the new regulation of consumer the payments and lending industries

Report
ANTITRUST ECONOMICS 2013
David S. Evans
University of Chicago, Global Economics Group
Elisa Mariscal
CIDE, Global Economics Group
TOPIC 12: HORIZONTAL MERGERS AND THE ANALYSIS
OF COMPETITIVE EFFECTS
Date
Topic 12| Part 1
10 October 2013
Overview
2
Part 1
Legal and
Economic
Background of
Mergers
Merger
Screening
Unilateral Effects:
Economic Theory
Part 2
Coordinated
Effects:
Economic
Theory and
Evidence
Two-Sided
Platforms
Empirical
Methods
Key Issues
3
Mergers can generate various efficiencies but can also increase
market power and ultimately harm consumers. Merger to monopoly
is an extreme example.
Economics can assess whether a merger is likely to harm consumers
by estimating price and other effects
Merger assessment is difficult because it makes predictions about the
future and economists (and the authorities) often lack good data
and the time (given time schedules) to analyze it.
4
Legal and Economic
Background on Mergers
Definition of Mergers
5
Merger results in transfer of ownership and
control of a company (or a part of a company)
Different types of combinations
• “Merger”: A and B agree to combine to form C
• “Acquisition”: B sells to A which combines B into A
• “Takeover”: A buys control of B from shareholders
against wishes of management
Use the term “merger” for convenience to refer
to mergers, acquisitions, and takeovers without
referring to an “acquirer” and “target” (A and B
respectively in the example above) .
Horizontal vs. Vertical Mergers
6
Horizontal merger of B and C in market of A, B, and C which have substitutable
products
A
C
B
f
e
Vertical merger of A and f where f supplies a downstream service to A.
Possible Benefits of Mergers
7
Permit the exchange of property to higher valued uses; ultimately mergers are
about selling property. (Why is Nokia selling and Microsoft buying Nokia’s handset
unit?)
Acquirer finds it cheaper to buy than to build (Why did Facebook buy
Instagram?)
Mergers can generate economies of scale and scope; reduce duplicative costs;
create synergies through complementary technologies. (One of the arguments
for EU efforts to break down national boundaries e.g. in capital markets.)
Encourage innovation and investment because an acquisition provides major
potential source of reward for the target. (eBay just bought processor Braintree
for $800 million providing exit for Braintree’s entrepreneurs and investors.)
Mergers and takeovers—and the mere threat of them—can discipline corporate
management and thereby solve separation of ownership and control problem
(CEO loses job in takeover and would like to avoid that.)
Possible Harm from Mergers
8
Mergers could increase prices and reduce output through
increased concentration. Merger to monopoly is the extreme
case.
Mergers could also reduce innovation, service, and other desirable
non-price aspects of competition.
Mergers could increase entry barriers and thereby result in durable
market power.
Vertical mergers could facilitate the extension of market power
from one market to another or help maintain market power
through control of essential suppliers.
The Policymaker’s Error Cost Conundrum
9
If competition authorities block too many mergers, the potential for
efficiency gains will be lost.
If competition authorities allow too many mergers the merged firms
may be able to exploit market power and thereby harm
consumers.
In both cases merger policy will affect how companies decide to
invest and innovate.
Merger policy can be seen as application of “error-cost
approach.” It balances false positives and false negatives in
applying merger screens and tests.
10
Merger Screening
Many Countries Have Legal Frameworks for
Clearing Mergers
11
Only “economically” significant mergers are reviewed where the
acquirer or target firm or both are large enough (reach certain
thresholds).
Screening measures are then used to quickly approve “nonproblematic mergers” (HHI and change in HHI).
Mergers are blocked if they lead to “significant impediment to
efficient competition” (EC) or “substantial lessening of competition”
(e.g. UK, US).
These legal frameworks are consistent with the view that mergers are
presumed to be procompetitive unless there are reasons to believe
otherwise.
The HHI Is A Common Method of Screening
12
Herfindahl-Hirschman Index (HHI) is calculated by squaring the
market share of each firm competing in a market, and then
summing the resulting numbers.
HHI = S12 + S22 + S32 + ... + Sn2 ; e.g. 302 + 302 + 402 =3400
The HHI can range from a minimum of close to 0 (a perfectly
competitive industry) to a maximum of 10,000 (a perfectly
monopolistic industry) if calculated by squaring percentages (like
10%) or between 0 and 1 if calculated by squaring fractional
shares like 0 .1 (so 3400 or .34 in the example above).
Change in HHI Indicates Change in Concentration
13
Suppose firms 1 and 2 merge
HHIpost = (S1 + S2 )2 + S32 + ... + Sn2
HHIpre = S12 + S22 + S32 + ... + Sn2
∆HHI = HHIpost – HHIpre= 2 S1S2
Example of Calculating HHI and Change in HHI
14
Pre-Merger
Post-Merger
Firm
Share
Firm
Share
A
50%
A
50%
B
25%
B+D
35%
C
15%
C
15%
D
10%
HHI
3450
HHI
3950
Change in HHI=2 x 25 x 10 = 500
OFT’s Merger Guidelines
15
HHI
Investigate?
