Segnalazione Atos Origin

Document Sample
Segnalazione Atos Origin Powered By Docstoc
   A study for DG Comp of the European Commission by
 Lear – Laboratorio di economia, antitrust, regolamentazione

    Presentation at NMa conference – Den Hague 18 October
                         The study
The Competition Directorate-General of the EU Commission asked Lear:

1.   to develop “a detailed methodological approach for the ex-post
     review of European Commission decisions in the field of merger
     control, namely for assessing the impact of the Commission
     decision on market developments”; and

2.   to apply this approach to the EU decision on the merger between
     the power cable producers Pirelli and BICC (Commission Decision
     of 19.07.2000).

The study can be found at
     The aims of the ex-post assessment
 We believe that the ex-post assessment of merger decisions should
 have 2 aims:

 1. to establish whether the decision pursues the economic goal
    of the EU Merger Regulation better than any alternative
    decision within the set of those that the Commission could
    have legally taken; and
 2. to assess whether the analysis adopted to reach the decision
    was complete and correct, given the information available at
    the time.
Both aims are important because the assessment should help to
improve the decision making process by minimising the number of
inappropriate decisions.                                             3
The assessment with respect to the ultimate
               economic goal
 What is the ultimate economic goal of the
The economic and legal debate on whether a consumer welfare
standard or a total welfare standard should be adopted in competition
policy and merger control is still unsettled.

The EU Commission adopts a consumer welfare standard in the
enforcement of competition law, and clears mergers only if they are likely
not to negatively affect final consumers.

The assessment with respect to the ultimate
              economic goal
 Would this goal have been better pursued
had the EU Commission adopted a different
We propose to answer this question by:

•   identifying all the alternative decisions available to the EU
•   evaluating the effects of the decision on consumer welfare;
•   comparing these effects with the ones that would have taken place
    in each possible counterfactual.

        The effect on consumer welfare

To reach a conclusion on the appropriateness of the decision it is
necessary to compare the impact on consumers of the market
development that followed the actual decision with the one that would
have resulted from each of the possible alternatives.

The level of consumer welfare in a market depends on:
• the price at which goods and services are exchanged
• the volume of these transactions
• the quality and variety of these goods and services and consumers’

    The effect on consumer welfare - 2

To understand how consumer welfare has changed following an
EU decision it is necessary to analyse how these market
variables have evolved, since the decision has been acted upon,
or would have, when considering a counterfactual.

However, time consequentiality does not imply causality

Hence, it is important to understand if the change in these
variables was due to decision or to some exogenous factors.

There are empirical techniques that can be used to separate the
effects of the decisions from other phenomena.

        The appropriateness of a clearance
In the case of a clearance the exercise is easier as the market evolution
after the merger often provides enough information to understand
whether the increase or decrease in consumer welfare was the
consequence of a change in level of competition or of the efficiency of
the merging firms in the market.

In this case one can conclude that if after the merger:
• The market has become more competitive or the merging firms more
  efficient, the decision was appropriate
• The market has become less competitive, the decision was inappropriate

However, if one cannot exclude that the change was due to some
exogenous factors, it is necessary to adopt some empirical techniques to
assess the what would have occurred in the absence of the merger
                 Testing key hypotheses
Is the change in production costs endogenous?
Survey of market players to find out if cost-change affected whole market
Estimate and compare price-cost margins of the merging firms and their
competitors by means of a structural model
Compare change in costs of the merging firms to those of control group
Is the change in the degree of competition endogenous?
Survey the market players to find out about major structural changes
Event study on market players’ stock prices
Is the shift in the demand curve endogenous?
Compare post-merger prices and sales of merging parties and competitors
Build a structural model with pre-merger supply and post-merger demand
data to compare estimated output with actual one
Survey market players to find out about changes in demand patterns and
quality of products                                                  9
    The appropriateness of a prohibition - 1

The actual evolution of the market is much less informative when
dealing with a prohibition, because it does not help in forming a prima
facie hypothesis that can be tested.

The approach proposed to assess the appropriateness of a clearance
is based on a series of tests aimed at understanding whether there is a
causal relationship between 2 events:
•    the actual evolution of the market; and
•    the actual merger.

This method cannot be applied for a prohibition because the second
event has not happened and, therefore, there is no actual causal link
that can be tested.
    The appropriateness of a prohibition - 2
If the decision is a prohibition it is only possible to define a potential
causal relationship between the (potential) merger and its (potential)
effects on the market and test it.
This problem is actually identical to the problem a competition authority
faces when it tries to predict the effects of a future merger.
All the techniques that are used in the ex-ante assessment of mergers
could be employed in this case, but lack of data can be a problem.
An alternative option is to rely on the expectation that some informed
market players formed on how the market would have evolved had the
proposed merger happened, which can be solicited through a
questionnaire or can be inferred from their actual market behaviour.

