Quantitative analysis in M

1
Quantitative analysis in MAs evaluation

• Ha Long, Vietnam 1314 August 2005
• Dr. Patrick Krauskopf, Vice-Director,
• Fanny Raess,
• Swiss Competition Commission

2
Structure of Presentation

• Introduction
• Role of quantitative techniques
• Frequently used quantitative analysis tools
• Price correlation analysis
• Own-price and Cross-price elasticity analysis
• SSNIP Test
• HHI
• Analysis of the modification of the competition
  structure
• 4. Conclusion

3
1. Introduction

• Objectives of the presentation
• Preliminary comment Description of the
  pre-merger situation to analyse the behavior of
  the competitors
• Perfect competition
• Monopolistic/Oligopolistic competition
• Contestable markets
• Dominance assessment (single vs collective)

4
2. Role of quantitative analysis

• Quantitative analyses can (and should) provide
  important evidence in an merger evaluation
• Quantitative evidence should be viewed as
  complementary to the qualitative evidence

5
3. Frequently-used quantitative analysis tools

• Using quantitative approach to
• Define the relevant market
• Evaluate the merger impact
• The choice of technique will depend on
• the specific circumstances of the case
• the availability of data
• It is possible that more than one technique could
  be usefully applied to any particular case

6
a) Price Correlation - definition

• Standard tool in evaluating product market
  definition
• Prices of products in the same market are
  expected to move together over time
• Basis of the approach
• Examine whether price levels move together
• Quantify the extent to which price levels move
  together over time. This is measured by the
  correlation coefficient
• The correlation coefficient can vary between 1
  and 1
• correlation of 1 gt prices moving perfectly
  together
• correlation of -1 gt prices moving perfecty
  inversely to one another
• correlation of 0 gt no statistical association
  between the two series

7
a) Price Correlation - Example

• Are still water, sparkling water and soft drinks
  in the same relevant product market?
• The price correlation analysis showed that
• Panel A Correlation between prices of water
  brands and soft drinks is weak
• Panel B Correlation between prices of still and
  sparkling water brands is high
• gt the market should include still and sparkling
  waters but should exclude soft drinks.

Illustration (source Lexecon) Panel A Low
correlation coefficient (lt 0.3) Panel B High
correlation coefficient (gt 0.9)
8
b) Own and cross-price elasticities (e) -
definitions

• Standard tools in evaluating product market
  definition
• The own-price elasticity measures the percentage
  change in quantity (Q) of good X for each
  percentage change in the price (P) of good X
• If e is high gt P increase is not likely to be
  profitable (ex Pepsi)
• If e is low gt P increase is likely to be
  profitable (ex Petrol)
• The cross-price elasticity measures the
  percentage change in quantity (Q) of good X for
  each percentage change in the price (P) of good Y

9
b) Own and cross-price elasticities example (1)

• Market Petrol
• Companies Shell, Agip, Esso, BP, Totalfina
• If Pshell increases gt Qshell decreases and
  Qotherbrands increases gt eShell is high
• If Pmarket increases gt Qmarket is only sligthly
  decreased gt e market is low

10
b) Own and cross-price elasticities example (2)

• Pepsi - Coca-Cola
• If Ppepsi increases gt consumers will switch
  to Coca-Cola gt Q Coca-cola will increase
• gt Pepsi and Coca-Cola are substitutes (e qp gt
  o)
• Wine - Water
• If P Wine increases gt consumer will not
  change their water consumption gt Q water will
  stay constant
• gt Wine and Water are not substitutes (e qp o)

11
c) SSNIP Test (small but significant
non-transitory increase in price) - definition

• Standard tool in evaluating product market
  definition
• It is a  hypothetical monopolist test 
• Tests whether a good or a set of goods define a
  relevant product market
• The question is  what type of constraints does
  the presence of other goods place on the producer
  of the good in question, when he decides to
  increase his price by 5-10? 

12
c) SSNIP Test - example

• 3 goods Apple (A), Pear (P) and Banana (B)
• Are they part of the same relevant product
  market?
• If all the producers of A can profitably
  introduce a small but significant and
  non-transitory increase in price, despite the
  existence of P and B gt the relevant market is
  made of A only
• However, if consumers, on reaction to this
  increase of price of A, significantly shift to P
  gt P are added to the relevant market, as the
  producers of A have no market power (unprofitable
  price increase)
• Same process with A and P together, until a
  profitable price increase is found

13
d) HHI - Herfindahl-Hirschman index defintion
(1)

• Measure of market concentration
• HHI Sum of the squares of the market shares
• The closer the market structure is to a monopoly,
  the higher the market concentration (and the
  lower the assumed degree of competition)
• In many juridictions, HHI is not considered as
  direct evidence, but as an indicator for further
  analysis

14
d) HHI as a screening device (EU)

• HHI lt 1000 gt Unlike to challenge
• 1000 lt HHI lt 1800 and ?HHI lt 100 gt Unlike to
  challenge
• 1000 lt HHI lt 1800 and ?HHI gt 100 gt Likely to
  challenge gt Further analysis needed
• HHI gt 1800 and ?HHI lt 50 gt Unlike to challenge
• HHI gt 1800 and lt 50 ?HHI lt 100 gt Likely to
  challenge gt Further analysis needed
• HHI gt 1800 and ?HHI gt 100 gt Likely to
  challenge

15
d) HHI - Example

• Pre-merger 5 firms
• A 30, B 25, C 25, D 10, E 10
• The HHI is 302 252 252 102 102 2350
• If B mergers with D, 4 firms
• A 30, B/D 35, C 25 , E 10
• The HHI is 352 302 252 102 2850
• ? HHI is 2850-2350 500
• Interpretation
• HHI gt 1800 and ?HHI gt 100 gt Likely to challenge

16
e) Analysis of the modification of the
competition structure

• Based on the market shares, is the merger likely
  to induce a structural change in the market
  structure?
• Examples
• Perfect ? Oligopolistic competition?
• Increased probability of collective dominance or
  cartellization?

17
4. Conclusion

• Empirical economic analysis lies at the heart of
  modern competition policy
• Econometric analyses does not come from out of
  the blue
• Potential weakness of the tools lack of data,
  asymmetry of data (firm/authority)
• Role of economists is complementary to the role
  of lawyers