Description
The purpose of this paper is to discuss and then analyze the interdependency between
bank and sovereign risk before, during and after the financial crisis
Journal of Financial Economic Policy
The eurozone financial crisis: role of interdependencies between bank and sovereign
risk
J ames R. Barth Apanard (Penny) Prabha Greg Yun
Article information:
To cite this document:
J ames R. Barth Apanard (Penny) Prabha Greg Yun, (2012),"The eurozone financial crisis: role of
interdependencies between bank and sovereign risk", J ournal of Financial Economic Policy, Vol. 4 Iss 1 pp.
76 - 97
Permanent link to this document:http://dx.doi.org/10.1108/17576381211210203
Downloaded on: 24 January 2016, At: 21:43 (PT)
References: this document contains references to 11 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1011 times since 2012*
Users who downloaded this article also downloaded:
Wenling Lu, David A. Whidbee, (2013),"Bank structure and failure during the financial crisis", J ournal of
Financial Economic Policy, Vol. 5 Iss 3 pp. 281-299http://dx.doi.org/10.1108/J FEP-02-2013-0006
Richard J . Buttimer, (2011),"The financial crisis: imperfect markets and imperfect regulation", J ournal of
Financial Economic Policy, Vol. 3 Iss 1 pp. 12-32http://dx.doi.org/10.1108/17576381111116795
Paul Bennett, (2011),"The (revised) future of financial markets", J ournal of Financial Economic Policy, Vol. 3
Iss 2 pp. 109-122http://dx.doi.org/10.1108/17576381111133598
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The eurozone ?nancial crisis:
role of interdependencies
between bank and sovereign risk
James R. Barth
Auburn University, Auburn, Alabama, USA and
Milken Institute, Santa Monica, California, USA
Apanard (Penny) Prabha
Milken Institute, Santa Monica, California, USA, and
Greg Yun
Santa Monica, California, USA
Abstract
Purpose – The purpose of this paper is to discuss and then analyze the interdependency between
bank and sovereign risk before, during and after the ?nancial crisis.
Design/methodology/approach – The authors’ approach is based upon an examination of 44 large
banks headquartered in 13 countries; eight of these countries belong to the European Union, seven
belong to the eurozone, and the remaining ?ve belong to neither group. This provides a good
comparison group of countries.
Findings – Evidence is found supporting the existence of signi?cant bank andsovereign risklinkages.
There are, however, different patterns in the relationships across countries and even across banks
within the same country. Also, higher correlations between bank and sovereign risk are found in
countries in which the ratio of the assets of banks relative to their home country’s GDPis relatively high.
Research limitations/implications – Based upon the empirical results, allowing banks to invest
in sovereign debt without requiring them to hold any capital against the “true” risk of such debt
increases the likelihood of insolvency. This means that interdependencies between bank and sovereign
risk are extremely important when setting regulatory capital requirements and considering whether
action is needed to limit any increase in the likelihood of contagion.
Originality/value – The paper provides a new examination of the interdependencies between
individual bank risk and the sovereign risk of the countries in which they are headquartered, with
special emphasis on the recent global ?nancial and eurozone crises.
Keywords Banks, Capital, Fixed assets, Sovereign debt crisis, Credit default swap spreads,
Capital requirements, Risk-weighted assets, Bailout, Asset diversi?cation
Paper type Research paper
I. Introduction
A collapse in the ?nancial sectors of several important developed countries gave rise to
the global ?nancial crisis of 2007-2009. This situation, in turn, led to a severe downturn
in economic activity in these countries that spilled over to other countries around the
world. As governments in the developed countries attempted to restore economic
growth through bailouts of banks and other simulative ?scal actions, national de?cits
began rising to alarming levels. This exacerbated problems as investors in the sovereign
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – G01, G0, G, G21, G2, G28
JFEP
4,1
76
Journal of Financial Economic Policy
Vol. 4 No. 1, 2012
pp. 76-97
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211210203
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
debt issued by these countries increasingly demanded higher interest rates as the
perceived risk of defaults rose. These troubling events shifted much of the focus fromthe
riskiness of banks to the riskiness of sovereign debt. As a result, these two
interdependent factors, heightened bank and sovereign debt risk, contributed to serious
concerns about the stability of the global ?nancial system. The problemwas particularly
acute in the eurozone countries, where concerns were that the banking and sovereign
risk problems could lead to devastating contagion throughout the euro area as well as
spread to other countries through ?nancial and trade channels.
Understanding the interdependencies between bank and sovereign risk is important,
particularly when banking and debt crises are intertwined. Government debt has
historically played an important role in the banking system. Banks hold government
debt enabling them not only to earn interest income but also to hold less regulatory
capital against that debt than is the case for other assets. The reason for the lower capital
requirement is that government securities have typically been considered to be risk free,
which means such assets can be converted to cash at low-transaction costs and at face
value. These properties of government debt securities have made them widely accepted
as collateral in ?nancial systems everywhere. Most importantly, since government debt
securities are considered to bear no risk, regulators have generally not required banks to
hold any capital against such securities. This allows banks to operate with lower
capital-to-asset ratios, which contributes to higher returns on owner-contributed equity
capital.
The recent global ?nancial crisis clearly demonstrates that the role of governments in
domestic banking systems goes well beyond simply being a provider of a supposedly
risk-free security. Indeed, many governments around the globe were forced to rescue big
troubled banks that were deemed to be too systemically important to be allowed to fail.
As nations issued more bonds to ?nance bank bailouts and to combat recessions,
sovereign risk rose along with escalating debt-to-GDP ratios. In some countries,
moreover, such debt became riskier than even the risk of the banks holding it. More
speci?cally, the bailout of banks led to a shift of credit risk from the ?nancial sector to
national governments (Acharya et al., 2010). This situation was further aggravated due
to the economic recessions created by the ?nancial crises during which governments
issued even more debt to fund government spending to stimulate economic activity. The
inter-connectivity between bank and sovereign risk does not stop here. A recent report
by the Bank for International Settlements (Committee on the Global Financial System
(CGFS-BIS), 2011) describes various transmission channels through which increased
sovereign risk can spill over to the banking sector by increasing the funding costs of
banks holding sovereign debt. This means there is a potential vicious cycle linking bank
risk and sovereign risk in a potential downward spiral.
The purpose of this paper is to discuss and thenanalyze the interdependencybetween
bank and sovereign risk by examining 44 banks in 13 countries before, during and after
the ?nancial crisis. Brie?y, we ?nd evidence supporting the existence of signi?cant bank
and sovereign risk linkages. There are, however, different patterns in the relationships
across countries and even across banks within the same country. There are, for example,
banks with relatively little exposure to the government debt of the peripheral eurozone
countries (i.e. Greece, Ireland, Italy, Portugal and Spain – the so-called PIIGS) and with
more globally diversi?ed assets. In these cases, market indicators indicate such banks
are less risky and, therefore, not likely to be bailed out by their governments.
The eurozone
?nancial crisis
77
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Additionally, stronger relationships between bank and sovereign risk are found in
countries in which the ratio of the assets of banks relative to their home country’s GDPis
relatively high. This is consistent with the view that market participants expect
countries in such situations to be more likely to bail out banks in time of distress, thereby
increasing sovereign risk. We ?nd, however, that these patterns are quite different
between the non-crisis and crisis periods.
The remainder of the paper proceeds as follows. In Section II, we discuss the
relationship between sovereign risk and bank capital requirements. In Section III,
the interdependencies between bank and sovereign risk are discussed and analyzed. The
relationship between bank size, foreign diversi?cation and sovereign debt exposure are
examined in Section IV. The conclusion is provided in Section V.
II. Sovereign risk and bank capital requirements
Recent bank regulatory reforms both at the national level such as the US Dodd-Frank
Act and the international level such as Basel III impose stricter capital requirements on
banks. Under Basel I and II, the sovereign debt of a fairly large group of countries was
assigned a zero risk weight in terms of satisfying the regulatory capital requirement.
More speci?cally, Basel I considered the sovereign debt of all OECD countries to be
risk-free. Thus, banks in these countries were permitted to acquire such debt without
increasing their regulatory capital. Basel II, initially published in 2004 and revised
during 2005-2006, considered sovereign debt rated AA- or higher to be risk-free and,
therefore, subject to a zero risk weight. Basel III, initiated in 2010, assigns a
risk-weighting scheme for sovereign debt similar to that under Basel II.
More generally, Basel III recommends that banks be required to maintain a Tier 1
capital ratio of 6 per cent of risk-weighted assets, which is an increase from the previous
ratio of 4 per cent. There is also a liquidity coverage ratio, in which banks are required to
have high-quality liquid assets suf?cient to cover more than 100 per cent of a bank’s
30-day net cash out?ow. Domestic government debt and quali?ed foreign government
debt are considered to be as liquid as cash and, therefore, can be used to meet the
liquidity coverage requirement. Figure 1 shows Basel III’s sovereign creditworthiness
risk-weights and sovereign credit ratings under different risk-weight categories for the
13 countries in our sample. As the ?gure shows, some of the developed countries
(i.e. Greece, Ireland, Italy and Portugal) have lost their risk-free status for regulatory
capital purposes since the start of the European debt crisis in 2010.
Risk-weighting schemes for assets that heavily rely on rating agencies may work
well during normal times but can lead to a misallocation of a bank’s asset portfolio
during troubled times. When banks are not required to hold more capital for government
debt because it is considered risk-free by credit rating agencies or by national regulators,
they are more likely to acquire or continue holding the debt of poorly managed nations
and thus extend less credit to individuals and businesses. In many cases, moreover,
domestic sovereign debt may receive preferential treatment due to “national discretion”
in which such debt is assigned a lower weight in the risk-weighted asset scheme for
regulatory capital purposes[1]. This preferential treatment of national government debt
may be politically expedient, but is not consistent with prudential regulatory practices
whensuchdebt is not trulyrisk free. Of course, sucha practice enables a bank to leverage
to a greater degree and thereby increase its return on equity. It also enables a
government to borrow more cheaply due to the greater demand provided by banks.
JFEP
4,1
78
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In both cases, however, the ultimate cost may be an increase in the riskiness of a
country’s entire ?nancial system and thereby greater likelihood of the potential spread
of problems to other countries.
Risk-weighting schemes for banks also treat all risk-free sovereign debt (i.e. that with
a rating of AA- andhigher) as having the same risk. However, market participants do not
perceive the riskiness of all such debt to be the same, which is evidenced from ?nancial
instruments such as sovereign credit default swaps (CDSs) and spreads in interest rates
on government bonds for different countries. As shown in Figures 2 and 3, there are
Figure 1.
Timeline of Standard
& Poor’s sovereign credit
rating downgrades
RISK FREE ZONE
Australia, France, Germany, UK, Switzerland
USA
A+ to A-
20%
BBB+ to BBB-
50%
BB+ to B-
100%
Below B-
150%
2005 2010 2011 2007 2009 2008 2006
RISK WEIGHTS
Note: Domestic sovereign debt in domestic currency has 100% liquidity
coverage ratio despite its risk rating.
AA-
AA
AA+
AAA
USA
Spain Ireland
Spain Ireland
Ireland
Greece
Greece
Greece
Japan
Japan
Portugal
China
China
China
S&P downgrades Greece,
Portugal and Spain (Apr 10).
Greek fiscal problems are revealed by PM
speech, a month later the deficit number is
revised and doubled from previous
estimation (Nov 09).
Spain and Portugal accept
the austerity package
(May and June10)
Portugal
receives 78
billion bail-
out (May 11)
Italy
Italy
Japan
Portugal
Ireland Spain
Notes: Only the 13 sample countries are included (see Appendix Table AI for the list); countries
that received risk-free status, or a 0 per cent risk weight under Basel II and III, are shown in the
upper part of the figure
Source: Bloomberg; Milken Institute
Ireland
Ireland Portugal
Greece
Eurozone agrees to 30
billion Greece rescue
package (Apr 10)
Greece accepts 110
billion EU-IMF package
(May 10)
Irish banks share prices rebound
from the 20-yrs largest drop
after the Irish Govt. guarantees
bank deposits (Sep 08).
Irish Govt. injects 5.5 billion
in banks (Dec 08).
Irish accepts 68 billion
bail-out (Nov 10)
Figure 2.
Sovereign CDS spreads
and credit ratings
for countries
0
20
40
60
80
100
120
140
160
180
2007 2008 2009 2010 2011 2007 2008 2009 2010 2011
France (AAA)
UK (AAA)
Australia (AAA)
US (AAA)*
Germany (AAA)
Five-year senior sovereign CDS spreads (bp)
0
500
1,000
1,500
2,000
2,500
Greece
Ireland
Portugal
Italy
Spain
AAA
AA+
AA
AA–
A+
A
A–
BBB+
BBB
BBB–
BB+
BB
BB–
B+
B
B–
CCC+
CCC
CCC–
Five-year senior sovereign CDS spreads (bp) S&P sovereign credit ratings
Note: *The rating for US sovereign debt was downgraded from AAA to AA + on
August 5, 2011
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
79
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
signi?cant differences in both CDS spreads and interest rate spreads across countries,
which indicate that market participants to not perceive all sovereign debt to be equally
risky. Notice that the CDS spreads for the sovereign debt of France and Germany are not
the same, even though the debt of both is rated AAA. Furthermore, the CDS spreads for
the debt of both countries are higher than the spread for US debt, even after S&P’s
downgrade of US sovereign debt to AA þ (Figure 3). In general, credit ratings are
slower to adjust and adjust less to a ?nancial crisis as compared to CDS spreads (as
shown in the right panel in Figure 2), and often fail to effectively predict a government’s
risk of defaulting on its debt (Wall Street Journal, 12 August 2011).
