THE PERENNIAL CHALLENGE
TO COUNTER TOO-BIG-TO-FAIL IN BANKING:
EMPIRICAL EVIDENCE FROM THE NEW INTERNATIONAL REGULATION
DEALING WITH GLOBAL SYSTEMICALLY IMPORTANT BANKS
Forthcoming, Journal of Banking and Finance
Sebastian C. Moenninghoff, WHU
Steven Ongena, University of Zurich, SFI, Bangor University & CEPR
Axel Wieandt, WHU
The research question ...
... that drives us all ...
How to design regulation
to reduce the costs and risks
of Too-Big-to-Fail (TBTF)?
What we do (in this paper)
• Provide evidence on how the new international regulation
on Global Systemically Important Banks (G-SIBs)
impacts the market value of large banks
• Analyze the stock price reactions for the 300 largest
banks from 52 countries across 12 relevant regulatory
announcement and designation events
What we find
• The new regulation negatively affects the value of the
newly regulated banks
• But the official designation of banks as G-SIB itself has a
partly offsetting positive impact
• Likely due to a TBTF perception by investors
• Demonstrates the inherently paradoxical nature of the
new regulation
«Paradoxial nature» is reflected in ambivalent market
participants` statements.
Josef Ackermann, CEO Deutsche Bank
“Despite the additional regulatory
burden associated with G-SIB status,
glad that Deutsche Bank would likely
be on the list of G-SIBs as this would
allow for benefits in the refinancing
and depository business”
(Handelsblatt, June 27, 2011)
“The bank’s G-SIB status would reflect
‘the comparative advantages […] of the
Bank of China versus our domestic
peers’” (Financial Times, November 4,
2011)
“The opportunity outweighs
the challenge”(People’s Daily
Online, November 6, 2011)
LITERATURE REVIEW
en bref
Official government announcements or
the establishment of deposit insurance schemes
strengthen TBTF
• US Comptroller of the Currency’s announcement of a TBTF policy (after
the bailout of large US banks in the 1980s) leads to positive shareholder
wealth effects for large money center banks in O’Hara & Shaw (JF 1990)
• Insured deposit financing shielded US bank holdings from market
discipline in early 1990s as reflected in long-term credit rating actions in
Billett, Garfinkel & O’Neal (JFE 1998)
• Introduction of a deposit insurance in Russia increased banks’ risk taking
in Chernykh & Cole (JBF 2011)
• Japanese deposit insurance reform strengthens TBTF in Imai (JBF 2006)
IM
Rescue measures can contribute
to the emergence of TBTF
• Federal Reserve Bank’s coordination of the rescue of Long
Term Capital Management (LTCM): impact on
• bank stocks in Kabir & Hassan (JBF 2005)
• subordinated debt yield spreads in Balasubramnian & Cyree (JBF
2011): especially large banks, see also Sironi (JMCB 2003) for Europe
IM
Implicit government guarantees
• Seemingly exist, findings based on:
• long-term bank stock returns in Ghandi & Lustig (JF 2015): esp. large
banks
• credit ratings (gives funding advantage) in Ueda & Weder di Mauro
(JBF 2013)
• But not based on: uninsured CDs in Ellis & Flannery (JME
1992)
• Reflected in balance sheet structures in Nier & Baumann
(JFI 2006)
• Increases risk taking by banks in Dam & Koetter (RFS 2012)
Banks` reacting
• To obtain TBTF banks merge, as reflected in
• increasing returns for bonds of midsize banks in Penas & Unal (JFE
2004)
• significant bond yield spread tightening for midsize banks in Deng,
Elyasiani & Mao (JBF 2007)
• bondholders of banks conducting domestic mergers benefit in Ongena
& Penas (JFS 2009)
• competitive conduct by rivals in Gropp, Hakenes & Schnabel (RFS
2011)
• May undercut ability of regulator to limit TBTF?
