CrossBorder Bank M

1
Cross-Border Bank MAs and Bank Risk Evidence
from Bond Yield Spreads

• Sungho Choi, Bill B. Francis, and Iftekhar
  Hasan,
• Rensselaer Polytechnic Institute, Bank of
  Finland

2
Cross-Border Bank MAs - Motivation

• Popular Business Strategy
• Charting new territory
• Different portfolio management, new financing
  sources.
• Numerous barriers, e.g., distance, currencies,
  language, culture, regulatory and supervisory
  norms
• In case of MAs which countrys regulation
  should prevail? Lack of concrete rules of
  regulation!

3
Cross-Border Bank MAs - Motivation

• Uncertainty associated with the new environment
  changes the risk perspective of banks and
  therefore impacts all stakeholders -
  shareholders, bondholders, and regulators.
• Empirical evidence on the impact of cross-border
  MAs on bank stakeholders is sparse.

4
Research Questions

• What is the impact of cross-border MAs on bond
  holders?
• In addressing this question we use the changes in
  bond yield (or change in bank risk) as the
  metric.
• Additionally, we also trace the impact on the
  other stakeholders - shareholders and regulators
• We also examine how these changes are affected by
  institutional factors that are known to impact
  both bond and share value (moral hazard, investor
  protection, recovery rate, regulation and
  supervision, creditor rights, and information
  cost).

5
Preview of Results

• There is a significant increase in yield spread
  on the announcement of cross-border MAs.
• Investor protection, moral hazard, recovery rate,
  creditor rights, and income levels between
  acquirers and targets countries are important
  determinants of the impact on the abnormal
  changes in yield spread.
• For the full sample, we do not find any evidence
  of wealth transfer between bondholders and
  shareholders. However when sort on positive and
  negative abnormal changes in yield spread (AESP)
  there is evidence of wealth transfer
• - Positive AESP wealth transfer from
  bondholders to shareholders
• - Negative AESP - wealth transfer from
  regulators to bondholders
• Results show that investor protection and moral
  hazard, are important in explaining abnormal
  changes in bond yield spreads.

6
Cross Border MA and Risk

• Mixed views on the Impact of MA on Bank Risk
• Risk Decreasing
• Reduce risk of bank insolvency - Vander
  Vennet(1996), Berger (2000).
• Risk Increasing
• Distance and differences in languages and
  cultures, regulatory and supervisory norms
  (Berger DeYoung, and Udell (2001)).
• Diversification into new sectors or regions which
  are likely to perform worse Winton (1999).

7
Empirical Evidence

• Amihud, Delong, and Saunders. (2002)
  Cross-border MA announcements result in negative
  and significant CAR to share holders of bidder
  banks. These mergers do not lead acquirers to
  engage in post-merger risk shifting or risk
  increasing behavior.
• Penas and Unal (2004) Domestic mergers are
  risk reducing, evidence through bond return.
• Our paper provides much needed empirical evidence
  on this important issue.

8
Data

• Thomson Financial SDC Platinum, 1995 to 2002.
• Use only mergers where the acquirers stock is
  publicly traded and bond yield spreads are
  available through Datastream.
• Acquiring bank needs to have at least one bond
  outstanding with a remaining maturity of greater
  than 2 years.
• 147 cross-border bank MAs.

9
Data - Continued

• Fitch-IBCA Bankscope database.
• Individual bond data and the government bond data
  of acquiring banks country are from Datastream.
• Country specific, regulatory and supervisory
  variables are obtained from Barth, Caprio, Levine
  (2001) and the World Bank database (2004).

10
Moral Hazard

• Having an extensive deposit insurance may lead to
  greater moral hazard for bank managers
  (Bhattacharya and Thakor (1993), Bhattacharya,
  Boot, and Thakor (1998), and Demirguc-Kunt and
  Detragiache (2002).
• A generous deposit insurance of target countries
  against acquirer countries may increase the risk
  of bank failure, resulting in a higher yield
  spread.

11
Investor Protection

• Extent of legal protection of investor is a key
  determinant of asset valuation (La Porta et al.
  (2002))
• -- strong legal protection of investors is
  associated with higher valuation of corporate
  assets.
• -- support for the existence of expropriation
  of minority shareholders as well as the role of
  the law in limiting such expropriation.
• The higher investor rights could be interpreted
  as less risk to investor.
• -- assuming all things are equal, if a bank
  takes over a financial institution in a country
  where there is a higher investor legal
  protection, the change in the acquiring banks
  risk should at least be non-positive.

12
Recovery Rate

• The recovery rate is defined as how many cents on
  the dollar claimants recover from an insolvent
  firm (Djankov et al. (2005)).
• The credit spread of risky bonds and loans depend
  inversely on the recovery rates on the bonds
  under consideration (Acharya and Bharath (2004)).
  