  Techniques that can be used to assess a
 prohibition (when no remedies proposed)
   Survey       Can always be used      Relies on the expectations of
                                        well informed market players
Event studies   Can be used if firms     Relies on the expectation
                are quoted on stock    formed by stock market when
                      market              merger was announced
 Structural      Can be used if all    Simulates effects of the merger
  models         necessary data is     using ex-ante supply data and
                     available             ex-post demand data
 Evaluation      Can be used only      Derives conclusions from effect
  methods        when valid control    of a merger in a similar market
                market can be found
 Techniques that can be used to assess a
prohibition (when remedies were proposed)
 Survey        Can always be used,      Relies on the expectations of
             but the more hypothetic    well informed market players
              the scenario the more
              biased the responses
  Event      Can be used if firms are     Relies on the expectation
 studies     quoted on stock market     formed by stock market when
                                           remedies were proposed
Structural     Can be used if all          Simulates effects of the
 models        necessary data is         merger using ex-ante supply
                available and if        data and ex-post demand data
             remedies are structural
               and easy to model
                  Empirical techniques

The following empirical techniques can be used in the ex-post
•   structural models and simulations
•   evaluation methods
•   event studies
•   surveys

These techniques are not mutually exclusive and the best approach
consists in using more than one simultaneously.

        Structural models and simulations
The central idea behind this technique is to empirically estimate the
parameters of a set of structural equations that describe the market
under consideration.

Sometimes a structural model is built for the ex-ante assessment of a
merger, in these cases it is recommended to use it also ex-post.

This technique:
•   provides very detailed results
•   requires a large set of assumptions
•   needs a considerable amount of good quality data
•   presents a trade off between ease of applicability and precision
•   is not apt to consider decisions that involved behavioural remedies
                  Evaluation methods
This type of estimation techniques are all based on the comparison
between the behaviour of two groups: the control group and the
“treated” group.

The basic idea is that, ceteris paribus, the difference in performance is
an estimate of the effects of the “treatment”, i.e. the EU decision.

There is not much academic literature on the use of their use in merger
decision, even ex-ante
These techniques:
• require care in selecting the control group as the assignment to the
  treatment may not be random
• present a trade off between ease of applicability and precision
                       Event studies
This methodology consists in assessing the reaction of the stock
market to an “event”, in this case the EU decision.

Event studies rely on the assumption that the stock price of a firm
represents the discounted value of the expected flow of profit

Any change in the value of the stock price of the firms operating in the
affected market relative to the one that would have been observed
without the event provides information on its effects on the market
This technique:
• relies on very strong assumptions on how the stock market works
• require that all the affected agents are quoted firms
• is less reliable in case of multi-product firms
A surveys involves the collection of data from a representative sample
of the participants to the relevant market

It represents a good means to collect qualitative and quantitative data
about actual and potential developments in the affected market.

A survey is very useful in the case of a prohibition as it allows to infer
what would have happened had the merger been cleared

A survey:
•   allows to collect data that can be used with the other techniques
•   has to be carefully designed and implemented to avoid biases
•   relies on a good response rate, which can be difficult to achieve
•   requires to carefully address the confidentiality issue
         Ex-ante and ex-post evaluations
From the above it is clear that are strong similarities between the ex-ante
of a merger and ex-post evaluation of a merger decision:
•   the techniques that can be used are the same
•   the data required is often similar, but it refers to a different time period

However, in some cases, the ex-post assessment may involve the
collection of different data (e.g. new entry).

If the firms in the affected market are quoted it is possible to run event

A major difficulty in the ex-post data collection is that the market
participants are under no legal obligation to provide information and have
no interest in cooperating.                                             19
              Some final considerations
From the application of this methodology to a specific decision we
have learnt some useful lessons:
• the collection of the data is a difficult exercise as firms are under no
  obligations to cooperate, hence increasing awareness of usefulness of
  ex-post assessment can increase willingness to participate
• the ex-post evaluation should be performed by an independent body,
  but the support of the Commission in the data collection exercise is
• since some of the information required can be commercially sensitive,
  we suggest to allow 5 years to pass before the assessment and to
  guarantee maximum confidentiality in the storage and publication of
  the data
• be aware that it is difficult to obtain good-quality and complete data

Shared By: