which are based on market power and price might be
undermined by a situation in which the user cannot figure out what the
price is? As an example, Chase Manhattan Bank, which has branches
now every two blocks on the Upper West Side, put out a fifty-page
bulletin about their fee structure, with twenty different kinds of banking
retail packages, which was totally incomprehensible in terms of how
much we would actually be paying.171
170. See Painter, supra note 161.
171. For more information on the products, services and other information that
would have been contained in the bulletin referenced by the speaker, please visit the
Chase Online Banking website, http://www.chase.com (last visited Mar. 22, 2008).
DR. GRON: That’s a fantastic comment. You are absolutely right.
Economic models and economic analyses assume that consumers
actually understand what they are charged. What is also true is, whether
people read that material or not, they don’t always seem to understand it.
One would hope that over the course of actually engaging in those
transactions, you would see whether you are spending more or less.
When you get a fifty-page document—if you get a two-page document,
with the size of type I get, there are about fifty pages in there—you can’t
tell what it’s going to mean for you. You start actually consuming the
product, and then you find out whether things are more or less costly,
just like whether you go to one grocery store or another. They may be
advertising a gallon of milk for a certain price, but it doesn’t tell you
how much your particular basket of goods is going to come out costing.
So banking is an experience good in that sense. To the extent that
they are charging you an interest rate and a set of fees, it’s the
responsibility of the consumer to understand, at some point, what it’s
costing them. To the extent, over time, you don’t analyze that, that’s
really something that the economic models would fail at. We actually
assume that either a consumer over time would understand that they are
spending a lot or a competitor will come in and say, “Hey, guys, you’re
spending, on average, this amount on your checking account.” Then you
get the simplified checking account solution guys coming in and telling
you, “No. Come to us.”
Your cell phone also has that attribute, to some extent—maybe not
yours; mine might. Your car—you buy an inexpensive car. It costs a
fortune to fix it. There are all those goods in life that have a certain
experience to them. What you are saying is that banking is no longer
something that you can look at from the outside, understand the price
it’s going to be, and just settle. It’s something that, while you consume
it, you are going to understand what it costs you and what the benefits
are. You are going to make your decision that way.
MR. FOER: I would say that obfuscation is a strategy of
differentiation. If all you looked at were interest rates, what kind of
competition would there be? Everybody would have the same interest
rate. So you have to differentiate it, and one of the ways is through all
these different fees and complications. What it says to me is that
consumer protection and antitrust have to go hand in hand. Unless you
have consumers who understand what they are buying, competition is
I just went through an experience with securitized loans. It was a
commercial loan, where I was the borrower. I started hearing things that
I had never known anything about—conduit loans. Those are loans
created in such a way that they can be sent off in securitized packages
and serviced by third parties out somewhere in the world. I was told that
there are certain risks to that. If you get into a little bit of trouble and
somebody is holding one of those loans, they don’t care anything about
you; they are going to foreclose very quickly. My benefit was that I
went with somebody that didn’t do that, and I paid a little bit more. This
happened a couple of days before the bubble burst on the subprimes.172
Instead of pulling the loan like they did with everybody else, I got the
Here’s the point. This was an incredibly complicated transaction
compared to the last time I did a refinancing of this sort, several years
back. Because of securitization, the loan documents are suddenly this
thick. They want all kinds of information, and there are all kinds of
third-party charges. The transaction costs of a loan—this was not a huge
loan—were the same as if it was a $1 billion loan. We are going to start
seeing some kind of reaction to these increased transaction costs, as this
whole process has gotten more and more complicated.
On the other side, as we see more and more transactions on the
Internet—when you borrow on the Internet or you buy something on the
Internet—you are going to be confronted with these lengthy disclosure
statements that you are not going to read.173 You can’t make the deal
until you have checked that you have read it, and you will check. That
brings us into a whole new area, because you will be giving away rights
left and right. You will be required to arbitrate. You might be required
not to participate in class actions.
We are going to have to resolve a lot of these questions that used to
come under the rubric of a contract of adhesion, because these are all
contracts of adhesion. How we are going to make them, in some way,
fair, or avoid the worst abuses that are likely? This is a challenge that is
going to affect banking and all other transactions.