• Maybe not: in Flannery & Sorescu (JF 1996) and in Schafer, Schnabel
& Weder di Mauro (RoF 2015)
G-SIB REGULATION
Time line
• 2009.04: G20 establishes Financial Stability Board (FSB) in
order to develop guidelines for additional regulation and
oversight of G-SIBs
• 2009:09: G20 tasks the FSB to develop concrete measures by
2012:10, w/ the Basel Committee on Banking Supervision
(BCBS)
• Supervision:
• more supervisory powers, an improved set of standards and methods, and
stricter assessment regimes, as well
independence, and appropriate resources
as
unambiguous
mandates,
• Regulation
• G-SIBs hold additional capital of 1.0‒2.5% Common Equity Tier 1,
depending on the individual institution’s systemic relevance as measured by
an indicator-based approach, in order to increase the loss-absorbency of GSIBs
• Resolution
• strengthening national resolution regimes, cross-border cooperation, and
resolution planning, and the enhanced resolvability of G-SIBs
Appendix A
Costs and benefits
• Net wealth effect of any new regulation ultimately depends on
the resulting equilibrium
• The additional regulatory measures for G-SIBs are widely
considered to involve costs for the affected institutions
• Capital surcharges imply higher average funding costs as bank equity
is perceived to be more costly than debt from the banks’ perspective
• Tax deductibility, information asymmetry
• Additional supervisory requirements
• The establishment of resolution regimes is also assumed to result in
significant administrative, legal, and operational costs
• Credible resolvability would lower implicit government support, which in turn
would increase funding costs for G-SIBs
• Value of designation as G-SIB and an explicit government
guarantee
WHAT WE DO
300 largest publicly listed banks
• Datastream -- asset size as of the end of 2009 -- financial
institutions” subcategory
• 265 commercial banks, 34 investment banks, 1 bancassurance
• 17 countries (5 continents)
Table 2
Table 3
Standard event study methodology
Rmt = national market index
Discussion
1.
Stock market closures and timing of news
• News occurring later on a business day in Europe or the US (e.g., afternoon
GMT) affects Asian stocks only on the next calendar day because the local stock
exchanges would already be closed
• Adjust the non-weekend event returns for time zones using the local stock
exchange closing times as provided by MSCI and event-specific estimates about
the time the information was released
2.
Use of (broad) local market indices as in Campbell, Cowan, and
Salotti (JBF 2010)
• Non-parametric tests
• Large individual banks may impact local index
• This effect seems limited for our sample
3.
Estimation window -150 to -1
4.
Event window: 1 day
• Precise news time stamp
• Crisis generates lots of noise
Robustness
(we want to mention immediately)
1. Calculate returns from the respective calendar days
instead of time zone-adjusted event returns
2. Include the MSCI World as a global market index
3. Use a pre-crisis “fixed” estimation window that consists
of the first 150 days in 2008 instead of a rolling
estimation window preceding the event dates
4. Employ a five day (+2/-2) event window instead of a one
day event window
WHAT WE FIND
Stockprices
Figure 2
Figure 1
Table 4
Table 4
Cross-sectional exercises
• the value of obtaining G-SIB status is lower for G-SIBs
with a higher government ownership stake
• we do not observe a clear relationship between home
country rating and value of G-SIB designation
• the degree of systemic relevance as expressed by the
required level of capital surcharge appears to have a
dampening effect on the returns on designation
Conclusion
• We observe negative absolute as well relative CARs for G-
SIBs compared to Non-G-SIBs and Other Banks upon
announcements related to the new regulatory measures
• This confirms the hypothesis that the new regulation for G-SIBs
involves costs for the affected banks.
• However, the official designation of particular banks as G-SIBs.
which leads to positive aggregate absolute and relative CARs
for G-SIBs across all three designation events
• Suggests that official designation as a G-SIB creates value for the
affected banks’ shareholders.
• The level of ownership of a bank by domestic government determines
the value of G-SIB designation, and so does the systemic relevance
Policy
• To the extent that the observed future costs from the new
regulation represent a reduction in implicit government
guarantees, our results confirm the effectiveness of the
announced reform proposals to limit TBTF
• Revealing the identities of G-SIBs eliminated ambiguity
about the presence of government guarantees, and
thereby may have run counter to the regulators’ intent to
contain the effects of TBTF
• Highlights the need for a conscious decision about the level of
supervisory transparency versus constructive ambiguity