The important point based on this data is that capital requirements based on existing
risk-weighting schemes may understate the risk of banks in some countries, particularly
during periods of heightening sovereign risk. The direct and immediate effect of the US
sovereign debt rating downgrade on the required capital to be held by banks was
non-existent since the downgrade from a AAA to AA þ rating did not affect the risk
weighting of assets. However, European banks, many of which had signi?cant exposure
to Greek, Italian, Irish and Portuguese government debt in late 2011 faced a sizeable hit
to capital. According to the European Banking Authority’s (EBA) stress test, the
risk-weighted assets of the 91 European banks that participated in the test increased
by 14 per cent under an adverse scenario as compared to the earlier 2010 ?gure
(EBA, 2011).
III. Bank and sovereign risk linkages
As already noted, CDS spreads may provide a better measure of credit risk than the
ratings of credit rating agencies because they capture the actions of market participants
and are available on a continuous basis so as to capture any changes in the economic and
?nancial environment. The spreads re?ect the cost of insurance against the risk of
default of ?xed-income debt securities. As Figure 4 shows, CDS spreads, as measures
Figure 3.
Sovereign CDS spreads
and bond yields of risk-free
rated sovereign debt
Notes: Data are for August 2011 after S&P’s rating downgrade of US sovereign debt; the
countries in the figure comprise all triple A-rated countries and countries considered risk-free
under Basel III’s risk-weighting scheme (i.e. those with ratings AA- and higher); a few
additional, non-sample countries with AAA sovereign credit ratings are also included for
comparative purposes
Source: Bloomberg; Milken Institute
352
159
108
102
94 90
79 79
71 71
69
63 57
53
40 36
0
S
p
a
i
n
(
A
A
)
F
r
a
n
c
e
A
u
s
t
r
i
a
C
h
i
n
a
(
A
A
-
)
J
a
p
a
n
(
A
A
-
)
D
e
n
m
a
r
k
G
e
r
m
a
n
y
U
K
A
u
s
t
r
a
l
i
a
T
h
e
N
e
t
h
e
r
l
a
n
d
s
H
o
n
g
K
o
n
g
F
i
n
l
a
n
d
U
S
(
A
A
+
)
S
w
e
d
e
n
N
o
r
w
a
y
S
w
i
t
z
e
r
l
a
n
d
S
p
a
i
n
(
A
A
)
F
r
a
n
c
e
A
u
s
t
r
a
l
i
a
A
u
s
t
r
i
a
C
h
i
n
a
(
A
A
-
)
J
a
p
a
n
(
A
A
-
)
F
i
n
l
a
n
d
D
e
n
m
a
r
k
G
e
r
m
a
n
y
U
K
T
h
e
N
e
t
h
e
r
l
a
n
d
s
H
o
n
g
K
o
n
g
U
S
(
A
A
+
)
S
w
e
d
e
n
N
o
r
w
a
y
S
w
i
t
z
e
r
l
a
n
d
100
200
300
400
Five-year CDS spreads (bp) 10-year government bond yield (%)
5.0
4.4
4.0
2.9
2.8
2.62.62.6
2.5
2.4
2.22.2
2.1
1.7
1.01.0
0
1
2
3
4
5
6
JFEP
4,1
80
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
of risk, for bank and sovereign debt tend to move together during a crisis period, though
sometimes the risk of banks is greater than the risk of sovereigns, while at other times
the reverse is the case.
Although there have been relatively few studies that examine the relationship
between bank and sovereign risk (measured by either bond spreads or CDS spreads),
interest in this research area is rapidly growing. Among recent studies are those by
Mody (2009), Acharya et al. (2010), Gerlach et al. (2010), Goldman Sachs Global
Economics (2010) and CGFS-BIS (2011). The general ?ndings fromthese studies indicate
there is an interrelationship between the two types of risk, but with the direction of the
causality between bank and sovereign risk being bi-directional. CGFS-BIS (2011), for
example, argue that greater sovereign risk pushes up the cost of bank funding, which in
turn could lead to greater risk-taking behavior by banks. Acharya et al. (2010),
in contrast, ?nd that the link between bank and sovereign risk becomes stronger in
periods of crisis. This is because when a government bails out banks, risk is shifted from
the banks to the sovereign. Speci?cally, using CDS data for banks and sovereigns from
March 2007 to March 2010, they ?nd on average a signi?cant increase in bank CDS
spreads prior to the ?rst of?cial announcement of a bank bailout (i.e. September 2008)
Figure 4.
Bank and sovereign risk
tend to move together
during a crisis
0
50
100
150
200
250
300
350
400
450
Spain
Banco Santander
Banco Bilbao Vizcaya Argentaria
0
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
500
1,000
1,500
2,000
2,500
3,000
Greece
National Bank of Greece
EFG Eurobank Ergasias
CDS spreads (bp) CDS spreads (bp)
CDS spreads (bp) CDS spreads (bp)
0
50
100
150
200
250
300
350
400
Italy
UniCredit
Intesa Sanpaolo
0
200
400
600
800
1,000
1,200
1,400
1,600
Portugal
Banco Commercial Portugues
Banco Espirito Santo
Note: Risk is measured by CDS spreads
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
81
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
in their western European sample of banks and countries. At the time of the bailout
announcement, they ?nd a signi?cant increase in sovereign CDS spreads, but a decline
in bank CDS spreads. Shortly after the bailout, bank and sovereign CDS spreads tend to
move together.
Our analysis also examines the bank and sovereign risk relationship, but for a
different geographical sample of countries than these other studies. Even though we
focus on selected eurozone countries, for comparative purposes, we also include banks
in China, Japan and the USA( the three home countries for the largest global banking
groups (see Appendix Table AI for the complete list of countries and banks in our
sample). In addition, we investigate the relationship of the riskiness of individual
banks – rather than averages for banking systems as a whole – and sovereigns to
obtain more detailed information about any interrelationships. We ?nd that the bailout
of individual banks contributes to an increase in sovereign CDS spreads, while bank
CDS spreads decline only among those banks that were bailed out. Interestingly,
sovereign CDS spreads rise in all sample countries regardless of whether or not the
countries provided ?nancial assistance to their distressed banks, suggesting the
spreading of problems across national borders.
To examine the effect of bank bailouts on bank and sovereign risk, we calculate the
change in the average CDS spread for both banks and sovereigns one month before and
one month after the ?rst bank bailout in each country. For example, we use October 28,
2008 as the ?rst bailout in the USA as that was when the US government provided
$US25 billion to JPMorgan Chase. The dates of individual bank bailouts and amount of
government ?nancial support to each bank in each country are shown in Table I.
Changes in CDS spreads before and after the bailouts are shown in Figure 5 for
countries in which governments extended ?nancial assistance to domestic banks and
in Figure 6 for those countries that did not provide such support. For comparison
purposes, the periods before and after the end of September 2008 are used to calculate
changes in CDS spreads in Figure 6.
The pattern of changes in CDS spreads in Figure 5 shows that bank bailouts shift
credit risk from the banking sector to the sovereign, which is generally consistent with
the ?ndings of earlier studies (Acharya et al., 2010). However, by looking at the impact
at an individual bank level, we ?nd that the decline in bank CDS spreads after a bailout
announcement occurs only if the banks received ?nancial assistance. The bank CDS
spreads do not decline for banks in the non-bailout country group (Figures 5 and 6).
Other interesting observations are as follows. First, in the case of Barclays, the CDS
spread declined even though the bank was not bailed out. We believe this may be due to
the market’s expectation of a further commitment of support by the sovereign coupled
with the globally diversi?ed operation of the bank. Second, for Societe Generale, Bank of
America, JPMorgan and Citigroup, their bank CDS spreads actually increased after
bailouts, but the increases were relatively small in comparison to the increases in
sovereign risk. Finally, for those countries in which governments did not bail out their
banks, sovereign risk increased nonetheless. This may be due to the interconnectivity of
the global economy and, therefore, some degree of sovereign debt contagion.
Additionally, the credit risk transfer frombailed-out banks to sovereigns seems to be
temporary because as time passes there is a continual increase in the bank and sovereign
risk linkages. The impact of an increase in bank CDS spreads on sovereign CDS spreads
seems to be substantially higher in the post-bailout period (i.e. after September 2008),
JFEP
4,1
82
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
which is illustrated by higher beta coef?cients and R
2
from the regressions in the
post-bank bailout period as compared to those in the pre-bailout period (Table II). (More
detailed results based on several additional sub-periods are provided in Appendix
Table AII).
An increase in sovereign risk and the interconnection between bank and sovereign
risk in most countries became particularly apparent in late-2008 and thereafter. Prior to
the ?nancial crisis, bank CDS spreads had always been higher than sovereign spreads.
The average value of sovereign CDS spreads for developed countries in our sample
was less than 11 basis points in the pre-crisis period (i.e. prior to the summer of 2007),
while the sovereign CDS spreads for emerging markets such as China were somewhat
higher. The average sovereign CDS spreads in these latter countries went up only
slightly after the crisis started, but surged after the period of bank bailouts in late-2008
(Table III). The spreads indicate that investors would have had to pay, on average,
$276,000 a year to insure $10 million of such sovereign debt against default, as
compared to $7,000 a year prior to the crisis. Of the sample of 44 banks in 13 countries,
there are 20 cases, or 45 per cent in which the sovereign CDS spreads increased and
ultimately exceeded bank CDS spreads based on a matching of banks and countries.
Almost all of these cases occurred in the peripheral eurozone countries.
Country Bank Bailout date Bailout amount (US$ billions)
France BNP Paribas March 20, 2009 3.45
BNP Paribas October 22, 2008 3.38
Societe Generale January 21, 2009 2.20
Societe Generale October 21, 2008 2.28
Credit Agricole October 21, 2008 4.23
Germany Commerzbank January 9, 2009 13.64
Commerzbank November 3, 2008 10.43
IKB Deut Indbank February 13, 2008 3.08
Ireland Allied Irish Bank December 31, 2010 4.94
Allied Irish Bank November 28, 2010 6.97
Allied Irish Bank February 11, 2009 4.52
Bank of Ireland November 28, 2010 2.91
Bank of Ireland February 11, 2009 8.04
Anglo Irish Bank Janaury 21, 2009 4.23
UK RBS December 22, 2009 41.09
Lloyd November 3, 2009 9.51
RBS January 19, 2009 7.37
Lloyds January 13, 2009 1.50
Lloyds October 13, 2008 7.66
RBS October 13, 2008 25.53
USA Bank of America January 16, 2009 20.00
Bank of America January 9, 2009 10.00
Bank of America October 26, 2008 15.00
Citibank October 29, 2008 55.00
Wells Fargo October 29, 2008 25.00
JPMorgan October 28, 2008 25.00
Note: We also use the end of September as a cutoff date for Ireland because the actual bailout occurred
at a much later date than other countries and data indicate the risk transferred earlier
Source: Bloomberg
Table I.
Date of bailout and
bank support
The eurozone
?nancial crisis
83
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 5.
Change in CDS spreads
pre- and post-bank
bailouts: countries
with bank bailouts
–50
–30
–10
10
30
50
70
90
110
Before bailout After bailout
Notes: A change of CDS spreads “before bailout” is the difference between the average CDS
spreads one month before the first bailout date in each country (Table I) and the average CDS
spreads since July 2007 up to that month; a change of CDS spreads “after bailout” is the
difference between the monthly average CDS spreads for the month after the bailout and the
monthly average spreads for the month before the bailout; in our data, UBS in Switzerland
received a bailout on October 16, 2008 ($7.3 billion), but it is excluded in this figure due to
unavailability of the sovereign CDS spreads for this time period
Source: Bloomberg; Milken Institute
F
r
a
n
c
e
G
e
r
m
a
n
y
I
r
e
la
n
d
U
n
it
e
d
K
in
g
d
o
m
U
n
it
e
d
S
t
a
t
e
s
B
N
P
P
a
r
ib
a
s
C
r
e
d
it
A
g
r
ic
o
le
S
o
c
ie
t
e
G
e
n
e
r
a
le
C
o
m
m
e
r
z
b
a
n
k
B
a
n
k
o
f
I
r
e
la
n
d
A
llie
d
I
r
is
h
B
a
n
k
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
la
n
d
L
lo
y
d
s
B
a
n
k
in
g
G
r
o
u
p
B
a
n
k
o
f
A
m
e
r
ic
a
C
o
r
p
.
J
P
M
o
r
g
a
n
C
h
a
s
e
&
c
o
.
C
it
ig
r
o
u
p
N
a
t
ix
is
D
e
u
t
s
c
h
e
B
a
n
k
H
S
B
C
B
a
n
k
B
a
r
c
la
y
s
B
a
n
k
S
t
a
n
d
a
r
d
C
h
a
r
t
e
r
e
d
Change in CDS spreads
Sovereign Bailed-out bank Non bailed-out bank
Figure 6.
Change in CDS spreads
pre- and post-bank
bailouts: countries with
no bank bailouts
–150
–100
–50
0
50
100
150
Before bailout After bailout
Change in CDS spreads
Sovereign
Notes: See note in the previous figure on the calculation of a change in average CDS spreads;
banks whose CDS data are unavailable from the end of 2007 are not included here
Source: Bloomberg; Milken Institute
C
h
in
a
G
r
e
e
c
e
J
a
p
a
n
I
t
a
ly
P
o
r
t
u
g
a
l
S
p
a
in
B
a
n
k
o
f
C
h
in
a
N
a
t
in
a
l
B
a
n
k
o
f
G
r
e
e
c
e
M
it
s
u
b
is
h
i
U
F
J
M
iz
u
h
o
B
a
n
k
S
u
m
it
o
m
o
M
it
s
u
i
U
n
iC
r
e
d
it
I
n
t
e
s
a
S
a
n
p
a
o
lo
B
a
n
c
a
M
o
n
t
e
B
a
n
c
o
P
o
p
o
la
r
e
B
a
n
c
o
C
o
m
m
e
r
c
ia
l
P
o
r
t
u
g
u
e
s
B
a
n
c
o
E
s
p
ir
it
o
S
a
n
t
o
B
a
n
c
o
B
P
I
B
a
n
c
o
S
a
n
t
a
n
d
e
r
B
a
n
c
o
B
ilb
a
o
V
iz
c
a
y
a
A
r
g
e
n
t
a
r
ia
B
a
n
c
o
P
o
p
u
la
r
E
s
p
a
n
o
l
JFEP
4,1
84
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As explained by Reinhart and Rogoff (2008) and Gerlach et al. (2010), among others, the
recent increase in sovereign risk is due primarily to the government bailout of domestic
banks, which in turn leads to an increase in public debt and ultimately the risk
of sovereigns honoring their debt obligations. This pattern re?ects the increasing
Sample All All PIIGS Others
Change in bank
CDS spreads 0.157
* *
(0.011) 0.157
* *
(0.011) 0.178
*
(0.086) 0.090
* *
(0.013)
Bailout 0.295
* * *
(0.003) 0.408
* *
(0.046) 0.094
* * *
(0.007)
European debt
crisis 0.512
* *
(0.012) 1.082
* *
(0.035) 0.083
*
(0.092)
Constant 0.043
* * *
(0.000) 20.320
* *
(0.013) 0.943
* * *
(0.006) 20.037 (0.191)
Country dummies Yes Yes Yes Yes
Bank dummies Yes Yes Yes Yes
No. of observations 60,436 60,436 24,529 35,907
R
2
0.044 0.044 0.046 0.055
Notes: Signi?cance at:
*
10,
* *
5 and
* * *
1 per cent levels; the numbers in parentheses are p-values,
based on robust standard errors clustered at the country level; the dependent variable is the change in
sovereign CDS spreads; the regressions are based on the sample of banks and countries listed in the
Appendix 1 (All ¼ all sample, PIIGS ¼ Greece, Ireland, Italy, Portugal and Spain, Others ¼ the
rest of countries in the sample); bailout and European debt crisis are dummy variables equal to one for
the period from September 30, 2008 to April 30, 2010, and from May 1, 2010 to August 3, 2010,
respectively; a dummy for the pre-bailout period is excluded to avoid perfect multi-collinearity
Table II.
Impact of bank risk
on sovereign risk
(1)
Pre-
crisis
(2)
Crisis/
pre-
bailout
(3)
Bailout
period
(4)
Post bailout/pre-
European
sovereign debt
(5)
Eurozone
sovereign
debt crisis
(6)
Eurozone sovereign
debt crisis plus US debt
ceiling talk
Australia n/a 20.35 47.09 69.87 48.65 53.14
China 23.67 47.34 146.38 114.56 75.01 76.92
France 2.83 8.17 32.78 43.21 80.05 86.64
Germany 3.76 8.45 25.65 36.76 42.51 48.30
Greece 10.20 32.76 97.64 220.20 843.23 1327.30
Ireland n/a 23.10 85.09 184.94 369.42 675.26
Italy 9.04 26.48 85.29 114.31 189.27 187.10
Japan 5.16 11.81 34.91 60.21 72.17 85.82
Portugal 6.13 25.72 73.35 98.68 352.90 640.93
Spain 3.28 23.88 74.40 100.17 241.59 266.45
Switzerland n/a n/a n/a 55.93 46.40 35.67
UK 1.89 13.33 47.28 84.48 70.21 62.22
USA n/a 11.40 31.36 43.53 40.84 47.08
Average 7.33 21.07 65.10 94.37 190.17 276.37
Notes: Period 1, pre-crisis (earliest available data – July 19, 2007); Period 2, crisis/pre-bailout ( July 20,
2007-September 29, 2008); Period 3, bailout period (September 30, 2008-January 31, 2009); Period 4, post
bailout/pre-Eurozone sovereign crisis (February 1, 2009-April 30, 2010); Period 5, Eurozone sovereign
debt crisis (May 1, 2010-December 31, 2010); Period 6, Eurozone sovereign debt crisis plus US debt
ceiling talk ( January 1, 2011-August 5, 2011)
Source: Bloomberg; Milken Institute
Table III.
Average value of
sovereign CDS spreads
for different time periods
The eurozone
?nancial crisis
85
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
linkages between the ?nancial and public sectors. In addition, sovereign risk increases
because government spending funded by debt issuance increases during recessions to
stimulate the economy. The result is a higher debt-to-GDP ratio. At the same time,
sovereign risk may increase for some countries due to contagion and a rising
perception of the government’s implicit guarantee to support the ?nancial sector.
The need to consider adjusting bank capital requirements applies directly to that
group of countries and banks for which bank and sovereign risk increase and become
highly correlated. The peripheral European countries fall into this category insofar as
their bank and sovereign risk increased and became highly correlated over time
(Figure 7). A home country bias generally incentivizes domestic banks to hold
government debt securities issued by their home country’s government. Under the
current capital requirement regime, the “true” risk of government debt may not be fully
re?ected in the sovereign credit ratings, and thereby the risk-weighted capital adequacy
of banks.
In the case of some countries, a positive interrelationship between sovereign and bank
risk is less clear. Australia, China, the UK and the USA, for example, experienced the
opposite phenomena in which the correlation between the risk of sovereign debt and the
risk of some banks decreased (Figure 8). A low correlation between bank and sovereign
CDS spreads suggest that any problems in the bank and public sectors in a country are
distinct and separate from one another. In this situation, market participants seem to
perceive that bank risk is unlikely to be shifted to the national government, or vice versa.
We now examine whether low or decreasing correlations can be explained by
bank-speci?c characteristics. The banks included in our sample are among the largest
banks ineachsample country. For the world’s largest economies (i.e. the USA, China, Japan,
France, Germany and the UK), the largest banks are among the top global banking groups
as well. These banks are especially important because they are generally considered
Figure 7.
Countries that experience
an increase in correlations
between bank and
sovereign risk
0
S
e
p
-
0
8
N
o
v
-
0
8
J
a
n
-
0
9
M
a
r
-
0
9
M
a
y
-
0
9
J
u
l
-
0
9
S
e
p
-
0
9
N
o
v
-
0
9
J
a
n
-
1
0
M
a
r
-
1
0
M
a
y
-
1
0
J
u
l
-
1
0
S
e
p
-
1
0
N
o
v
-
1
0
J
a
n
-
1
1
M
a
r
-
1
1
M
a
y
-
1
1
J
u
l
-
1
1
0.2
0.4
0.6
0.8
1
Greece vs.
NBG
Ireland vs.
Bank of Ireland
Italy vs.
UniCredit
Portugal vs.
BCP
Spain vs.
Santander
360-day rolling correlation Sovereign debt crisis: Correlation = 0.88
Correlation = 0.71
Note: The rolling correlation for the sovereign debt crisis period is calculated based on
the period from May 1, 2010 to August 3, 2011
Source: Milken Institute
JFEP
4,1
86
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
systemically important banks since a failure of any one of themhas the potential to spread
to other domestic ?nancial institutions as well as those in other countries.
Speci?cally, we examine next how the bank and sovereign risk relationship depends
on the size of a bank, measured by a bank’s assets relative to its home country’s GDP,
as well as the extent of a bank’s diversi?cation abroad, measured by the percentage of
assets (or net revenue) in foreign countries. And, since most of our sample banks are in
Europe, where the dominant foreign players are still mostly other European banks
(Allen et al., 2011), we also examine whether an increase in bank and sovereign risk
correlations is due to an exposure to sovereign risk in Greece, Ireland and Portugal.
IV. How bank and sovereign risk linkages depend on bank size, foreign
diversi?cation and exposure
4.1 Bank size and the bank and sovereign risk relationship
Closely related to our previous discussion of bank bailout risk transfer to sovereigns is
the issue of too-big-to fail. Over time major banking institutions have gotten even bigger.
Figure 8.
Countries that experience
a decline/mix in
correlations between bank
and sovereign risk
United States
Bank of America
JP Morgan Chase
Citigroup
Wells Fargo
0
0.2
0.4
0.6
0.8
1
Australia
National Australia Bank
Common Wealth Bank of australia
Westpac Banking Corporation
Australia and New Zealand Banking Group
360-day rolling
correlation
360-day rolling
correlation
360-day rolling
correlation
360-day rolling
correlation
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.2
0
0.2
0.4
0.6
0.8
1
China
Industrial & Commercial Bank of China
Bank of China
Agricultural Bank of China
United Kingdom
HSBC Bank
Barclays Bank
Royal Bank of Scotland Group
Lloyds Banking Group
Standard Chartered
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
Note: The rolling correlation for the sovereign debt crisis period is calculated based on the
period from May 1, 2010 to August 3, 2011
Source: Milken Institute
The eurozone
?nancial crisis
87
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The assets for each of the top 10 biggest banks in our sample exceed more than two
trillion US dollars in 2010. The assets of the top banks in many countries, moreover, are
very large relative to their home country’s GDP. UBS and Credit Suisse of Switzerland,
Santander of Spain, BNP Paribas and Cre´dit Agricole of France, Bank of Ireland and
Allied IrishBanks of Ireland, andBarclays inthe UKare among banks inour sample that
have assets that are greater than 100 per cent of their home country’s GDP. The assets of
the biggest US and Japanese banks, in contrast, are less than their country’s GDP (see
Appendix Table AI).
Since our analysis focuses on the risk relationship between the banking and public
sectors, we measure the size of a bank by its asset size relative to the size of the
domestic economy. A failure of a bank that is large relative to national GDP has the
potential to adversely affect the entire economy. Generally, market participants would,
therefore, expect a government to provide implicit guarantees or ?nancial support for
such large banks in the event of severe ?nancial distress. At the same time, however,
bailing out a large bank imposes greater potential ?nancial risk to the sovereign
through the issuance of debt. Thus, correlations between bank and sovereign risk
should rise for big banks during crisis periods.
Figure 9 shows the relationship between bank size and bank-sovereign risk
correlations. There is a positive relationship, suggesting that when domestic banks are
large relative to the economy, instability of an individual bank could be enough to
increase sovereign risk, thereby producing a higher correlation. However, the positive
relationship is found to be relatively weak, insofar as there are some very large banks
whose risk is not highly correlated with home country risk, including banks in
Switzerland and the UK[2]. When these two countries are excluded, however, the result
is an increase in the marginal relationship and better ?t as measured by the beta
coef?cient and R
2
, respectively, (see the right chart of Figure 9). The correlations
between bank and sovereign risk rise, even more so after bailouts. This is most likely
due to the sovereign being in a worse ?nancial position and hence less likely to be able
to provide further ?nancial support, especially when a distressed bank is quite big.
When we expand the coverage to the bailout period of 2008 (the left chart of
Figure 9) the size effect becomes weaker, as shown by both a smaller regression
coef?cient and R
2
, which is likely due to the risk transfer from banks to the sovereign
Figure 9.
Bank size and the
bank-sovereign risk
correlations
y = 0.0714x + 0.4889
R
2
= 0.1084
Correlation
Correlation Correlation
2009-2010
11 countries Switzerland United Kingdom
y = 0.0092x + 0.6444
R
2
= 0.001
–0.8
–0.6
–0.4
–0.2
0
0
0.2
0.2
0.4
0.4
0.6
0.6
0.8
0.8
1
1
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6
Log (% Total Asset to GDP) Log (% Total Asset to GDP) Log (% Total Asset to GDP)
2008-2010
y = 0.0314x + 0.5922
R
2
= 0.0159
0 4
2009-2010
2 6 0 4 2 6
Note: The correlation between the daily sovereign and bank CDS spreads is computed
annually to be comparable with the bank size data which are available each year
Source: Milken Institute
JFEP
4,1
88
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
that occurred during the bank bailouts in most countries in late 2008 (i.e. the correlation
is lower during this bailout period because sovereign risk increased through its efforts
to decrease bank risk by providing ?nancial assistance).
4.2 Bank diversi?cation and the bank-sovereign risk relationship
In an increasingly globally integrated ?nancial system, banks have become more
global by expanding their cross-border activities. As shown in Figure 10, the top global
banking groups in our sample had on average approximately 40 per cent of their total
assets (and total net revenue) outside their home country in 2010. Santander in Spain
had the highest percentage of assets abroad in that year at 96 per cent.
Risk diversi?cation and economies of scale achieved through operations in multiple
national markets are generally considered to be bene?cial for banks. Diversifying assets
abroad makes a bank less exposed to a shock in its own country. Of course, foreign
shocks can spill over to domestic banks through globally integrated banking systems as
well. Arguably, the bene?ts of diversi?cation should outweigh the costs of the potential
risk of contagion via cross-border banking if banks have signi?cant portfolio
diversi?cation (e.g. investing in countries with healthy and stable economies).