172. See David Anderson, Year in Review 2007: The Subprime Lending Crisis, 71
TEX. B.J. 20 (2008)
(stating that the subprime bubble burst in 2007 when “housing
prices fell in many parts of the country at the same time rates on many subprime loans
adjusted upward and borrowers were required to start repaying principal”)
173. See generally Jeffrey P. Taft, Internet-Based Payment Systems: An Overview of
the Regulatory and Compliance Issues, 56 CONSUMER FIN. L.Q. REP. 42 (2002).
5. For more information, please visit the Fordham University School of Law website, http://law.fordham. edu (last visited Mar . 23 , 2008 ).
6. For more information, please visit the K&L Gates website , http://www.klgates. com (last visited Mar. 1 , 2008 ).
7. For more information, please visit the American Antitrust Institute website , http://www.antitrustinstitute. org (last visited Mar. 1 , 2008 ).
8. 374 U.S. 321 ( 1963 ).
9. Id. at 325.
10. Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994 , Pub. L. No. 103 - 328 , 108 Stat. 2338 (codified in scattered sections of 12 U .S.C.).
11. See 12 U.S.C. § 1831u ( 1999 ) (codification of § 102(a) of the Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994 ).
12. For more information, please visit the Fordham Journal of Corporate & Financial Law website , http://law.fordham.edu/jcfl. htm (last visited Mar . 10 , 2008 ).
13. American Antitrust Institute, Our People: Officers, Directors, Fellows, and the Advisory Board, http://www.antitrustinstitute.org/About_Us/People/index. ashx (last visited Mar. 1 , 2008 ).
14. See id.
15. See id.
16. See id.
17. See id.
18. NERA Economic Consulting, Experts, http://www.nera.com/Expert.asp?e_ID= 1482 (last visited Mar. 1 , 2008 ).
19. See id.
20. See id.
21. See id.
31. See Paul J. Stancil, Atomism and the Private Merger Challenge , 78 TEMP. L. REV. 949 , 969 ( 2005 ).
32. Bank Merger Act of 1960 , Pub. L. No. 86 - 463 , 74 Stat. 129; Bank Merger Act of 1966 , Pub. L. No. 89 - 356 , 80 Stat. 7; Bank Holding Company Act of 1956, Pub. L. No. 110 - 131 , 70 Stat. 133 (codified in scattered sections of 12 U .S.C.); Bank Holding Company Act Amendments of 1970 , Pub. L. No. 91 - 607 , 84 Stat. 1760 (codified in scattered sections of 12 U .S.C.).
33. See Jonathan R. Macey , Bank Failures , Risk Monitoring, and the Market for Bank Control , 88 COLUM. L. REV. 1153 , 1215 - 23 ( 1988 ) (discussing the regulatory scheme of banks under the various acts ).
34. See id.
35. See id.
36. 12 U.S.C. § 1828(c)(5)(B ) ( 2007 ).
37. Herfindahl refers to the Herfindahl-Hirschman Index (HHI), a commonly accepted measure of market concentration . See David J. Saylor, Cable Television Law 2008 : Competition in Video, Internet & Telephony , 925 PLI/PAT. 423, 444 ( 2008 ).
38. NAT'L COMM'N ON ELEC. FUND TRANSFERS , EFT IN THE UNITED STATES: POLICY RECOMMENDATIONS AND THE PUBLIC INTEREST: THE FINAL REPORT OF THE NAT'L COMM'N ON ELEC . FUND TRANSFERS 1 ( 1977 ) [hereinafter FINAL REPORT] .
39. Id .
40. See generally id .
41. Electronic Funds Transfer Act, 15 U.S.C. §§ 1693 - 1693r ( 1978 ).
42. See FINAL REPORT, supra note 38 , at 12-13.
43. See John T. Scott , Nonprice Competition in Banking Markets, 44 S. ECON . J. 594 ( 1978 ).
44. Compare Bank of America, Global Reach, http://corp.bankofamerica.com/pub lic/public.portal?_pd_page_label=products/regions/emea (last visited Mar. 1 , 2008 ), with Citigroup, Global Locations, http://www.citigroup.com/citigroup/global/index. htm (last visited Mar. 1 , 2008 ) ( Note that both banks have branches in the same locations, indicating competition with each other in multiple markets .).
45. See ALLEN N. BERGER ET AL., DIV. OF RESEARCH & STATISTICS AND MONETARY AFFAIRS , FED. RESERVE BD., THE EFFECTS OF COMPETITION FROM LARGE, MULTIMARKET FIRMS ON THE PERFORMANCE OF SMALL, SINGLE-MARKET FIRMS: EVIDENCE FROM THE BANKING INDUSTRY 8-9 ( 2005 ), available at http://www.federalreserve.gov/PUBS/FEDS/ 2005 /200515/200515pap.pdf.
