Antitrust problems can compromise or even derail an otherwise
well-conceived merger or acquisition. An unanticipated challenge
from antitrust enforcers can send an acquirer scrambling for
remedies. Often, this leads to divestitures at fire-sale prices and
a serious erosion of shareholder value. If no acceptable remedies
can be devised, the transaction may even have to be abandoned.
Yet serious as it is, antitrust risk is being managed with
growing success. Antitrust enforcement is no longer a realm of
ill-defined standards and unpredictable outcomes. Spurred by
advances in analytic methods as well as computing power, economists
both inside and outside the enforcement agencies have brought
considerable science to bear in analyzing the competitive impact of
proposed transactions. Thanks to the increasing use of econometrics
- the application of sophisticated statistical techniques to
analyzing economic data - questions once debated using qualitative
information and even pure rhetoric can now be answered on the basis
of hard statistical evidence.
The development of objective methods for addressing antitrust
issues has taken much of the guesswork out of the assessment of
antitrust risk. Rather than waiting for the Department of Justice,
the Federal Trade Commission or the European Commission's Merger
Task Force to review a transaction, senior business leaders can
have their legal and economic consultants pre-screen it early in
the planning process. This pre-screening can provide some
confidence that the deal is likely to sail through antitrust review
or attract close scrutiny.
Pre-deal screening provides business leaders with critical
insights. In some cases, senior executives may be willing to accept
a relatively high level of antitrust risk because a deal has the
potential to provide correspondingly high rewards. In other cases,
alternative transactions that offer lower financial returns may
actually make better strategic sense after antitrust risk has been
taken into account. Pre-deal screening gives senior executives the
chance to make these strategic choices with open eyes.
Econometrics comes of age
Antitrust laws have existed in the US for more than 100 years.
Under the terms of the Sherman Act (1890): "Every contract,
combination in the form of trust or otherwise, or conspiracy, in
restraint of trade or commerce among the several States, or with
foreign nations, is declared to be illegal." The Clayton Act (1914)
prohibits mergers and acquisitions where the effect "may be
substantially to lessen competition, or to tend to create a
monopoly." Under the Hart-Scott-Rodino Act, a 1976 amendment to the
Clayton Act, anyone considering a large acquisition must notify
both the Federal Trade Commission and the Department of Justice in
advance of the transaction. In Europe, the Merger Task Force
conducts enforcement procedures under merger regulations passed in
1990.
In the US, the explicit purpose of antitrust enforcement is to
protect consumers. In the words of the Federal Trade Commission,
practices that significantly restrict competition and that have no
overriding business justification are likely to harm consumers by
"increasing prices, reducing availability of goods or services,
lowering quality or service, or significantly stifling
innovation."
For decades, anyone in the US contemplating a merger that might
arouse antitrust concerns could be certain of review by the Federal
government and, possibly, by individual states, too. Before the
mid-1990s, however, antitrust enforcement had much in common with
the enforcement of pornography and decency laws: the standards were
largely in the eyes of the beholder. This made it very difficult to
predict whether a given transaction would trigger antitrust
suspicions or, in fact, actually be proscribed.
A decade ago, antitrust analysis was essentially qualitative. If
company A wanted to merge with company B, economists looked at
their shares of some market and tried to decide whether the
combined share would be too big. While nominally a numerical
calculation, the quantitative aspects of the analysis were actually
trivial. A much more important question was how to define the
relevant market. If companies A and B were in the auto business,
was the relevant market all cars in the US, all cars worldwide,
small cars, big cars, cars plus trucks? And once the shares were
added up, what did it mean? Is a 40% market share too much? Is 50%
too much? Why? And, more particularly, why might 40% be too much in
one industry and 50% be acceptable in another?
Answering such questions was traditionally a matter of argument.
The parties to the transaction brought one set of evidence, the
enforcers brought a different set, and the two sides fought it out.
As often as not, the outcome depended more on the relative
persuasiveness of the two sides than on market realities.
Competitive questions were posed in a way that was not susceptible
to quantification. And even if they had been, the available
computing power would have been inadequate to answer them.
Resolving an antitrust issue empirically typically involves truly
complicated calculations. It is only in the last 10 years that
technological advances have made this kind of analysis
feasible.
Today, it is practical to determine how prices in a given market
are actually set at a nuts-and-bolts quantitative level. Computing
power is just one of the prerequisites for such analysis. Equally
important is the availability of relevant data. For example, if the
parties to a merger produce consumer goods of the sort sold in
supermarkets, retail scanner data will surely be available on the
sales of these products. These data are available from vendors such
as Nielsen and IRI, which track actual supermarket and convenience
store checkout data on a sample of stores in each geographic area.