Figure 10.
Percentage of bank assets
(and net revenue) abroad:
top global banks, 2010
96%
B
a
n
c
o
S
a
n
t
a
n
d
e
r
U
B
S
A
G
B
a
r
c
l
a
y
s
P
l
c
D
e
u
t
s
c
h
e
B
a
n
k
C
i
t
i
g
r
o
u
p
I
n
c
H
S
B
C
H
o
l
d
i
n
g
s
B
N
P
P
a
r
i
b
a
s
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
J
P
M
o
r
g
a
n
C
h
a
s
e
S
o
c
i
é
t
é
G
é
n
é
r
a
l
e
B
a
n
k
o
f
A
m
e
r
i
c
a
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
B
a
n
k
o
f
C
h
i
n
a
L
i
m
i
t
e
d
I
n
d
&
C
o
m
m
.
B
a
n
k
o
f
C
h
i
n
a
A
g
r
i
c
u
l
t
u
r
a
l
B
a
n
k
o
f
C
h
i
n
a
B
a
n
c
o
S
a
n
t
a
n
d
e
r
U
B
S
A
G
B
a
r
c
l
a
y
s
P
l
c
D
e
u
t
s
c
h
e
B
a
n
k
C
i
t
i
g
r
o
u
p
I
n
c
H
S
B
C
H
o
l
d
i
n
g
s
B
N
P
P
a
r
i
b
a
s
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
J
P
M
o
r
g
a
n
C
h
a
s
e
S
o
c
i
é
t
é
G
é
n
é
r
a
l
e
B
a
n
k
o
f
A
m
e
r
i
c
a
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
B
a
n
k
o
f
C
h
i
n
a
L
i
m
i
t
e
d
I
n
d
&
C
o
m
m
.
B
a
n
k
o
f
C
h
i
n
a
A
g
r
i
c
u
l
t
u
r
a
l
B
a
n
k
o
f
C
h
i
n
a
69% 67%
62%
62%
53% 52%
36%
29% 28%
14% 12%
5% 3% 1%
Bank assets abroad Banks' domestic assets
95%
60%
56%
74%
46%
64% 65%
36%
22%
50%
21%
0% 2% 3% 1%
Banks's net revenues abroad Banks' domestic revenues
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
89
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The opposite would be the case if banks allocate their assets to countries with relatively
high ?nancial and economic risk with the expectation of receiving better investment
returns.
We next examine whether bank asset diversi?cation across borders can reduce
spillover effects of home sovereign risk to banks, and vice versa. This will enable us to
determine whether the bene?ts of diversi?cation can shield banks from home sovereign
shocks. Since distinguishing between strong diversi?cation fromweak diversi?cation is
dif?cult, we simply compare the impact of foreign-asset diversi?cation on the
bank-sovereign risk relationship during a tranquil period to the impact during the crisis
period.
The relationship between bank-asset diversi?cation and the bank-sovereign risk
linkage is shown in Figure 11. The sample is divided into three sub-periods: pre-crisis
(2005-2006), crisis (2007-2008) and European sovereign debt crisis (2009-2010).
A negative relationship indicates foreign asset diversi?cation provides bene?ts by
reducing the bank-sovereign risk linkage.
In the pre-crisis and crisis periods, bank risk is found to be negatively correlated with
sovereign risk for more globally diversi?ed banks, suggesting the bene?cial effect of
cross-border banking. The relationship turns positive during the European sovereign
debt crisis period, however. A possible explanation for this is that market participants
may perceive bank and sovereign risk as being systemic – the risk of individual banks
in Europe cannot be easily distinguished fromone another. Likewise, sovereignrisk may
not be easily separable formthe riskiness of the domestic banking system. Figure 11 also
shows that the bene?ts of foreign diversi?cation in reducing spillover effects are lesser
during the crisis period (the middle chart) than the pre-crisis period (the left chart)[3].
Figure 12 shows the third chart of Figure 11 but with more bank-speci?c detail for
the period 2008-2010. It is worth noting that banks in the upper-right corner are all
European banks that are large and known to be global. These globally diversi?ed
banks are likely to be considered risky because of the recent uncertainty over the
?nancial condition of the European banking system, and also because investors ?nd it
dif?cult to differentiate the extent to which the majority of the banks overseas assets
Figure 11.
Impact of bank asset
diversi?cation on bank
and sovereign risk
correlations
Deutsche Bank
Bank of China
y = –0.0083x + 0.3844
R
2
= 0.2228
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.8 –0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100
Correlation Correlation Correlation
Foreign Asset/Total Asset
0 20 40 60 80 100
Foreign Asset/Total Asset
0 20 40 60 80 100
Foreign Asset/Total Asset
Pre-crisis period
y = 0.0019x + 0.5151
R
2
= 0.0355
European sovereign debt crisis period
Pre-Crisis
Financial Crisis,
Bailout
Post Bailout,
European
Sovereign
Crisis
Citigroup
y = –0.0092x + 0.8107
R
2
= 0.2807
Crisis/bailout period
Notes: The pre-crisis period is from 2005 to 2006, crisis period is from 2007 to 2008, and
European sovereign debt crisis is from 2009 to 2010; the y-axis is the correlation between
sovereign and bank CDS spreads; the correlation is computed annually (i.e. a correlation
coefficient for 2005 refers to CDS data points during January 1, 2005-December 31, 2005)
to be comparable with the bank diversification data which are available each year
Source: Bloomberg; Milken Institute
Banco Santander
JFEP
4,1
90
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
are exposed to the bank and sovereign risk of troubled European countries versus
other parts of the world. As shown in the ?gure, Chinese and US banks still exhibit a
negative relationship between foreign diversi?cation and the correlation between bank
and sovereign risk.
4.3 Bank exposure to the PIIGS sovereign debt and the bank-sovereign risk relationship
The recent spike in sovereign risk in the peripheral eurozone countries (i.e. Greece,
Ireland, Italy, Portugal and Spain; or PIIGS) generates concerns about the riskiness of
banks in other countries that hold sovereign bonds issued by the PIIGS. Cross-country
sovereign exposure occurs insofar as banks headquartered in one country ?nd it
pro?table to hold sovereign debt issued by other countries due to its higher yield, and
this effect is re-enforced to the extent that banks are not required to hold capital against
such debt (see Section 2).
Data for European sovereign debt exposure by European banks are available from
the EBA stress test[4], as the test required the 91 participating European banks to
release their holdings of sovereign debt in 21 European countries. The exposure data
reveal that national banks tend to hold the bonds of their own governments. In other
words, the majority of the PIIGS sovereign debt is held by these countries’ domestic
banks. Even though the negative impact of sovereign defaults would fall primarily on
the domestic banking sector, there could be devastating effects for some individual
banks in other countries of the European Union (Barth et al., forthcoming).
Figure 12.
Bank and sovereign risk
are highly correlated after
the bailout period
regardless of foreign asset
diversi?cation, 2008-2010
BNP Paribas
Deutsche Bank
HSBC
Barclays
Bank of America
JP Morgan
ICBC
Banco Santander
Bank of China
UBS
BNP Paribas
Deutsche Bank
HSBC
Barclays
Banco Santander
UBS
Bank of America
JP Morgan
Bank of China
ICBC
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 10 30 40 50 6 0 70 80 90 100
Correlation
Foreign Asset/Total Asset
2010
2009
2008
Source: Bloomberg; Milken Institute
20
The eurozone
?nancial crisis
91
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
According to EBA (2011), the amount of Greek sovereign debt outstanding is
EUR82.7 billion, of which59per cent of this amount is heldbyGreekbanks as of December
2010 (Table IV). The amount outstanding for Ireland and Portugal is EUR15.8 and
37.6 billion, of which 64 and 50 per cent is held domestically, respectively. (Comparable
data for the individual banks in our sample are provided in Appendix Table AIII).
We make use of these newly released exposure data on a gross basis to assess the
extent to which bank riskiness correlates with Irish, Greek and Portuguese sovereign
risk. As shown in Table IV, although the majority of the PIIGS sovereign debt is held
by domestic banks, the proportion that is held by banks in other European countries is
not a negligible amount, particularly for banks in Germany and France. The positive
relationship shown in Figure 13 shows that bank risk becomes more highly correlated
with the sovereign risk for those banks with greater sovereign exposure to these three
countries, demonstrating market perceptions of greater risk of banks that are more
exposed to this sovereign debt[5].
V. Conclusion
Based upon our empirical results, allowing banks to invest in sovereign debt without
requiring them to hold any capital against the “true” risk of such debt increases
the likelihood of insolvency. This means that interdependencies between bank
Sovereign exposure by counterparty country (as of
December 2010), percentage of total debt outstanding
Greece Ireland Italy Portugal Spain
Participating banks in
Austria 0.56 0.33 0.42 0.36 0.08
Belgium 4.72 1.71 7.19 5.55 1.08
Cyprus 7.03 2.29 0.01 0.00 0.02
Denmark 0.09 2.47 0.15 0.16 0.05
Finland 0.00 0.25 0.00 0.00 0.00
France 11.57 8.77 14.35 10.88 3.51
Germany 9.24 5.99 11.48 9.18 6.45
Greece 58.50 0.11 0.04 0.00 0.00
Ireland 0.05 64.34 0.30 0.65 0.13
Italy 1.66 1.09 55.53 0.95 1.12
Luxembourg 0.10 0.00 0.83 0.48 0.06
Malta 0.01 0.04 0.00 0.01 0.00
The Netherlands 1.41 2.75 2.87 2.06 0.78
Portugal 1.71 3.28 0.36 50.37 0.10
Slovenia 0.02 0.09 0.03 0.04 0.01
Spain 0.54 0.50 2.31 14.44 84.07
Sweden 0.15 0.01 0.13 0.35 0.06
UK 2.62 5.97 4.00 4.53 2.49
Total (91 European banks, billion euros) 82.7 15.8 286.2 37.6 264.4
Notes: Data are net direct positions (gross exposures (long) net of cash short position of sovereign
debt to other counterparties only where there is maturity matching) of 91 European banks (summed up
by country) to PIIGS sovereign debt; the italicized numbers indicates sovereign debt held by domestic
banks
Source: European Banking Authority; Milken Institute
Table IV.
PIIGS sovereign exposure
by counterparty country
JFEP
4,1
92
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and sovereign risk are extremely important when setting regulatory capital
requirements and considering whether action is needed to limit any increase in the
likelihood of contagion.
Notes
1. As Bolton and Jeanne (2011, p. 10) point out:
The European Central Bank modi?ed its rules and declared that Greek bonds would remain
eligible as collateral even with junk status, before announcing a policy of supporting the
price of certain government debts through open market purchases.
2. In addition, a relatively weak relationship is likely to be due to our bank sample selection
which includes only large banks in each country.
3. We also performed a similar analysis for foreign revenue diversi?cation.
The conclusions are very similar to those for foreign asset diversi?cation described in
this subsection.
4. We use the data from the second stress test which became available to the public in July 2011
(the data are as of December 2010). The ?rst stress test was completed in the summer of
2010.
5. Although this relationship is not surprising, it must be interpreted with caution as this
positive link could be driven by relatively the few banks which have sovereign debt holdings
of these three countries in their portfolios. The data for bank exposure to European
sovereign debt are, unfortunately, limited for non-European banks. We assume that
non-European banks’ direct exposure to the peripheral European debts is relatively minor.
Since the majority of European cross-border banking is concentrated within the European
region, this is a reasonable assumption.
Figure 13.
Impact of bank exposure
to Irish, Greek and
Portuguese sovereign debt
on the bank-sovereign risk
correlation
y = 0.0981x – 0.0244
R
2
= 0.4848
–1
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12
Correlations
Log (Exposure)
Notes: The vertical axis is the correlation between domestic bank
CDS spreads and Greek (Irish or Portuguese) sovereign CDS spreads;
the correlation is computed annually from October 2008-August 2011
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
93
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
References
Acharya, V., Drechsler, I. and Schnabl, P. (2010), “A pyrrhic victory? Bank bailouts and
sovereign credit”, Working Paper No. 17136, National Bureau of Economic Research,
Cambridge, MA.
Allen, F., Beck, T., Carletti, E., Lane, P., Schoenmaker, D. and Wagner, W. (2011), Cross-border
Banking in Europe: Implications for Financial Stability and Macroeconomic Policies, Centre
for Economic Policy Research, London.
Barth, J., Li, T. and Prabhavivadhana, A. (2011), “Greece’s ‘unpleasant arithmetic’: containing the
threat to the global economy”, Global Economy Journal, Vol. 11 No. 4, pp. 1-13.
Bolton, P. and Jeanne, O. (2011), “Sovereign default risk and bank fragility in ?nancially
integrated economies”, Working Paper No. 16899, National Bureau of Economic Research,
Cambridge, MA.
CGFS-BIS (2011), “The impact of sovereign credit risk on bank funding conditions”, Committee
on the Global Financial System Papers No. 43, Bank for International Settlements, Basel.
EBA (2011), EU-wide Stress Test Aggregate Report, European Banking Authority, London.
Gerlach, S., Schulz, A. and Wolff, G.B. (2010), “Banking and sovereign risk in the euro area”,
Discussion Paper No. 9, Deutsche Bundesbank, Frankfurt am Main.