46. See Statement of Janet L. Yellen, 81 Fed. Res. Bull. 1093 , 1100 - 02 ( 1995 ).
47. See Duncan E. Alford , Core Principles for Effective Banking Supervision: An Enforceable International Financial Standard? , 28 B.C. INT'L & COMP. L. REV . 237 , 237 ( 2005 ).
48. See generally Patrick Mulloy & Cynthia Lasker, The Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994: Responding to Global Competition, 21 J. LEGIS. 255 ( 1995 ) (providing a summary of the history of banking law).
49. See Thomas A. Garrett et al., A Spatial Analysis of State Banking Regulation 1 (Fed. Reserve Bank of St . Louis, Working Paper 2003-44C, 2003 ), available at http://research.stlouisfed.org/wp/2003/2003- 044 .pdf (“ By 1994, all states except Iowa permitted statewide branching through the acquisition of existing bank offices . . . [and ] all states except Hawaii permitted some entry by out-of-state holding companies .”).
50. Michael J. de la Merced and Peter Edmonston , Builder of Sallie Mae Deal Has a Daring History , N.Y. TIMES , Apr. 18 , 2007 , at C5.
51. OECD, Roundtable on Competition and Regulation in Retail Banking, at 4 , DAF/COMP/WP2/WD( 2006 )60, http://www.ftc.gov/bc/international/docs/Banking_U S.pdf.
52. See id.
53. Arthur E. Wilmarth , Jr., Too Big to Fail, Two Few to Serve? The Potential Risks of Nationwide Banks, 77 IOWA L . REV. 957 , 1038 - 39 ( 1992 ) (“More than ninety percent of smaller firms rely on local commercial banks as their primary source of financial services . . . .”).
54. See Ash Demirgüç-Kunt et al., The Impact of Bank Regulations, Concentration, and Institutions on Bank Margins 27 (World Bank, Policy Research Working Paper 3030, 2003 ), http://www-wds.worldbank.org/external/default/WDSContentServer/IW3 P/IB/ 2003 /05/30/000094946_03052004111117/Rendered/PDF/multi0page.pdf.
55. Id .
56. See Gina M. Killian , Bank Mergers and the Department of Justice's Horizontal Merger Guidelines: A Critique and Proposal, 69 NOTRE DAME L . REV. 857 , 857 ( 1994 ) (stating that “[t]he Federal Reserve Board (“Fed”) and the Department of Justice (“DOJ”) are two agencies with regulatory jurisdiction over bank mergers .”).
57. United States v. Phila. Nat'l Bank , 374 U.S. 321 , 334 ( 1963 ) (stating that in “the case of a merger, consolidation, acquisition of assets, or assumption of liabilities, the appropriate agency shall also take into consideration the effect of the transaction on competition . . . .”).
58. See Killian, supra note 56 , at 865.
59. See J. Robert Kramer II , Antitrust Review In Banking and Defense , 11 GEO. MASON L. REV. 111 , 117 ( 2002 ) (stating that cooperation between the DOJ and Federal Reserve Board has grown, and that they “routinely hold joint meetings with the parties and collectively give parties feedback on competitive concerns”).
60. See Catherine M. Bejerana , Capitalist Manifesto: The Inadequacy of Antitrust Laws In Preventing the Cannibalism of Competition, 2 ASIAN-PAC . L. & POL'Y J . 144 , 160 - 63 ( 2001 ) (stating that “[t]he Department of Justice's analysis for bank mergers is essentially similar to that of other industries” and providing the other laws and rules applicable to bank mergers).
61. 15 U.S.C. § 18 ( 1996 ).
62. Phila . Nat'l Bank , 374 U.S. at 321.
63. Id . at 330.
64. United States v. Phila. Nat'l Bank , 201 F. Supp . 348 , 372 (E.D. Pa . 1962 ).
65. Phila . Nat'l Bank , 374 U.S. at 331 ( stating that “the trend toward concentration
87. See id.
88. See Susan Harriman, Note, Parens Patriae Actions on Behalf of Indirect Purchasers: Do They Survive Illinois Brick? , 34 HASTINGS L.J. 179 , 184 ( 1982 ) (stating that one purpose of the Hart-Scott-Rodino Act was “to deter future antitrust violations” (quoting H.R. REP . NO. 94 - 499 , at 3 ( 1975 ))).