< 1,000
NO
> 1,000
and
<1,800
YES IF  HHI > 100
NO OTHERWISE
> 1,800
YES IF  HHI > 50
NO OTHERWISE
HHI and Change in HHI Serve as Screening Devices
Is the HHI Screen Based on Economics?
16
The HHI framework is based on theory that assumes Cournot
competition and homogeneous goods. These assumptions are
unlikely for many mergers.
The HHI framework also does not have significant empirical support.
The HHI framework also requires the definition of a market. As we’ve
seen, economics does not support defining hard market boundaries.
Errors in market definition could result in too many or too few mergers
passing screen.
Difficulties with HHI and market definition have lead some economists
to argue in favor of taking a preliminary look at competitive effects.
That has its own problems. See Upward Pricing Pressure Tests later.
Best practice is to use a variety of evidence to assess whether a
merger is sufficiently likely to cause problems that warrant more
investigation.
Key Steps in Regulatory Analysis of Mergers
17
Define markets
Assess overlaps between products of merging firms
Apply HHI (or other) screens and continue only if there is a possible
problem
Determine whether overlaps result in significant harm to
competition such as significant price increases
Consider remedies that could eliminate competitive problems
through divestitures
Decide whether to block a merger or approve possibly with
remedies
Unilateral vs. Coordinated Effects
18
Unilateral effects: The acquiring firm may raise price because the
merger gives it more market power.
Coordinated effects: The merger either strengthens or makes
possible tacit collusion among firms thereby leading to an increase
in price.
A given merger could involve either or both types of effects.
Commonly Used Economic Framework for Analysis
of Mergers
19
Market definition based on hypothetical monopolist test
Estimation of change in prices as a result of the merger, ignoring
efficiency effects using diversion ratios, demand estimation, natural
experiments, or company documents.
Estimation and consideration of efficiency effects (in practice this is
often not taken seriously).
Assessment of whether prices increase by small and significant nontransitory amount after accounting for change in competitive
conditions and efficiencies or whether benefits of efficiencies
outweigh harm from increased prices.
Mergers are presumed to be pro-competitive
20
Mergers are commonplace in the market economy.
Most are not examined by competition authorities because they do
not involve large enough acquirers or targets.
Most mergers that are examined by competition authorities are
cleared without conditions. A few are cleared after the divestiture of
some overlapping products.
Mergers are an integral part of the market process and are presumed
to foster competition and innovation.
In practice, merger prohibitions are a small fraction of all mergers—
but without rules there would likely be many more problematic
mergers. Firms seldom propose 3-2 mergers, for example.
Most Mergers before EC Are Cleared
21
10.0%
9.0%
8.0%
7.0%
6.0%
Percent of case for Phase II
Percent of cases withdrawn or prohibited
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
Most mergers are cleared quickly in Phase I. Most mergers that go into Phase II are cleared
although often subject to some divestitures. Only mergers that are large enough to be “noticed”
under the guidelines are considered here. So an even larger fraction of all mergers are approved.
22
Unilateral Effects: Economic
Theory
Internalization of Price Effects Is Key Issue
23
Suppose Firms A and B sell two products that compete with each other
When Firm A raises its price it loses some sales and profits to Firm B. In
effect Firm B partly benefits when Firm A raises its price because some
customers are going to switch to it and Firm B makes profits from them
instead of Firm A.
Now suppose Firm A buys Firm B but operates A and B as separate
divisions.
The new firm will capture the benefits of profits from a higher price from
both A and B so the benefits from raising prices increase. The same
logic applies to B raising its price.
The new owner of A+B will find that it can make more money by raising
the prices of both A and B because it is also capturing the “spillover”
effects between the products. Economists say the acquirer and target
“internalize” the full spillovers between each other.
Limitations on Raising Price
24
Virtually all models of market behavior show that when Firms A and B
sell substitute products their combination will always tend to increase
price. The only question is whether there are factors that will limit that
increase (in addition to efficiencies considered later).
Factors include:
• Degree of substitution between the products (close or distant?)
• Availability of substitutes of other products (in practice market shares
provides a proxy for substitutes)
• Change in supply of substitutes through entry or expansion (could new
suppliers come in?)
• Countervailing buyer power (a big buyer like Tesco has greater ability to
prevent price increases than does a small business or consumer)
• Efficiencies that offset tendency to raise price (e.g. a firm passes through
enough of the cost savings to offset the price effect).