Goldman Sachs Global Economics (2010), “Rising funding costs in the periphery”, European
Weekly Analyst Research Report No. 10/13, Goldman Sachs Global Economics,
New York, NY, November 4.
Mody, A. (2009), “From Bear Stearns to Anglo Irish: how eurozone sovereign spreads related to
?nancial sector vulnerability”, Working Paper No. 10, International Monetary Fund,
Washington, DC.
Reinhart, C. and Rogoff, K. (2008), “Banking crises: an equal opportunity menace”,
Working Paper No. 14587, National Bureau of Economic Research, Cambridge, MA.
Wall Street Journal (2011), “Raters fail to see defaults coming: history shows ?rms rarely
anticipate sovereign failures”, August 12 (online.wsj.com).
About the authors
James R. Barth, PhD, is the Senior Finance Fellow at the Milken Institute, as well as the Lowder
Eminent Scholar in Finance at Auburn University.
Apanard (Penny) Prabha, PhD, is an Economist at the Milken Institute.
Apanard (Penny) Prabha is the corresponding author and can be contacted at:
[email protected]
Greg Yun was a former intern at the Milken Institute and currently is an undergraduate
Economics major at the University of Chicago.
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
JFEP
4,1
94
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Appendix
Bank name Asset to GDP (%) Total assets ($US billions)
Australia
National Australia Bank 53.5 662
Westpac Banking Corporation 48.2 597
Commonwealth Bank of Australia 44.2 547
Australia and New Zealand Banking Group 41.5 514
China
Industrial & Commercial Bank of China 34.7 2,042
Bank of China 27.0 1,587
Agricultural Bank of China 26.7 1,569
France
BNP Paribas 104.2 2,671
Cre´dit Agricole 83.1 2,130
Socie´te´ Ge´ne´rale 59.0 1,513
Natixis 23.9 612
Germany
Deutsche Bank 77.5 2,547
Commerzbank 30.7 1,008
IKB Deutsche Industriebank 1.5 48
Greece
National Bank of Greece 52.8 161
EFG Eurobank Ergasias 38.2 117
Ireland
Bank of Ireland 52.8 161
Allied Irish Banks 38.2 117
Italy
UniCredit 60.5 1,242
Intesa Sanpaolo 42.8 881
Banca Monte dei Paschi di Siena 15.9 327
Banco Popolare 8.8 181
Japan
Bank of Tokyo-Mitsubishi UFJ 34.4 1,876
Mizuho Financial Group 30.6 1,672
Sumitomo Mitsui Banking Corp 24.1 1,318
Resona Bank 8.0 436
Portugal
Banco Comercial Portugueˆs 58.3 134
Banco Espirito Santo 48.8 112
Banco BPI 26.6 61
Spain
Banco Santander 115.4 1,627
Banco Bilbao Vizcaya Argentaria 52.4 739
Banco Popular Espanol 12.3 174
Banco de Sabadell 9.2 130
Switzerland
UBS 267.2 1,411
(continued)
Table AI.
Banks and countries
included in sample (data
as of year-end 2010)
The eurozone
?nancial crisis
95
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Bank name Asset to GDP (%) Total assets ($US billions)
Credit Suisse Group 209.3 1,105
UK
Barclays Bank 103.2 2,323
Royal Bank of Scotland Group 100.7 2,266
Lloyds Banking Group 68.7 1,546
HSBC Bank 55.6 1,250
Standard Chartered 23.0 517
USA
Bank of America Corporation 15.6 2,265
JP Morgan Chase & Co. 14.6 2,118
Citigroup 13.2 1,914
Wells Fargo & Company 8.7 1,258
Notes: UK’s largest traded bank is the HSBC holding companies, but the CDS data are available for
the HSBC bank. Japan’s largest traded bank is Mitsubishi UFJ Financial group, but CDS data are
available for Bank of Tokyo-Mitsubishi UFJ the core retail and commercial banking arm of the
Mitsubishi UFJ Financial Group
Source: Bloomberg; World Economic Outlook, International Monetary Fund; Milken Institute Table AI.
All All PIIGS Others
Change bank CDS 0.157
* *
(0.011) 0.157
* *
(0.011) 0.179
*
(0.086) 0.090
* *
(0.013)
Crisis/pre-bailout 0.066 (0.224) 0.005 (0.809) 0.026 (0.159)
Bailout 2.074
* * *
(0.000) 2.968
* * *
(0.000) 1.478
* * *
(0.001)
Post-bailout 0.220
* *
(0.038) 0.267 (0.193) 0.025 (0.394)
Eurozone sovereign
dept in the second
half of 2010 0.467
* * *
(0.007) 0.915
* *
(0.028) 0.076 (0.108)
Eurozone sovereign
dept 2011 0.617
* *
(0.027) 1.275
*
(0.086) 0.114
* *
(0.015)
Constant 0.043
* * *
(0.000) 20.352
* *
(0.013) 0.934
* * *
(0.007) 20.051
* *
(0.037)
Country dummies Yes Yes Yes Yes
Bank dummies Yes Yes Yes Yes
No. of observation 60,436 60,436 24,529 35,907
R
2
0.044 0.045 0.047 0.060
Notes: Signi?cance at:
*
10,
* *
5 and
* * *
1 per cent levels. The numbers in parentheses are p-values,
based on the robust standard errors clustered at the country level. The dependent variable is change in
sovereign CDS spreads. The regressions are based on the sample of banks and countries listed in the
Appendix Table AI (All, all sample; PIIGS, Greece, Ireland, Italy, Portugal and Spain; others, the rest of
countries in the sample). Crisis/pre-bailout, bailout, etc. refer to a dummy which is assigned a value of
one in each speci?ed period (see note in Table III). The dummy of the pre-crisis period is excluded to
avoid the perfect multicollinearity
Table AII.
Impact of bank risk
on sovereign risk
JFEP
4,1
96
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
E
x
p
o
s
u
r
e
t
o
P
I
I
G
S
c
o
u
n
t
r
i
e
s
s
o
v
e
r
e
i
g
n
d
e
b
t
(
a
s
o
f
D
e
c
e
m
b
e
r
2
0
1
0
)
i
n
m
i
l
l
i
o
n
e
u
r
o
s
C
o
u
n
t
r
y
B
a
n
k
s
P
o
r
t
u
g
a
l
I
r
e
l
a
n
d
I
t
a
l
y
G
r
e
e
c
e
S
p
a
i
n
P
I
I
G
S
F
r
a
n
c
e
B
N
P
P
a
r
i
b
a
s
2
,
0
3
3
4
9
7
2
4
,
1
1
4
4
,
9
9
6
3
,
9
0
0
3
5
,
5
4
0
C
r
e ´
d
i
t
A
g
r
i
c
o
l
e
1
,
1
0
9
1
3
5
1
0
,
1
2
3
6
5
5
2
,
7
7
2
1
4
,
7
9
4
S
o
c
i
e ´
t
e ´
G
e ´
n
e ´
r
a
l
e
6
3
1
4
4
2
3
,
3
4
1
2
,
6
5
1
2
,
2
2
0
9
,
2
8
5
G
e
r
m
a
n
y
D
e
u
t
s
c
h
e
B
a
n
k
8
6
4
7
7
5
,
3
3
6
1
,
5
1
0
2
,
0
8
2
9
,
4
9
1
C
o
m
m
e
r
z
b
a
n
k
9
7
6
2
6
1
0
,
1
3
2
3
,
0
4
3
3
,
1
6
5
1
7
,
3
4
2
G
r
e
e
c
e
E
F
G
E
u
r
o
b
a
n
k
E
r
g
a
s
i
a
s
0
0
1
0
0
8
,
7
4
0
0
8
,
8
4
0
N
a
t
i
o
n
a
l
B
a
n
k
o
f
G
r
e
e
c
e
0
1
8
0
1
2
,
8
8
3
0
1
2
,
9
0
1
I
r
e
l
a
n
d
A
l
l
i
e
d
I
r
i
s
h
B
a
n
k
s
2
4
3
5
,
0
4
3
8
1
6
4
0
3
3
5
6
,
4
7
7
B
a
n
k
o
f
I
r
e
l
a
n
d
0
3
,
2
6
9
3
0
0
0
3
,
2
9
9
I
t
a
l
y
I
n
t
e
s
a
S
a
n
p
a
o
l
o
7
0
1
1
4
5
7
,
6
2
2
6
1
9
7
6
4
5
9
,
1
8
9
U
n
i
C
r
e
d
i
t
8
4
5
8
4
7
,
4
4
5
6
3
7
1
,
9
1
0
5
0
,
1
3
4
B
a
n
c
a
M
o
n
t
e
d
e
i
P
a
s
c
h
i
d
i
S
i
e
n
a
2
0
2
0
3
2
,
0
1
8
8
2
8
4
3
2
,
5
1
2
B
a
n
c
o
P
o
p
o
l
a
r
e
0
0
1
1
,
7
5
9
8
7
0
1
1
,
8
4
5
P
o
r
t
u
g
a
l
B
a
n
c
o
C
o
m
e
r
c
i
a
l
P
o
r
t
u
g
u
e ˆ
s
5
,
8
2
9
2
1
3
5
0
7
2
7
0
6
,
8
1
8
B
a
n
c
o
E
s
p
i
r
i
t
o
S
a
n
t
o
2
,
6
8
3
0
0
3
0
9
5
5
3
,
0
4
8
B
a
n
c
o
B
P
I
3
,
8
9
6
2
8
3
9
7
2
3
2
5
0
5
,
4
7
5
S
p
a
i
n
B
a
n
c
o
S
a
n
t
a
n
d
e
r
3
,
6
3
2
0
2
6
1
1
7
7
4
1
,
8
0
7
4
5
,
8
7
7
B
a
n
c
o
B
i
l
b
a
o
V
i
z
c
a
y
a
A
r
g
e
n
t
a
r
i
a
6
4
7
0
3
,
8
9
7
1
2
7
5
3
,
4
5
2
5
8
,
1
2
3
B
a
n
c
o
P
o
p
u
l
a
r
E
s
p
a
n
o
l
6
4
3
0
2
1
0
0
8
,
8
7
4
9
,
7
2
7
B
a
n
c
o
d
e
S
a
b
a
d
e
l
l
9
1
3
8
0
0
7
,
2
9
6
7
,
4
2
5
U
K
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
G
r
o
u
p
2
0
8
4
0
2
4
,
6
5
4
1
,
1
5
5
3
7
9
6
,
7
9
8
H
S
B
C
B
a
n
k
3
2
0
1
3
4
3
,
8
5
6
9
1
9
6
3
6
5
,
8
6
5
B
a
r
c
l
a
y
s
B
a
n
k
1
,
1
7
4
4
0
7
2
,
9
1
5
9
3
5
,
4
9
6
1
0
,
0
8
5
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
0
0
3
2
0
6
2
9
4
S
o
u
r
c
e
:
E
u
r
o
p
e
a
n
B
a
n
k
i
n
g
A
u
t
h
o
r
i
t
y
Table AIII.
PIIGS sovereign exposure
by European bank
The eurozone
?nancial crisis
97
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Maria Cantero-Saiz, Sergio Sanfilippo-Azofra, Begoña Torre-Olmo, Carlos López-Gutiérrez. 2014.
Sovereign Risk And The Bank Lending Channel In Europe. Journal of International Money and Finance
. [CrossRef]
2. Markus K. Brunnermeier, Martin OehmkeBubbles, Financial Crises, and Systemic Risk 1221-1288.
[CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_922225459.pdf
The purpose of this paper is to discuss and then analyze the interdependency between
bank and sovereign risk before, during and after the financial crisis
Journal of Financial Economic Policy
The eurozone financial crisis: role of interdependencies between bank and sovereign
risk
J ames R. Barth Apanard (Penny) Prabha Greg Yun
Article information:
To cite this document:
J ames R. Barth Apanard (Penny) Prabha Greg Yun, (2012),"The eurozone financial crisis: role of
interdependencies between bank and sovereign risk", J ournal of Financial Economic Policy, Vol. 4 Iss 1 pp.
76 - 97
Permanent link to this document:http://dx.doi.org/10.1108/17576381211210203
Downloaded on: 24 January 2016, At: 21:43 (PT)
References: this document contains references to 11 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1011 times since 2012*
Users who downloaded this article also downloaded:
Wenling Lu, David A. Whidbee, (2013),"Bank structure and failure during the financial crisis", J ournal of
Financial Economic Policy, Vol. 5 Iss 3 pp. 281-299http://dx.doi.org/10.1108/J FEP-02-2013-0006
Richard J . Buttimer, (2011),"The financial crisis: imperfect markets and imperfect regulation", J ournal of
Financial Economic Policy, Vol. 3 Iss 1 pp. 12-32http://dx.doi.org/10.1108/17576381111116795
Paul Bennett, (2011),"The (revised) future of financial markets", J ournal of Financial Economic Policy, Vol. 3
Iss 2 pp. 109-122http://dx.doi.org/10.1108/17576381111133598
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The eurozone ?nancial crisis:
role of interdependencies
between bank and sovereign risk
James R. Barth
Auburn University, Auburn, Alabama, USA and
Milken Institute, Santa Monica, California, USA
Apanard (Penny) Prabha
Milken Institute, Santa Monica, California, USA, and
Greg Yun
Santa Monica, California, USA
Abstract
Purpose – The purpose of this paper is to discuss and then analyze the interdependency between
bank and sovereign risk before, during and after the ?nancial crisis.
