It expands the literature on cross-border acquisitions b

1
Cross-border Bank Acquisitions Is there a
Performance Effect?

• By Ricardo Correa
• Discussant Elijah Brewer III, DePaul University
  and the Federal Reserve Bank of Chicago

2
Outline

• Why I like the paper?
• Summary of setup
• Big picture questions
• Results
• Suggestions/clarifications probit specification
• Suggestions/clarifications difference-in-differe
  nce tests
• Assessment

3
Why I like this paper

• The paper combines two strands of the banking
  literature
• It expands the literature on cross-border
  acquisitions by analyzing both the determinants
  of financial foreign direct investment at the
  country level and the target specific
  characteristics that motivate cross-border
  acquisitions.
• It extends the operating performance analysis
  used to analyze performance effects post domestic
  MA acquisitions to a sample of deals that
  include targets in developed and emerging
  economies.

4
Summary of setups

• Probit specification to estimate the factors that
  distinguish banking organizations that have been
  targets of cross-border bids from those that have
  not using data from mid-1990s to the early 2000s.
  The dependent variable is equal to one, the year
  a firm is the target of a bid from a foreign
  financial institution, zero otherwise.
• Univariate tests are conducted for several
  measures of performance
• Comparing performance two-years before relative
  to two-years after merger.
• An equation to estimate the effects of economic
  integration and information costs on a targets
  performance post-cross-border acquisition

5
BIG questions

• Which factors influence cross-border
  acquisitions?
• Do cross-border acquisitions improve the targets
  performance?
• What factors can explain the change, if any, in
  performance?

6
Results

• First question
• The discrete choice estimates show that banking
  organizations are more likely to get acquired in
  cross-border deals if they are relative large,
  poor performers, located in a small country, and
  when the banking sector is more concentrated.
• Second question
• Post-acquisition performance for target banking
  organizations does not improve in the first two
  years relative to domestically-owned financial
  institutions. In fact, the multivariate
  regression shows that performance declines over
  the first several years after the acquisition by
  a foreign entity.

7
Suggestions, Clarificationsprobit specification
(1)

• Average performance measures over a longer
  horizon to smooth out short-run fluctuations in
  these measures that may not have anything to do
  with the long run performance of the firms.
• A risk measure
• Variability of ROA or ROE
• RWA/TA
• Cross-border deals might be motivated by the
  desire to gain access to a banking
  organizations retail operations. Why not control
  for this possibility by including the ratio of
  retail deposits to total deposits in the
  specification?
• Number of foreign banking organizations or
  banking organizations that are owned by foreign
  banks in the banks market area. Are you more
  likely to be a target if there are other banking
  organizations in your market that are foreign
  owned?
• Is a banking organization more likely to be a
  target if it has a foreign operation?

8
Suggestions, Clarificationsprobit specification
(2)

• Are firms in a country more likely to be targets
  of cross-border deals if the currency of the
  country is relatively weak? A country-fixed
  effect?
• Did you review the history of non-acquirers to
  identify firms whose objective function may
  include elements other than maximizing the value
  of the banking organizations to their
  shareholders? In particular, did you exclude
  organizations in any year in which they are
  government owned or owned as a part of a
  cooperative?

9
Suggestions, Clarifications probit specification
(3)

• To capture the importance of non-interest income,
  It might be useful to make use of
• The ratio of non-interest income to net interest
  income plus non-interest income
• In addition, recent studies seem to suggest that
  banking organizations have been shifting within
  non-interest income between traditional and
  non-traditional sources of non-interest revenue.
  So, why not use the following decomposition of
  non-interest income
• Traditional fee-based revenue
• is income from fiduciary activities plus service
  charges on deposit accounts.
• Nontraditional sources of non-interest revenue
• all other non-interest revenue (e.g., income from
  underwriting corporate debt and equity
  securities, selling and underwriting various
  forms of insurance, etc).

10
Suggestions, Clarificationsprobit specification
(4)

• What does the concentration measure actually tell
  us?
• In four countries (France, Germany, Spain, and
  the United States) that account for 20 percent
  of the deals and 37 percent of the observations
  the cross-country average three-bank
  concentration ratio is likely to be very low and
  not quite representative of the actual
  concentration in the countries. According to one
  study using data for the period 1989 to 1996 the
  cross-country average three-bank concentration
  ratio was 0.26. For countries like the United
  States the three-bank concentration ratio does
  not reflect the extent of competition in a
  banking market in which a particular firm
  operates. The importance of this is likely to
  vary with size of country.
• What does the Heritage Freedom measure and how is
  it expected to influence the probability that a
  firm will be target of a bid from a foreign
  entity? You need to discuss this and point out
  the limitation of this index.

11
Suggestions, Clarificationsprobit specification
(5)

• Room for greater interpretation of the
  coefficients. For example
• The Market CAP/GDP variable is included as a
  measure of the importance and market power of the
  banking system in the economy.
• Correa claims that a more developed equity market
  competes with the banking sector in the
  allocation of resources, reducing market power
  and makes entry less attractive for MNBs. An
  alternative hypothesis is that more competition
  primarily from capital markets may make for
  greater efficiencies and higher profits, making
  entry more attractive. A discussion a long these
  lines would fill out the paper nicely.

12
Suggestions, Clarificationsprobit specification
(6)

• Econometric issues
• Do non-acquirers appear multiple times? If so,
  do you have panel data problems that Petersen
  (2007) discusses in Estimating Standard Errors
  in Finance Panel Data Sets Comparing
  Approaches.
• Are the factors influencing cross-border mergers
  constant over time? The number of deals seem to
  peak in 2000, declining thereafter.

13
Suggestions, Clarifications difference-in-differen
ce tests (1)

• Control sample issues
• Does the control sample include banking
  organizations that are undergoing some domestic
  restructuring that could make it a less than
  perfect control for overall changes in the
  banking activity at the country level?
• If there is a problem with the control sample,
  this could not only influence the
  difference-in-difference tests but it could
  influence the performance and economic
  integration tests that are performed latter on in
  the paper.
• I suggest that Correa try several other
  country-level indices
• One for those firms within the country that are
  involved in domestic mergers.
• Another for those firms within the country that
  are not involved in any type of merger
  activities.

14
Suggestions, Clarifications difference-in-differen
ce tests (2)

• Compares performance of the standalone entity
  over the two year period before the merger with
  the entity that is a subsidiary of the acquirer
  over the two year period after the merger. Are we
  comparing apples and oranges here?

15
Assessment

• Correa have considered two BIG picture questions
• Which factors influence cross-border
  acquisitions?
• Do cross-border acquisitions improve the targets
  performance?
• Uses deals in both emerging markets and developed
  countries. Provides a nice addition to the
  literature.
• Spending a little more time on the RHS variables
  in the probit regression seems worthwhile.
• Developing appropriate benchmarks to compare
  banking organizations performance before and
  after cross-border acquisitions.
• Correa has already found some very interesting
  correlations in the data and I have no doubt
  there is much to learn from considering these two
  BIG picture questions.