• Higher recovery rates imply a high probability of
  a bank recouping its loans, thus mitigating the
  risk to lenders.

13
Regulation and Supervision

• Regulation and supervision significantly
  influence cross-border bank MAs (Berger,
  DeYoung, Genay, Udell (2000), Focarelli and
  Pozzolo (2001), Buch and Delong (2004), Jayaratne
  and Strahan (1998)).
• -- governmental regulations and supervision
  may reduce information symmetries and are often
  essential to ensure the solvency of whole banking
  system.
• -- higher and tougher regulations and
  supervisions in acquiring banks countries
  relative to target countries increase the chance
  of better performance and thus reduce the risk of
  bank failure, resulting in lower yield spreads.

14
Methodology

• Yield spreads - the difference between the yield
  on a bank bond and a government security of that
  country with comparable maturity (Gande, Puri and
  Saunders (1999)).
• Based on weekly yield data.
• Combine all firms bond yields into a single
  yield by value weighted averages of individual
  bonds.
• Methodology to measure the abnormal announcement
  effect on bond yield spreads is adopted and
  modified from Eckbo Maksimovic and Williams
  (1990).

15
Methodology Continued

• Abnormal effects on yield spreads are estimated
  directly as the parameter ßj in the model
• SPjt aj
  ßjdjt ejt (1)
• - where - SPjt is the market value weighted
  average yield spread of a bank js bond djt is a
  dummy variable which takes on a value of 1 if
  week t is the week of the announcement.
• Estimation period typically (-30, -4) weeks
  event window (-1,1) weeks.
• Abnormal effects are averaged with equal weights
  across banks to form the average abnormal
  effects.

16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
Regression of Announcement Week Abnormal Effects
on Yield Spreads

• Abnormal effects on yield spreads are estimated
  directly as the parameter ßj in the model
  
• AESPj aj ß1Investori ß2Recoveryi
  ß3Hazardi ß4Toughi
• ß5Transi ß6 Concentration S ßj Zj
  ej
• -Where AESPj is the announcement abnormal
  effect on yield spread and is estimated as the
  parameter ßj in equation 1. We use the natural
  logarithm of one plus parameter ßj since it is
  characterized by high kurtosis.
• We estimate AESP based on several announcement
  windows as well as estimation windows of (-30,
  -4) and (-52, -4) weeks.

25
Independent Variables

• Investor Protection 1 if Target has Better
  Investor Protection than Acquirer otherwise0
• Recovery 1 if Target has Better Recovery Rate
  than Acquirer otherwise0
• Hazard Index World Bank
• Tough Index on Supervision World Bank
• TransRelative transparency (target
  index/acquirer index) World bank
• Concent1 if target has a higher banking
  concentration
• Zothers including Creditor Rights
  (Target/acquire index), Law (target acquirer
  same legal origin1) Income1 if same income
  level DFX (target acquirer same currency1
  D100CS1 if all payments is cash transaction
  LogTAcqLogarithm of Total Assets of the
  Acquiring Bank ROEacqROE of the acquiring bank,
  TETAcqequity to assets past year of acquiring
  bank LLRacqLoan Loss Reserve prior year to Total
  Loan of Acquiring Bank

26
Regression Results

• Investor Protection Negative to in all 10
  models
• Hazard Positive to in all models
• Tough, Creditor Rights, Concentration
  Positive, Marginal Significance in Some of the
  Equations
• Recovery, Trans, DFX, Law, Income, Size, ROE,
  Equity/Asset Not Significant
• D100CS Positive to in all models
• LLR Positive in Most Equations

27
Summary of Results

• There is a significant increase in yield spread
  on the announcement of cross-border MAs.
• Investor protection, moral hazard, recovery rate,
  creditor rights, and income levels between
  acquirers and targets countries are important
  determinants of the impact on the abnormal
  changes in yield spread.
• For the full sample, we do not find any evidence
  of wealth transfer between bondholders and
  shareholders. However when sort on positive and
  negative abnormal changes in yield spread (AESP)
  there is evidence of wealth transfer
• - Positive AESP wealth transfer from
  bondholders to shareholders
• - Negative AESP - wealth transfer from
  regulators to bondholders
• Results show that investor protection (-) and
  moral hazard () are important in explaining
  abnormal changes in bond yield spreads.

28
Implications

• Since cross-border bank MAs increase the risk of
  acquiring banks, the home country regulator(s)
  may require acquiring banks to increase their
  reserves to better protect the banks
  stakeholders - depositors, bond holders, and
  shareholders.
• Regulators should especially consider
  institutional characteristics in both the home-
  and the host-country in judging the sufficiency
  of the banks reserve positions.