Design/methodology/approach – The authors’ approach is based upon an examination of 44 large
banks headquartered in 13 countries; eight of these countries belong to the European Union, seven
belong to the eurozone, and the remaining ?ve belong to neither group. This provides a good
comparison group of countries.
Findings – Evidence is found supporting the existence of signi?cant bank andsovereign risklinkages.
There are, however, different patterns in the relationships across countries and even across banks
within the same country. Also, higher correlations between bank and sovereign risk are found in
countries in which the ratio of the assets of banks relative to their home country’s GDPis relatively high.
Research limitations/implications – Based upon the empirical results, allowing banks to invest
in sovereign debt without requiring them to hold any capital against the “true” risk of such debt
increases the likelihood of insolvency. This means that interdependencies between bank and sovereign
risk are extremely important when setting regulatory capital requirements and considering whether
action is needed to limit any increase in the likelihood of contagion.
Originality/value – The paper provides a new examination of the interdependencies between
individual bank risk and the sovereign risk of the countries in which they are headquartered, with
special emphasis on the recent global ?nancial and eurozone crises.
Keywords Banks, Capital, Fixed assets, Sovereign debt crisis, Credit default swap spreads,
Capital requirements, Risk-weighted assets, Bailout, Asset diversi?cation
Paper type Research paper
I. Introduction
A collapse in the ?nancial sectors of several important developed countries gave rise to
the global ?nancial crisis of 2007-2009. This situation, in turn, led to a severe downturn
in economic activity in these countries that spilled over to other countries around the
world. As governments in the developed countries attempted to restore economic
growth through bailouts of banks and other simulative ?scal actions, national de?cits
began rising to alarming levels. This exacerbated problems as investors in the sovereign
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – G01, G0, G, G21, G2, G28
JFEP
4,1
76
Journal of Financial Economic Policy
Vol. 4 No. 1, 2012
pp. 76-97
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/17576381211210203
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
debt issued by these countries increasingly demanded higher interest rates as the
perceived risk of defaults rose. These troubling events shifted much of the focus fromthe
riskiness of banks to the riskiness of sovereign debt. As a result, these two
interdependent factors, heightened bank and sovereign debt risk, contributed to serious
concerns about the stability of the global ?nancial system. The problemwas particularly
acute in the eurozone countries, where concerns were that the banking and sovereign
risk problems could lead to devastating contagion throughout the euro area as well as
spread to other countries through ?nancial and trade channels.
Understanding the interdependencies between bank and sovereign risk is important,
particularly when banking and debt crises are intertwined. Government debt has
historically played an important role in the banking system. Banks hold government
debt enabling them not only to earn interest income but also to hold less regulatory
capital against that debt than is the case for other assets. The reason for the lower capital
requirement is that government securities have typically been considered to be risk free,
which means such assets can be converted to cash at low-transaction costs and at face
value. These properties of government debt securities have made them widely accepted
as collateral in ?nancial systems everywhere. Most importantly, since government debt
securities are considered to bear no risk, regulators have generally not required banks to
hold any capital against such securities. This allows banks to operate with lower
capital-to-asset ratios, which contributes to higher returns on owner-contributed equity
capital.
The recent global ?nancial crisis clearly demonstrates that the role of governments in
domestic banking systems goes well beyond simply being a provider of a supposedly
risk-free security. Indeed, many governments around the globe were forced to rescue big
troubled banks that were deemed to be too systemically important to be allowed to fail.
As nations issued more bonds to ?nance bank bailouts and to combat recessions,
sovereign risk rose along with escalating debt-to-GDP ratios. In some countries,
moreover, such debt became riskier than even the risk of the banks holding it. More
speci?cally, the bailout of banks led to a shift of credit risk from the ?nancial sector to
national governments (Acharya et al., 2010). This situation was further aggravated due
to the economic recessions created by the ?nancial crises during which governments
issued even more debt to fund government spending to stimulate economic activity. The
inter-connectivity between bank and sovereign risk does not stop here. A recent report
by the Bank for International Settlements (Committee on the Global Financial System
(CGFS-BIS), 2011) describes various transmission channels through which increased
sovereign risk can spill over to the banking sector by increasing the funding costs of
banks holding sovereign debt. This means there is a potential vicious cycle linking bank
risk and sovereign risk in a potential downward spiral.
The purpose of this paper is to discuss and thenanalyze the interdependencybetween
bank and sovereign risk by examining 44 banks in 13 countries before, during and after
the ?nancial crisis. Brie?y, we ?nd evidence supporting the existence of signi?cant bank
and sovereign risk linkages. There are, however, different patterns in the relationships
across countries and even across banks within the same country. There are, for example,
banks with relatively little exposure to the government debt of the peripheral eurozone
countries (i.e. Greece, Ireland, Italy, Portugal and Spain – the so-called PIIGS) and with
more globally diversi?ed assets. In these cases, market indicators indicate such banks
are less risky and, therefore, not likely to be bailed out by their governments.
The eurozone
?nancial crisis
77
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Additionally, stronger relationships between bank and sovereign risk are found in
countries in which the ratio of the assets of banks relative to their home country’s GDPis
relatively high. This is consistent with the view that market participants expect
countries in such situations to be more likely to bail out banks in time of distress, thereby
increasing sovereign risk. We ?nd, however, that these patterns are quite different
between the non-crisis and crisis periods.
The remainder of the paper proceeds as follows. In Section II, we discuss the
relationship between sovereign risk and bank capital requirements. In Section III,
the interdependencies between bank and sovereign risk are discussed and analyzed. The
relationship between bank size, foreign diversi?cation and sovereign debt exposure are
examined in Section IV. The conclusion is provided in Section V.
II. Sovereign risk and bank capital requirements
Recent bank regulatory reforms both at the national level such as the US Dodd-Frank
Act and the international level such as Basel III impose stricter capital requirements on
banks. Under Basel I and II, the sovereign debt of a fairly large group of countries was
assigned a zero risk weight in terms of satisfying the regulatory capital requirement.
More speci?cally, Basel I considered the sovereign debt of all OECD countries to be
risk-free. Thus, banks in these countries were permitted to acquire such debt without
increasing their regulatory capital. Basel II, initially published in 2004 and revised
during 2005-2006, considered sovereign debt rated AA- or higher to be risk-free and,
therefore, subject to a zero risk weight. Basel III, initiated in 2010, assigns a
risk-weighting scheme for sovereign debt similar to that under Basel II.
More generally, Basel III recommends that banks be required to maintain a Tier 1
capital ratio of 6 per cent of risk-weighted assets, which is an increase from the previous
ratio of 4 per cent. There is also a liquidity coverage ratio, in which banks are required to
have high-quality liquid assets suf?cient to cover more than 100 per cent of a bank’s
30-day net cash out?ow. Domestic government debt and quali?ed foreign government
debt are considered to be as liquid as cash and, therefore, can be used to meet the
liquidity coverage requirement. Figure 1 shows Basel III’s sovereign creditworthiness
risk-weights and sovereign credit ratings under different risk-weight categories for the
13 countries in our sample. As the ?gure shows, some of the developed countries
(i.e. Greece, Ireland, Italy and Portugal) have lost their risk-free status for regulatory
capital purposes since the start of the European debt crisis in 2010.
Risk-weighting schemes for assets that heavily rely on rating agencies may work
well during normal times but can lead to a misallocation of a bank’s asset portfolio
during troubled times. When banks are not required to hold more capital for government
debt because it is considered risk-free by credit rating agencies or by national regulators,
they are more likely to acquire or continue holding the debt of poorly managed nations
and thus extend less credit to individuals and businesses. In many cases, moreover,
domestic sovereign debt may receive preferential treatment due to “national discretion”
in which such debt is assigned a lower weight in the risk-weighted asset scheme for
regulatory capital purposes[1]. This preferential treatment of national government debt
may be politically expedient, but is not consistent with prudential regulatory practices
whensuchdebt is not trulyrisk free. Of course, sucha practice enables a bank to leverage
to a greater degree and thereby increase its return on equity. It also enables a
government to borrow more cheaply due to the greater demand provided by banks.
JFEP
4,1
78
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In both cases, however, the ultimate cost may be an increase in the riskiness of a
country’s entire ?nancial system and thereby greater likelihood of the potential spread
of problems to other countries.
Risk-weighting schemes for banks also treat all risk-free sovereign debt (i.e. that with
a rating of AA- andhigher) as having the same risk. However, market participants do not
perceive the riskiness of all such debt to be the same, which is evidenced from ?nancial
instruments such as sovereign credit default swaps (CDSs) and spreads in interest rates
on government bonds for different countries. As shown in Figures 2 and 3, there are
Figure 1.
Timeline of Standard
& Poor’s sovereign credit
rating downgrades
RISK FREE ZONE
Australia, France, Germany, UK, Switzerland
USA
A+ to A-
20%
BBB+ to BBB-
50%
BB+ to B-
100%
Below B-
150%
2005 2010 2011 2007 2009 2008 2006
RISK WEIGHTS
Note: Domestic sovereign debt in domestic currency has 100% liquidity
coverage ratio despite its risk rating.
AA-
AA
AA+
AAA
USA
Spain Ireland
Spain Ireland
Ireland
Greece
Greece
Greece
Japan
Japan
Portugal
China
China
China
S&P downgrades Greece,
Portugal and Spain (Apr 10).
Greek fiscal problems are revealed by PM
speech, a month later the deficit number is
revised and doubled from previous
estimation (Nov 09).
Spain and Portugal accept
the austerity package
(May and June10)
Portugal
receives 78
billion bail-
out (May 11)
Italy
Italy
Japan
Portugal
Ireland Spain
Notes: Only the 13 sample countries are included (see Appendix Table AI for the list); countries
that received risk-free status, or a 0 per cent risk weight under Basel II and III, are shown in the
upper part of the figure
Source: Bloomberg; Milken Institute
Ireland
Ireland Portugal
Greece
Eurozone agrees to 30
billion Greece rescue
package (Apr 10)
Greece accepts 110
billion EU-IMF package
(May 10)
Irish banks share prices rebound
from the 20-yrs largest drop
after the Irish Govt. guarantees
bank deposits (Sep 08).
Irish Govt. injects 5.5 billion
in banks (Dec 08).
Irish accepts 68 billion
bail-out (Nov 10)
Figure 2.
Sovereign CDS spreads
and credit ratings
for countries
0
20
40
60
80
100
120
140
160
180
2007 2008 2009 2010 2011 2007 2008 2009 2010 2011
France (AAA)
UK (AAA)
Australia (AAA)
US (AAA)*
Germany (AAA)
Five-year senior sovereign CDS spreads (bp)
0
500
1,000
1,500
2,000
2,500
Greece
Ireland
Portugal
Italy
Spain
AAA
AA+
AA
AA–
A+
A
A–
BBB+
BBB
BBB–
BB+
BB
BB–
B+
B
B–
CCC+
CCC
CCC–
Five-year senior sovereign CDS spreads (bp) S&P sovereign credit ratings
Note: *The rating for US sovereign debt was downgraded from AAA to AA + on
August 5, 2011
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
79
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
signi?cant differences in both CDS spreads and interest rate spreads across countries,
which indicate that market participants to not perceive all sovereign debt to be equally
risky. Notice that the CDS spreads for the sovereign debt of France and Germany are not
the same, even though the debt of both is rated AAA. Furthermore, the CDS spreads for
the debt of both countries are higher than the spread for US debt, even after S&P’s
downgrade of US sovereign debt to AA þ (Figure 3). In general, credit ratings are
slower to adjust and adjust less to a ?nancial crisis as compared to CDS spreads (as
shown in the right panel in Figure 2), and often fail to effectively predict a government’s
risk of defaulting on its debt (Wall Street Journal, 12 August 2011).
The important point based on this data is that capital requirements based on existing
risk-weighting schemes may understate the risk of banks in some countries, particularly
during periods of heightening sovereign risk. The direct and immediate effect of the US
sovereign debt rating downgrade on the required capital to be held by banks was
non-existent since the downgrade from a AAA to AA þ rating did not affect the risk
weighting of assets. However, European banks, many of which had signi?cant exposure
to Greek, Italian, Irish and Portuguese government debt in late 2011 faced a sizeable hit
to capital. According to the European Banking Authority’s (EBA) stress test, the
risk-weighted assets of the 91 European banks that participated in the test increased
by 14 per cent under an adverse scenario as compared to the earlier 2010 ?gure
(EBA, 2011).
III. Bank and sovereign risk linkages
As already noted, CDS spreads may provide a better measure of credit risk than the
ratings of credit rating agencies because they capture the actions of market participants
and are available on a continuous basis so as to capture any changes in the economic and
?nancial environment. The spreads re?ect the cost of insurance against the risk of
default of ?xed-income debt securities. As Figure 4 shows, CDS spreads, as measures
Figure 3.