89. Albert A. Foer , Toward Guidelines for Merger Remedies, 52 CASE W. RES. L. REV . 211 , 212 ( 2001 ) (stating that the “Hart-Scott-Rodino Antitrust Improvements Act of 1976 (“HSR”) has had the largely unanticipated effect of moving the merger law development and enforcement policy process from a regime of post hoc adjudication to ad hoc regulation and pre hoc administrative negotiation”) (footnote omitted).
90. Robert Pitofsky , Chairman, Fed. Trade Comm'n, Remarks at the Cutting Edge Antitrust Conference Law Seminars International (Feb. 17 , 2000 ), available at http://www.ftc.gov/speeches/pitofsky/restruct.shtm (stating that “few merger cases are fully litigated in the United States courts or the FTC's administrative process”).
91. See WILLIAM C. HOLMES , ANTITRUST LAW HANDBOOK § 6:6 (West 2007 ) (defining the Guidelines as guidelines jointly issued by the Justice Department and the Federal Trade Commission which set forth case selection criteria that both agencies use when screening mergers for potential challenge under section 7 of the Clayton Act, and stating that the “[g]uidelines employ a basic five-part analysis to determine whether a particular merger threatens to create or further entrench market power”).
92. See Phila . Nat'l Bank , 374 U.S. at 363-64 ( holding that “the concentration of firms . . . is so inherently likely to lessen competition substantially that it must be enjoined in the absence of evidence clearly showing that the merger is not likely to have such anticompetitive effects”) (emphases added).
93. See United States v. Phillipsburg Nat'l Bank & Trust Co ., 399 U.S. 350 , 377 (recognizing two aspects of market structure that the court thought might well rebut the 105 . See Notice, supra note 99, § 1 . 11 .
106. See United States v. Phila. Nat'l Bank , 374 U.S. 321 , 356 ( 1963 ).
107. See id. at 356-57.
108. Compare id. at 321 (stating that Phila . Nat'l Bank was decided on June 17 , 1963 ), with Baker, supra note 94 , at 148 ( stating that the modern era of merger regulation began with the creation of the 1982 Merger Guidelines ).
109. See Notice, supra note 99, § 1.2.
110. See Bejerana, supra note 60 , at 160.
111. See Jonathan R. Macey , Commercial Banking and Democracy: The Illusive Quest for Deregulation, 23 YALE J . ON REG. 1 , 1 ( 2006 ).
112. See Yvonne S. Quinn , Practical Aspects of Defending Bank Mergers Before The Federal Reserve Board and the Department of Justice, 62 ANTITRUST L .J. 91 , 94 ( 1993 ).
113. See Kramer, supra note 59 , at 116.
114. See Killian, supra note 56 , at 857.
115. See Kramer, supra note 59 , at 116.
116. See United States v. Phila. Nat'l Bank , 374 U.S. 321 , 356 ( 1963 ) (defining the relevant geographic market as the section of the country “in which to appraise the probable competitive effects” of a proposed merger).
117. See id. (defining the relevant product market as the line of commerce “in which to appraise the probable competitive effects” of a proposed merger).
118. See generally Killian, supra note 56 (explaining the process that the Fed and the DOJ use to analyze a bank merger).
119. See Orson Swindle , Comm'r , Fed. Trade Comm'n, Remarks Before the 126th Annual Meeting of the New York State Bar Association (Jan. 23 , 2003 ), available at http://www.ftc.gov/speeches/swindle/030123osremarks.shtm (stating that “the Commission announced its acceptance of a consent agreement in settlement of its challenge to Wal-Mart's acquisition of Supermercados Amigo, the largest supermarket chain in Puerto Rico” ).
120. See Complaint at 1, Advanced Micro Devices , Inc. v. Intel Corp., Civ . Action No. 05 -441 -JJF (D. Del . June 27, 2005 ).
121. See id.
123. FDIC, FDIC STATEMENT OF POLICY ON BANK MERGER TRANSACTIONS 5145 ( 2003 ), available at http://www.fdic.gov/regulations/laws/rules/5000- 1200 . html (stating that “[t]he relevant geographic market(s) includes the areas in which the offices to be acquired are located and the areas from which those offices derive the predominant portion of their loans, deposits, or other business”).