They capture literally everything that gets scanned through the
checkout aisles of these stores. This jumble of transactions is
then sorted into product categories, brands, package sizes and so
on, and aggregated up to the weekly level. This amounts to a
mind-numbing abundance of data.
Outside the supermarket arena, there are troves of third-party
data in such unexpected markets as agricultural chemicals. Even in
markets where there has been no syndicated research, individual
companies may have conducted their own custom-designed surveys.
Often, these generate enough data to support econometric
evaluation.
The empirical analysis of real economic data has already gained
currency in the US, and it is gaining currency in Europe. Today,
econometric analysis is persuasive and sometimes dispositive in
resolving antitrust issues. This is not to suggest that complete
unanimity has been achieved: we cannot yet put data into a black
box, turn the crank, and read the results. Nonetheless, while
reasonable people may still disagree about what the data mean, the
areas of disagreement are far more circumscribed than they were as
little as 10 years ago.
When a transaction involves hundred of millions, if not
billions, of dollars or euros, the antitrust enforcers are certain
to examine it. If anything gives them pause, they are sure to look
at the relevant data. In the light of this reality, senior business
leaders should arm themselves with the same capability.
The bread market in Philadelphia
A merger in the food industry illustrates the kind of
perspective pre-deal screening can provide. Weston Foods, which
owned the Stroehmann and Maier brands of bread, wanted to buy
Bestfoods Baking, which owned the Freihofer's and Arnold brands.
The acquirer anticipated that antitrust enforcers would be
concerned that the transaction would endow the merged company with
the market power needed to raise prices unilaterally on its bread
products. Evaluating this kind of problem is known as a unilateral
effects analysis.
Philadelphia was one of four cities to be analyzed. Weston's
Stroehmann and Maier brands accounted for over 40% of the white
bread sold in Philadelphia, while Bestfoods' Freihofer's brand was
relatively small, with about 4% of sales. With this market
structure, a simple unilateral effects model based solely on market
shares would predict hefty merger-induced price increases on
Freihofer's bread. While such increases would undoubtedly cause the
Freihofer's brand to lose some sales, the model would predict that
Weston's brands, with over 40% of the market, would recapture a
substantial portion of these lost sales.
Was this hypothesis correct? The answer hinged on whether the
Freihofer's and Weston brands were, in fact, particularly close
substitutes for one another. This is precisely the kind of question
that econometric analysis is designed to answer. In economic terms,
it is a question that revolves around own-price elasticities and
cross-price elasticities of demand.
The result of the analysis was good news for the merging
parties. After analyzing extensive supermarket sales data, NERA
Economic Consulting found that the brands in question were not
close substitutes for one another. In other words, if the price of
the Freihofer's brand bread were to be raised, relatively few of
the lost sales would go to Stroehmann or Maier. Freihofer's would
lose sales to private label brands and other brands, but not to
Stroehmann or Maier, their combined market share notwithstanding.
The merged firm would, therefore, have little incentive to raise
prices. Ultimately, the Department of Justice accepted NERA's
reasoning about the transaction's potential impact on
competition.
Defining a market
Today, as in the past, antitrust issues often revolve around
market definition. Here, too, econometrics can play an extremely
helpful role.
Several years ago, Brown & Williamson acquired American
Tobacco. A key question was whether premium and discount cigarettes
competed with each other. As the two companies were much more
prominent in discount cigarettes than they were in the premium
segment, this question had important implications for the
measurement of market shares. If all cigarettes are one large
market, the parties' shares were modest. On the other hand, if the
discount segment were a market unto itself, the transaction would
make for substantial increases in concentration. Which was the
relevant market?
Econometrics is well suited to address this question. Under
merger guidelines issued by the Department of Justice and the
Federal Trade Commission, a market is defined as the smallest group
of products for which a hypothetical monopolist could profitably
impose a small but non-transitory price increase of, say, 5% or
10%.
Deciding whether discount and premium cigarettes constitute
separate markets translates into a question about elasticity of
demand: Do consumers respond to price changes by substituting one
type of cigarette for the other? Based on the profit margin and the
size of the hypothetical price increase, a relatively simple
formula gives the critical elasticity of demand. In the case of
discount cigarettes, this number turned out to be 1.82. If the true
elasticity were greater than 1.82, discount cigarettes would not
constitute a market. If the elasticity were less than 1.82, the
discount segment would be a relevant market unto itself. In fact,
econometric analysis found that the true elasticity was 2.5, which
was substantially greater than 1.82. Discount cigarettes, we
showed, should not be considered a market. This analysis helped
remove a serious impediment to the acquisition.