Sovereign CDS spreads
and bond yields of risk-free
rated sovereign debt
Notes: Data are for August 2011 after S&P’s rating downgrade of US sovereign debt; the
countries in the figure comprise all triple A-rated countries and countries considered risk-free
under Basel III’s risk-weighting scheme (i.e. those with ratings AA- and higher); a few
additional, non-sample countries with AAA sovereign credit ratings are also included for
comparative purposes
Source: Bloomberg; Milken Institute
352
159
108
102
94 90
79 79
71 71
69
63 57
53
40 36
0
S
p
a
i
n
(
A
A
)
F
r
a
n
c
e
A
u
s
t
r
i
a
C
h
i
n
a
(
A
A
-
)
J
a
p
a
n
(
A
A
-
)
D
e
n
m
a
r
k
G
e
r
m
a
n
y
U
K
A
u
s
t
r
a
l
i
a
T
h
e
N
e
t
h
e
r
l
a
n
d
s
H
o
n
g
K
o
n
g
F
i
n
l
a
n
d
U
S
(
A
A
+
)
S
w
e
d
e
n
N
o
r
w
a
y
S
w
i
t
z
e
r
l
a
n
d
S
p
a
i
n
(
A
A
)
F
r
a
n
c
e
A
u
s
t
r
a
l
i
a
A
u
s
t
r
i
a
C
h
i
n
a
(
A
A
-
)
J
a
p
a
n
(
A
A
-
)
F
i
n
l
a
n
d
D
e
n
m
a
r
k
G
e
r
m
a
n
y
U
K
T
h
e
N
e
t
h
e
r
l
a
n
d
s
H
o
n
g
K
o
n
g
U
S
(
A
A
+
)
S
w
e
d
e
n
N
o
r
w
a
y
S
w
i
t
z
e
r
l
a
n
d
100
200
300
400
Five-year CDS spreads (bp) 10-year government bond yield (%)
5.0
4.4
4.0
2.9
2.8
2.62.62.6
2.5
2.4
2.22.2
2.1
1.7
1.01.0
0
1
2
3
4
5
6
JFEP
4,1
80
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
of risk, for bank and sovereign debt tend to move together during a crisis period, though
sometimes the risk of banks is greater than the risk of sovereigns, while at other times
the reverse is the case.
Although there have been relatively few studies that examine the relationship
between bank and sovereign risk (measured by either bond spreads or CDS spreads),
interest in this research area is rapidly growing. Among recent studies are those by
Mody (2009), Acharya et al. (2010), Gerlach et al. (2010), Goldman Sachs Global
Economics (2010) and CGFS-BIS (2011). The general ?ndings fromthese studies indicate
there is an interrelationship between the two types of risk, but with the direction of the
causality between bank and sovereign risk being bi-directional. CGFS-BIS (2011), for
example, argue that greater sovereign risk pushes up the cost of bank funding, which in
turn could lead to greater risk-taking behavior by banks. Acharya et al. (2010),
in contrast, ?nd that the link between bank and sovereign risk becomes stronger in
periods of crisis. This is because when a government bails out banks, risk is shifted from
the banks to the sovereign. Speci?cally, using CDS data for banks and sovereigns from
March 2007 to March 2010, they ?nd on average a signi?cant increase in bank CDS
spreads prior to the ?rst of?cial announcement of a bank bailout (i.e. September 2008)
Figure 4.
Bank and sovereign risk
tend to move together
during a crisis
0
50
100
150
200
250
300
350
400
450
Spain
Banco Santander
Banco Bilbao Vizcaya Argentaria
0
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
J
u
n
-
0
7
J
u
n
-
0
8
J
u
n
-
0
9
J
u
n
-
1
0
J
u
n
-
1
1
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
D
e
c
-
0
7
500
1,000
1,500
2,000
2,500
3,000
Greece
National Bank of Greece
EFG Eurobank Ergasias
CDS spreads (bp) CDS spreads (bp)
CDS spreads (bp) CDS spreads (bp)
0
50
100
150
200
250
300
350
400
Italy
UniCredit
Intesa Sanpaolo
0
200
400
600
800
1,000
1,200
1,400
1,600
Portugal
Banco Commercial Portugues
Banco Espirito Santo
Note: Risk is measured by CDS spreads
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
81
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
in their western European sample of banks and countries. At the time of the bailout
announcement, they ?nd a signi?cant increase in sovereign CDS spreads, but a decline
in bank CDS spreads. Shortly after the bailout, bank and sovereign CDS spreads tend to
move together.
Our analysis also examines the bank and sovereign risk relationship, but for a
different geographical sample of countries than these other studies. Even though we
focus on selected eurozone countries, for comparative purposes, we also include banks
in China, Japan and the USA( the three home countries for the largest global banking
groups (see Appendix Table AI for the complete list of countries and banks in our
sample). In addition, we investigate the relationship of the riskiness of individual
banks – rather than averages for banking systems as a whole – and sovereigns to
obtain more detailed information about any interrelationships. We ?nd that the bailout
of individual banks contributes to an increase in sovereign CDS spreads, while bank
CDS spreads decline only among those banks that were bailed out. Interestingly,
sovereign CDS spreads rise in all sample countries regardless of whether or not the
countries provided ?nancial assistance to their distressed banks, suggesting the
spreading of problems across national borders.
To examine the effect of bank bailouts on bank and sovereign risk, we calculate the
change in the average CDS spread for both banks and sovereigns one month before and
one month after the ?rst bank bailout in each country. For example, we use October 28,
2008 as the ?rst bailout in the USA as that was when the US government provided
$US25 billion to JPMorgan Chase. The dates of individual bank bailouts and amount of
government ?nancial support to each bank in each country are shown in Table I.
Changes in CDS spreads before and after the bailouts are shown in Figure 5 for
countries in which governments extended ?nancial assistance to domestic banks and
in Figure 6 for those countries that did not provide such support. For comparison
purposes, the periods before and after the end of September 2008 are used to calculate
changes in CDS spreads in Figure 6.
The pattern of changes in CDS spreads in Figure 5 shows that bank bailouts shift
credit risk from the banking sector to the sovereign, which is generally consistent with
the ?ndings of earlier studies (Acharya et al., 2010). However, by looking at the impact
at an individual bank level, we ?nd that the decline in bank CDS spreads after a bailout
announcement occurs only if the banks received ?nancial assistance. The bank CDS
spreads do not decline for banks in the non-bailout country group (Figures 5 and 6).
Other interesting observations are as follows. First, in the case of Barclays, the CDS
spread declined even though the bank was not bailed out. We believe this may be due to
the market’s expectation of a further commitment of support by the sovereign coupled
with the globally diversi?ed operation of the bank. Second, for Societe Generale, Bank of
America, JPMorgan and Citigroup, their bank CDS spreads actually increased after
bailouts, but the increases were relatively small in comparison to the increases in
sovereign risk. Finally, for those countries in which governments did not bail out their
banks, sovereign risk increased nonetheless. This may be due to the interconnectivity of
the global economy and, therefore, some degree of sovereign debt contagion.
Additionally, the credit risk transfer frombailed-out banks to sovereigns seems to be
temporary because as time passes there is a continual increase in the bank and sovereign
risk linkages. The impact of an increase in bank CDS spreads on sovereign CDS spreads
seems to be substantially higher in the post-bailout period (i.e. after September 2008),
JFEP
4,1
82
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
which is illustrated by higher beta coef?cients and R
2
from the regressions in the
post-bank bailout period as compared to those in the pre-bailout period (Table II). (More
detailed results based on several additional sub-periods are provided in Appendix
Table AII).
An increase in sovereign risk and the interconnection between bank and sovereign
risk in most countries became particularly apparent in late-2008 and thereafter. Prior to
the ?nancial crisis, bank CDS spreads had always been higher than sovereign spreads.
The average value of sovereign CDS spreads for developed countries in our sample
was less than 11 basis points in the pre-crisis period (i.e. prior to the summer of 2007),
while the sovereign CDS spreads for emerging markets such as China were somewhat
higher. The average sovereign CDS spreads in these latter countries went up only
slightly after the crisis started, but surged after the period of bank bailouts in late-2008
(Table III). The spreads indicate that investors would have had to pay, on average,
$276,000 a year to insure $10 million of such sovereign debt against default, as
compared to $7,000 a year prior to the crisis. Of the sample of 44 banks in 13 countries,
there are 20 cases, or 45 per cent in which the sovereign CDS spreads increased and
ultimately exceeded bank CDS spreads based on a matching of banks and countries.
Almost all of these cases occurred in the peripheral eurozone countries.
Country Bank Bailout date Bailout amount (US$ billions)
France BNP Paribas March 20, 2009 3.45
BNP Paribas October 22, 2008 3.38
Societe Generale January 21, 2009 2.20
Societe Generale October 21, 2008 2.28
Credit Agricole October 21, 2008 4.23
Germany Commerzbank January 9, 2009 13.64
Commerzbank November 3, 2008 10.43
IKB Deut Indbank February 13, 2008 3.08
Ireland Allied Irish Bank December 31, 2010 4.94
Allied Irish Bank November 28, 2010 6.97
Allied Irish Bank February 11, 2009 4.52
Bank of Ireland November 28, 2010 2.91
Bank of Ireland February 11, 2009 8.04
Anglo Irish Bank Janaury 21, 2009 4.23
UK RBS December 22, 2009 41.09
Lloyd November 3, 2009 9.51
RBS January 19, 2009 7.37
Lloyds January 13, 2009 1.50
Lloyds October 13, 2008 7.66
RBS October 13, 2008 25.53
USA Bank of America January 16, 2009 20.00
Bank of America January 9, 2009 10.00
Bank of America October 26, 2008 15.00
Citibank October 29, 2008 55.00
Wells Fargo October 29, 2008 25.00
JPMorgan October 28, 2008 25.00
Note: We also use the end of September as a cutoff date for Ireland because the actual bailout occurred
at a much later date than other countries and data indicate the risk transferred earlier
Source: Bloomberg
Table I.
Date of bailout and
bank support
The eurozone
?nancial crisis
83
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Figure 5.
Change in CDS spreads
pre- and post-bank
bailouts: countries
with bank bailouts
–50
–30
–10
10
30
50
70
90
110
Before bailout After bailout
Notes: A change of CDS spreads “before bailout” is the difference between the average CDS
spreads one month before the first bailout date in each country (Table I) and the average CDS
spreads since July 2007 up to that month; a change of CDS spreads “after bailout” is the
difference between the monthly average CDS spreads for the month after the bailout and the
monthly average spreads for the month before the bailout; in our data, UBS in Switzerland
received a bailout on October 16, 2008 ($7.3 billion), but it is excluded in this figure due to
unavailability of the sovereign CDS spreads for this time period
Source: Bloomberg; Milken Institute
F
r
a
n
c
e
G
e
r
m
a
n
y
I
r
e
la
n
d
U
n
it
e
d
K
in
g
d
o
m
U
n
it
e
d
S
t
a
t
e
s
B
N
P
P
a
r
ib
a
s
C
r
e
d
it
A
g
r
ic
o
le
S
o
c
ie
t
e
G
e
n
e
r
a
le
C
o
m
m
e
r
z
b
a
n
k
B
a
n
k
o
f
I
r
e
la
n
d
A
llie
d
I
r
is
h
B
a
n
k
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
la
n
d
L
lo
y
d
s
B
a
n
k
in
g
G
r
o
u
p
B
a
n
k
o
f
A
m
e
r
ic
a
C
o
r
p
.
J
P
M
o
r
g
a
n
C
h
a
s
e
&
c
o
.
C
it
ig
r
o
u
p
N
a
t
ix
is
D
e
u
t
s
c
h
e
B
a
n
k
H
S
B
C
B
a
n
k
B
a
r
c
la
y
s
B
a
n
k
S
t
a
n
d
a
r
d
C
h
a
r
t
e
r
e
d
Change in CDS spreads
Sovereign Bailed-out bank Non bailed-out bank
Figure 6.
Change in CDS spreads
pre- and post-bank
bailouts: countries with
no bank bailouts
–150
–100
–50
0
50
100
150
Before bailout After bailout
Change in CDS spreads
Sovereign
Notes: See note in the previous figure on the calculation of a change in average CDS spreads;
banks whose CDS data are unavailable from the end of 2007 are not included here
Source: Bloomberg; Milken Institute
C
h
in
a
G
r
e
e
c
e
J
a
p
a
n
I
t
a
ly
P
o
r
t
u
g
a
l
S
p
a
in
B
a
n
k
o
f
C
h
in
a
N
a
t
in
a
l
B
a
n
k
o
f
G
r
e
e
c
e
M
it
s
u
b
is
h
i
U
F
J
M
iz
u
h
o
B
a
n
k
S
u
m
it
o
m
o
M
it
s
u
i
U
n
iC
r
e
d
it
I
n
t
e
s
a
S
a
n
p
a
o
lo
B
a
n
c
a
M
o
n
t
e
B
a
n
c
o
P
o
p
o
la
r
e
B
a
n
c
o
C
o
m
m
e
r
c
ia
l
P
o
r
t
u
g
u
e
s
B
a
n
c
o
E
s
p
ir
it
o
S
a
n
t
o
B
a
n
c
o
B
P
I
B
a
n
c
o
S
a
n
t
a
n
d
e
r
B
a
n
c
o
B
ilb
a
o
V
iz
c
a
y
a
A
r
g
e
n
t
a
r
ia
B
a
n
c
o
P
o
p
u
la
r
E
s
p
a
n
o
l
JFEP
4,1
84
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
As explained by Reinhart and Rogoff (2008) and Gerlach et al. (2010), among others, the
recent increase in sovereign risk is due primarily to the government bailout of domestic
banks, which in turn leads to an increase in public debt and ultimately the risk
of sovereigns honoring their debt obligations. This pattern re?ects the increasing
Sample All All PIIGS Others
Change in bank
CDS spreads 0.157
* *
(0.011) 0.157
* *
(0.011) 0.178
*
(0.086) 0.090
* *
(0.013)
Bailout 0.295
* * *
(0.003) 0.408
* *
(0.046) 0.094
* * *
(0.007)
European debt
crisis 0.512
* *
(0.012) 1.082
* *
(0.035) 0.083
*
(0.092)
Constant 0.043
* * *
(0.000) 20.320
* *
(0.013) 0.943
* * *
(0.006) 20.037 (0.191)
Country dummies Yes Yes Yes Yes
Bank dummies Yes Yes Yes Yes
No. of observations 60,436 60,436 24,529 35,907
R
2
0.044 0.044 0.046 0.055
Notes: Signi?cance at:
*
10,
* *
5 and
* * *
1 per cent levels; the numbers in parentheses are p-values,
based on robust standard errors clustered at the country level; the dependent variable is the change in
sovereign CDS spreads; the regressions are based on the sample of banks and countries listed in the
Appendix 1 (All ¼ all sample, PIIGS ¼ Greece, Ireland, Italy, Portugal and Spain, Others ¼ the
rest of countries in the sample); bailout and European debt crisis are dummy variables equal to one for
the period from September 30, 2008 to April 30, 2010, and from May 1, 2010 to August 3, 2010,
respectively; a dummy for the pre-bailout period is excluded to avoid perfect multi-collinearity
Table II.