124. See U.S. DEP'T OF JUSTICE , BANK MERGER COMPETITIVE REVIEWINTRODUCTION AND OVERVIEW ( 2000 ), http://www.usdoj.gov/atr/public/guidelines/647 2.pdf.
125. See Saylor, supra note 37.
126. Concentration ratio is a percentage figure representing the aggregate market share of a given number of leading firms in an industry . See Michael O . Finkelstein & Richard M. Friedberg , The Application of an Entropy Theory of Concentration to the Clayton Act, 76 YALE L .J. 677 ( 1967 ) (stating that the measure of economic concentration most frequently used by lawyers and economists in the horizontal merger cases is the concentration ratio ).
127. See U.S. Dep 't of Justice, The Herfindahl-Hirschman Index (HHI), 131. NOW stands for Negotiable Order of Withdrawal. See James D. Lawlor, Power of Savings Bank or Similar Institution to Provide Checking Facilities or Negotiable Orders of Withdrawal (NOW) to Customers, 64 A.L.R. 3d 1314 ( 1975 ).
132. Dr . Anne Gron, Bank Merger Analysis: Defining the Relevant Market (Nov. 13 , 2007 ) (unpublished PowerPoint slides, on file with the Fordham Journal of Corporate and Financial Law) .
138. See id.
139. Community Reinvestment Act of 1977 , 12 U.S.C. §§ 2901 - 2908 ( 1997 ).
140. The structure-conduct-performance approach to antitrust analysis sets elements of performance as the main goals . See Leonard W. Weiss, The Structure-ConductPerformance Paradigm and Antitrust, 127 U. PA. L. REV . 1104 ( 1979 ).
141. See Thomas W. Hazlett, Symposium, Is Antitrust Anticompetitive?, 9 HARV.
146. For an explanation of market-structure studies, see Almarin Phillips, Market Concentration and Performance: A Survey of the Evidence, 61 NOTRE DAME L . REV. 1099 , 1105 ( 1986 ).
147. See id.
158. For more information, please visit the Freddie Mac website , http://www.freddiemac. com (last visited Mar . 22 , 2008 ).
159. For more information, please visit the Ginnie Mae website , http://www.ginniemae. gov (last visited Mar . 22 , 2008 ).
160. For more information, please visit the Fannie Mae website , http://www.fanniemae. com (last visited Mar . 22 , 2008 ).
161. See R. Stephen Painter , Jr., Subprime Lending , Suboptimal Bankruptcy: A Proposal to Amend §§ 522(F)(1)(B) and 548(A)(1)(B) of the Bankruptcy Code to Protect Subprime Mortgage Borrowers and Their Unsecured Creditors, 38 LOY. U. CHI. L.J. 81 ( 2006 ) (defining a subprime mortgage as a loan catered to higher risk borrowers).
162. See Wayne Passmore et al., GSEs , Mortgage Rates, and the Long-Run Effects of Mortgage Securitization 1 n.2 (Fed . Reserve Bd., Working Paper No. 2001 - 26 , 2001 ), available at http://ssrn.com/abstract=275008.
163. REIT is short for real estate investment trust, which is a “financial vehicle that allows investors to pool their capital for participation in real estate ownership or mortgage financing .” Jack H. McCall , A Primer on Real Estate Trusts: The Legal Basics of REITs, 2 TRANSACTIONS: TENN . J. BUS . L. 1 , 2 ( 2001 ).
166. See Fred Galves, Might Does Not Make Right: The Call for Reform of the Federal Governments D'Oench, Duhme and 12 U.S.C. § 1823(E) Superpowers in Failed Bank Litigation , 80 MINN. L. REV. 1323 , 1333 - 39 ( 1996 ) (stating that the U.S. government bailed out multiple failing savings and loan associations in the 1980s and 1990s ).
167. See generally Tim McCarthy, Refining Product Market Definition in the Antitrust Analysis of Bank Mergers, 46 DUKE L .J. 865 ( 1997 ).
168. Nelson D. Schwartz , Perhaps Too Many Acts in the Corporate Tent , N.Y. TIMES , Nov. 11 , 2007 , § 4, at 5.
169. See Press Release, Citigroup Inc., Citibank Agency and Trust Announces Opening of a New Office in Orange County , California (Oct. 12 , 2004 ), available at http://www.citibank.com/citigroup/press/2004/data/041012a.htm.