Merger of Firms in an Oligopoly Industry
with Homogeneous Demand
25
In an oligopoly industry (with
Cournot competition) market
price is lower than a
monopolist would charge,
but higher than the
competitive price so each
firm has some degree of
market power
P
D(P)
PM
PCOURNOT
MONOPOLY
COURNOT
COMPETITION
MC
Deadweight loss is also lower
than that in the monopoly
case in the same market, but
still positive
qMA,B Q COURNOT
qC
Q
The Price-Cost Margin for Industry Increases With
Concentration Under Cournot Competition
26
n
HHI 
S
2
i
m 
p
i
Homogenous product industry with Cournot
competition and symmetric costs:
pc
m 
HHI
EM
Consequences of a Reduction in the Number of
Firms with Homogeneous Products
27
Basic economic theory shows that reducing the number of firms
results in higher prices (Cournot Model)
P
D(P)
After a merger, we see:
• Higher prices
• Lower output
MONOPOLY
N=2
PM
N=3
N=4
COMPETITION
But:
MC
• How significant is the effect?
• Is it counteracted by
efficiencies?
qM
qC
Q
Mergers of Firms that Produce Differentiated
products
28
Key issue is how substitutable the merging products are — we are
obviously more concerned with the combination of two small car
producers products than we would be with the merger of small car
producer with a large car producer.
The “diversion ratio” — the portion of sales that gets “diverted”
between the merging parties rather than to non-merging rivals —and
the cross-elasticity of demand are critical although related measures of
substitution to consider.
Market shares are less informative for this analysis since the market will
ordinarily include products that vary in their degree of substitution with
the merging products. Adding apples and oranges or at least
McIntonish and Golden Delicious apples.
With differentiated products shares should be based on “value” rather
than units because at least “value” takes quality differences into
account.
Example of differentiated products
29
In differentiated products, the closeness
of competition matters
• In the event of a 5% price rise, firm A loses
20% of its customers
• The post-merger firm A+B only loses 10%
of its customers after a price rise because
it captures 10% through common
ownership
• If the critical loss for A is between 10-20%,
a merger with B will make profitable a
price rise which would not have been
profitable pre-merger
• If the critical loss is between 16-20%, a
merger with C will also make a price rise
profitable, and so on
Overall market shares matter less than
the diversion ratio
Can result in a small but significant price
increase even if modest change in HHI
Response of customers of Firm
A to a 5% price rise
Stay with firm A
80%
Switch to
B
10%
C
4%
D
2%
E
2%
F
2%
Back of the Envelope Merger Analysis for
Differentiated Products
30
It is possible to estimate the price change for firm A from a proposed
merger from the diversion ratio between firm A and B (D) and the
price-cost margin of firm A (m)
• Percent price change = mD/(2(1-D))
• In the case of linear demand with constant unit costs.
If margin is 20% and diversion ratio is10% then the price change is:
.2x.1/2x.9=.02/1.8=.0111 or 1.1%.
• If the margin is 50% and the diversion ratio is 20%.
• Then the price change is .5x.2/2x.8=.0625 or 6.3%
Can derive similar formulas with different assumptions and factor in
efficiency effects. Can also extend to accounting for increasing
prices for both products
Upward Pricing Pressure Test
31
Farrell and Shapiro have proposed the use of a variant of this to assess
whether a merger should be given more scrutiny.
The basic idea is to give some credit for possible efficiencies (E) and
measure whether given these efficiencies there will be a tendency for
prices to rise. If there is, then further scrutiny is warranted.
One version of the test is if Dm – E(1-m) > 0 then the merger is subject
to scrutiny where D is the symmetric diversion ratio between the two
firms, m is the common price margin, and E is efficiencies as a
percent of cost.
If the efficiency credit is 10% then with a diversion ratio of 10% and a
margin of 50% the equation yields .1 x .5 - .1 x .5 = 0 so there is no
pricing pressure. If the diversion ratio was greater than 10% or the
margin was greater than 50% the merger would be subject to
scrutiny.
Efficiencies and Merger to Monopoly
32
The firms in a competitive industry with p=MC are all bought up and
merged into monopoly.
Surplus transferred from
consumers to producers
P $7
Merger
Surplus from
cost savings
$6
$5
PPost
Consider also when
marginal cost
decreases from $3.00
to $2.50
Dead Weight Loss
$4
Pre-merger Marginal Cost
PPre $ 3
Post-merger Marginal Cost
$2
$1
Demand
$0
0
1
2
3
QPost
4
5
6
QPre
7
8
9
10
Q
MR
In this example price to consumers increases and social welfare declines
less than consumer welfare because of efficiencies.
Efficiencies and Merger to Monopoly
33
In this example price increases, consumer welfare declines, but social
welfare increases because the efficiency benefits outweigh the
deadweight losses to society.
Surplus transferred from
consumers to producers
P $7
Merger
Surplus from
cost savings
$6
$5
PPost
Consider also when
marginal cost
decreases from $3.00
to $1.50
Dead Weight Loss
$4
Pre-merger Marginal Cost
PPre $ 3
$2
Post-merger Marginal Cost
$1
Demand
$0
0
1
2
3
QPost
MR
4
5
6
QPre
7
8
9
10
Q
End of Part 1, Next Class Part 2
34
Part 1
Legal and
Economic
Background of
Mergers
Merger
Screening
Unilateral Effects:
Economic Theory
Part 2
Coordinated
Effects:
Economic
Theory and
Evidence
Two-Sided
Platforms
Empirical
Methods

similar documents