Developing antitrust remedies
Of course, econometric analysis does not always produce such
favourable results. At times, we find that the merging firms'
products do, in fact, compete head-to-head with one another. Even
so, all is not necessarily lost. Econometric analysis, run in
reverse, can also be used to identify suitable fixes for the
problem. Identifying a divestiture or other remedy early in the
planning process can contribute substantially to a merger's overall
success. When divestitures take place under severe pressure,
sellers are often forced to settle for disappointing prices. With
advance notice and greater time to plan, they can generally achieve
better results.
Some years ago, Vail Resorts Inc owned just two ski areas in
Colorado - Vail itself, and Beaver Creek. When a competitor put the
Breckinridge, Keystone and Arapahoe Basin ski areas on the market,
Vail Resorts was eager to buy them. The company already had
first-class skiing at Beaver Creek and Vail. Of the new areas,
Breckinridge had decent slopes in an attractive setting, Keystone
was more family-oriented, and only Arapahoe Basin had truly
superlative skiing.
Skiers who fly to their vacations have a wide range of choice in
Wyoming, Utah and elsewhere, so antitrust concerns were focused on
Denver-area skiers who drive to slopes within some 60 miles of
home. Because no more than 10 such slopes were available to these
skiers, the US, Department of Justice expressed concerns that
combining five of them under one roof would represent a serious
concentration of ownership.
In contrast to the markets for bread and cigarettes, the ski
resort market has no third-party sources of data. Therefore, we
therefore conducted our own survey of Denver-area skiers. We
learned that they are highly price-sensitive consumers who will
choose the ski area that offers lift tickets at the lowest
available price.
Our econometric analysis indicated that a single divestiture -
of Arapahoe Basin - would eliminate antitrust concerns raised by
the acquisition. Only Arapahoe Basin would add to Vail Resort's
share in the market for truly top-level skiing. While the other
resorts would broaden its portfolio, they were not the kind of
close substitutes for other Vail holdings that would raise concerns
about unilateral price increases. Thanks to this divestiture, Vail
Resorts was able to satisfy the antitrust enforcers and reap most
of the benefits of the merger.
Pre-deal uses of econometric analysis
Whether a company is a buyer or seller, knowing something about
how the econometrics are going to play out is helpful in many ways,
including managing the antitrust risk and choosing between
alternative potential deals.
One big advantage is in timing. If data are readily available
either in-house or from a third-party source, a buyer can evaluate
a proposed acquisition before making a first approach to the
target. Particularly in cases where the superficial facts suggest
that there might be antitrust issues, econometric analysis of the
marketplace data may be a prerequisite for getting the target to
agree to discuss a deal.
A firm that is trying to sell assets can use econometric
analysis to handicap the antitrust risks posed by a variety of
alternative buyers. This handicapping can help determine which
offer is, in economic terms, the best one. Depending on the
seller's imperatives, a bid that is 10% higher but presents
substantially more antitrust exposure may simply not be rewarding
enough to compensate for the added risk.
Pre-screening does not necessarily entail the kind of full-blown
analysis that precedes a presentation to an enforcement agency or
court. Some of the potential transactions may raise such obvious
antitrust concerns that they really have no chance of success.
These can often be eliminated through a relatively brief analysis.
A more feasible set of targets can then be evaluated in greater
detail. Detailed analysis can ensue once the field has been
narrowed further.
For many reasons, mergers and acquisitions often fail to deliver
the anticipated boost to shareholder value. By using econometrics
to pre-screen proposed transactions, both buyers and sellers can
now approach deals with a clear view of the antitrust risks they
are likely to face. Forewarned is forearmed.
This article first appeared in "Viewpoint, the Marsh &
McLennan Companies journal". MMC owns the copyright for the
article.
Author biography
Sumanth
Addanki
NERA Economic
Consulting
Sumanth Addanki specializes in antitrust, intellectual property,
and the evaluation of commercial damages.
In the antitrust area, Addanki has analyzed the competitive
consequences of numerous mergers in a wide range of industries,
from agricultural products and chemicals through consumer products,
medical devices, pharmaceuticals and semiconductors, among many
others; he has frequently presented the results of his analyses to
US antitrust agencies and courts. He has addressed liability and
damages issues involving allegations of predatory pricing,
monopolization, price discrimination, resale price maintenance,
tying and other antitrust violations, again in a variety of
different industries. Many of his antitrust inquiries have focused
specifically on intellectual property and its unique role in the
analysis of market power and competitive effects.
Addanki has testified before state and federal courts, the
Federal Trade Commission and various other regulatory bodies. His
articles and speeches have been invited and published by the
American Bar Association, the Practicing Law Institute, European
Competition Law Review and many other comparable institutions.
NERA Economic Consulting
50 Main Street, 14th
Floor
White Plains, New York 10606
United States
Tel: +1 914 448 4000
Fax: +1 914 448 4040
Web: www.nera.com