Impact of bank risk
on sovereign risk
(1)
Pre-
crisis
(2)
Crisis/
pre-
bailout
(3)
Bailout
period
(4)
Post bailout/pre-
European
sovereign debt
(5)
Eurozone
sovereign
debt crisis
(6)
Eurozone sovereign
debt crisis plus US debt
ceiling talk
Australia n/a 20.35 47.09 69.87 48.65 53.14
China 23.67 47.34 146.38 114.56 75.01 76.92
France 2.83 8.17 32.78 43.21 80.05 86.64
Germany 3.76 8.45 25.65 36.76 42.51 48.30
Greece 10.20 32.76 97.64 220.20 843.23 1327.30
Ireland n/a 23.10 85.09 184.94 369.42 675.26
Italy 9.04 26.48 85.29 114.31 189.27 187.10
Japan 5.16 11.81 34.91 60.21 72.17 85.82
Portugal 6.13 25.72 73.35 98.68 352.90 640.93
Spain 3.28 23.88 74.40 100.17 241.59 266.45
Switzerland n/a n/a n/a 55.93 46.40 35.67
UK 1.89 13.33 47.28 84.48 70.21 62.22
USA n/a 11.40 31.36 43.53 40.84 47.08
Average 7.33 21.07 65.10 94.37 190.17 276.37
Notes: Period 1, pre-crisis (earliest available data – July 19, 2007); Period 2, crisis/pre-bailout ( July 20,
2007-September 29, 2008); Period 3, bailout period (September 30, 2008-January 31, 2009); Period 4, post
bailout/pre-Eurozone sovereign crisis (February 1, 2009-April 30, 2010); Period 5, Eurozone sovereign
debt crisis (May 1, 2010-December 31, 2010); Period 6, Eurozone sovereign debt crisis plus US debt
ceiling talk ( January 1, 2011-August 5, 2011)
Source: Bloomberg; Milken Institute
Table III.
Average value of
sovereign CDS spreads
for different time periods
The eurozone
?nancial crisis
85
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
linkages between the ?nancial and public sectors. In addition, sovereign risk increases
because government spending funded by debt issuance increases during recessions to
stimulate the economy. The result is a higher debt-to-GDP ratio. At the same time,
sovereign risk may increase for some countries due to contagion and a rising
perception of the government’s implicit guarantee to support the ?nancial sector.
The need to consider adjusting bank capital requirements applies directly to that
group of countries and banks for which bank and sovereign risk increase and become
highly correlated. The peripheral European countries fall into this category insofar as
their bank and sovereign risk increased and became highly correlated over time
(Figure 7). A home country bias generally incentivizes domestic banks to hold
government debt securities issued by their home country’s government. Under the
current capital requirement regime, the “true” risk of government debt may not be fully
re?ected in the sovereign credit ratings, and thereby the risk-weighted capital adequacy
of banks.
In the case of some countries, a positive interrelationship between sovereign and bank
risk is less clear. Australia, China, the UK and the USA, for example, experienced the
opposite phenomena in which the correlation between the risk of sovereign debt and the
risk of some banks decreased (Figure 8). A low correlation between bank and sovereign
CDS spreads suggest that any problems in the bank and public sectors in a country are
distinct and separate from one another. In this situation, market participants seem to
perceive that bank risk is unlikely to be shifted to the national government, or vice versa.
We now examine whether low or decreasing correlations can be explained by
bank-speci?c characteristics. The banks included in our sample are among the largest
banks ineachsample country. For the world’s largest economies (i.e. the USA, China, Japan,
France, Germany and the UK), the largest banks are among the top global banking groups
as well. These banks are especially important because they are generally considered
Figure 7.
Countries that experience
an increase in correlations
between bank and
sovereign risk
0
S
e
p
-
0
8
N
o
v
-
0
8
J
a
n
-
0
9
M
a
r
-
0
9
M
a
y
-
0
9
J
u
l
-
0
9
S
e
p
-
0
9
N
o
v
-
0
9
J
a
n
-
1
0
M
a
r
-
1
0
M
a
y
-
1
0
J
u
l
-
1
0
S
e
p
-
1
0
N
o
v
-
1
0
J
a
n
-
1
1
M
a
r
-
1
1
M
a
y
-
1
1
J
u
l
-
1
1
0.2
0.4
0.6
0.8
1
Greece vs.
NBG
Ireland vs.
Bank of Ireland
Italy vs.
UniCredit
Portugal vs.
BCP
Spain vs.
Santander
360-day rolling correlation Sovereign debt crisis: Correlation = 0.88
Correlation = 0.71
Note: The rolling correlation for the sovereign debt crisis period is calculated based on
the period from May 1, 2010 to August 3, 2011
Source: Milken Institute
JFEP
4,1
86
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
systemically important banks since a failure of any one of themhas the potential to spread
to other domestic ?nancial institutions as well as those in other countries.
Speci?cally, we examine next how the bank and sovereign risk relationship depends
on the size of a bank, measured by a bank’s assets relative to its home country’s GDP,
as well as the extent of a bank’s diversi?cation abroad, measured by the percentage of
assets (or net revenue) in foreign countries. And, since most of our sample banks are in
Europe, where the dominant foreign players are still mostly other European banks
(Allen et al., 2011), we also examine whether an increase in bank and sovereign risk
correlations is due to an exposure to sovereign risk in Greece, Ireland and Portugal.
IV. How bank and sovereign risk linkages depend on bank size, foreign
diversi?cation and exposure
4.1 Bank size and the bank and sovereign risk relationship
Closely related to our previous discussion of bank bailout risk transfer to sovereigns is
the issue of too-big-to fail. Over time major banking institutions have gotten even bigger.
Figure 8.
Countries that experience
a decline/mix in
correlations between bank
and sovereign risk
United States
Bank of America
JP Morgan Chase
Citigroup
Wells Fargo
0
0.2
0.4
0.6
0.8
1
Australia
National Australia Bank
Common Wealth Bank of australia
Westpac Banking Corporation
Australia and New Zealand Banking Group
360-day rolling
correlation
360-day rolling
correlation
360-day rolling
correlation
360-day rolling
correlation
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.2
0
0.2
0.4
0.6
0.8
1
China
Industrial & Commercial Bank of China
Bank of China
Agricultural Bank of China
United Kingdom
HSBC Bank
Barclays Bank
Royal Bank of Scotland Group
Lloyds Banking Group
Standard Chartered
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
S
e
p
-
0
8
S
e
p
-
1
0
D
e
c
-
0
8
D
e
c
-
0
9
D
e
c
-
1
0
J
u
n
-
0
9
M
a
r
-
0
9
S
e
p
-
0
9
M
a
r
-
1
0
J
u
n
-
1
0
M
a
r
-
1
1
J
u
l
-
1
1
Note: The rolling correlation for the sovereign debt crisis period is calculated based on the
period from May 1, 2010 to August 3, 2011
Source: Milken Institute
The eurozone
?nancial crisis
87
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The assets for each of the top 10 biggest banks in our sample exceed more than two
trillion US dollars in 2010. The assets of the top banks in many countries, moreover, are
very large relative to their home country’s GDP. UBS and Credit Suisse of Switzerland,
Santander of Spain, BNP Paribas and Cre´dit Agricole of France, Bank of Ireland and
Allied IrishBanks of Ireland, andBarclays inthe UKare among banks inour sample that
have assets that are greater than 100 per cent of their home country’s GDP. The assets of
the biggest US and Japanese banks, in contrast, are less than their country’s GDP (see
Appendix Table AI).
Since our analysis focuses on the risk relationship between the banking and public
sectors, we measure the size of a bank by its asset size relative to the size of the
domestic economy. A failure of a bank that is large relative to national GDP has the
potential to adversely affect the entire economy. Generally, market participants would,
therefore, expect a government to provide implicit guarantees or ?nancial support for
such large banks in the event of severe ?nancial distress. At the same time, however,
bailing out a large bank imposes greater potential ?nancial risk to the sovereign
through the issuance of debt. Thus, correlations between bank and sovereign risk
should rise for big banks during crisis periods.
Figure 9 shows the relationship between bank size and bank-sovereign risk
correlations. There is a positive relationship, suggesting that when domestic banks are
large relative to the economy, instability of an individual bank could be enough to
increase sovereign risk, thereby producing a higher correlation. However, the positive
relationship is found to be relatively weak, insofar as there are some very large banks
whose risk is not highly correlated with home country risk, including banks in
Switzerland and the UK[2]. When these two countries are excluded, however, the result
is an increase in the marginal relationship and better ?t as measured by the beta
coef?cient and R
2
, respectively, (see the right chart of Figure 9). The correlations
between bank and sovereign risk rise, even more so after bailouts. This is most likely
due to the sovereign being in a worse ?nancial position and hence less likely to be able
to provide further ?nancial support, especially when a distressed bank is quite big.
When we expand the coverage to the bailout period of 2008 (the left chart of
Figure 9) the size effect becomes weaker, as shown by both a smaller regression
coef?cient and R
2
, which is likely due to the risk transfer from banks to the sovereign
Figure 9.
Bank size and the
bank-sovereign risk
correlations
y = 0.0714x + 0.4889
R
2
= 0.1084
Correlation
Correlation Correlation
2009-2010
11 countries Switzerland United Kingdom
y = 0.0092x + 0.6444
R
2
= 0.001
–0.8
–0.6
–0.4
–0.2
0
0
0.2
0.2
0.4
0.4
0.6
0.6
0.8
0.8
1
1
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6
Log (% Total Asset to GDP) Log (% Total Asset to GDP) Log (% Total Asset to GDP)
2008-2010
y = 0.0314x + 0.5922
R
2
= 0.0159
0 4
2009-2010
2 6 0 4 2 6
Note: The correlation between the daily sovereign and bank CDS spreads is computed
annually to be comparable with the bank size data which are available each year
Source: Milken Institute
JFEP
4,1
88
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
that occurred during the bank bailouts in most countries in late 2008 (i.e. the correlation
is lower during this bailout period because sovereign risk increased through its efforts
to decrease bank risk by providing ?nancial assistance).
4.2 Bank diversi?cation and the bank-sovereign risk relationship
In an increasingly globally integrated ?nancial system, banks have become more
global by expanding their cross-border activities. As shown in Figure 10, the top global
banking groups in our sample had on average approximately 40 per cent of their total
assets (and total net revenue) outside their home country in 2010. Santander in Spain
had the highest percentage of assets abroad in that year at 96 per cent.
Risk diversi?cation and economies of scale achieved through operations in multiple
national markets are generally considered to be bene?cial for banks. Diversifying assets
abroad makes a bank less exposed to a shock in its own country. Of course, foreign
shocks can spill over to domestic banks through globally integrated banking systems as
well. Arguably, the bene?ts of diversi?cation should outweigh the costs of the potential
risk of contagion via cross-border banking if banks have signi?cant portfolio
diversi?cation (e.g. investing in countries with healthy and stable economies).
Figure 10.
Percentage of bank assets
(and net revenue) abroad:
top global banks, 2010
96%
B
a
n
c
o
S
a
n
t
a
n
d
e
r
U
B
S
A
G
B
a
r
c
l
a
y
s
P
l
c
D
e
u
t
s
c
h
e
B
a
n
k
C
i
t
i
g
r
o
u
p
I
n
c
H
S
B
C
H
o
l
d
i
n
g
s
B
N
P
P
a
r
i
b
a
s
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
J
P
M
o
r
g
a
n
C
h
a
s
e
S
o
c
i
é
t
é
G
é
n
é
r
a
l
e
B
a
n
k
o
f
A
m
e
r
i
c
a
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
B
a
n
k
o
f
C
h
i
n
a
L
i
m
i
t
e
d
I
n
d
&
C
o
m
m
.
B
a
n
k
o
f
C
h
i
n
a
A
g
r
i
c
u
l
t
u
r
a
l
B
a
n
k
o
f
C
h
i
n
a
B
a
n
c
o
S
a
n
t
a
n
d
e
r
U
B
S
A
G
B
a
r
c
l
a
y
s
P
l
c
D
e
u
t
s
c
h
e
B
a
n
k
C
i
t
i
g
r
o
u
p
I
n
c
H
S
B
C
H
o
l
d
i
n
g
s
B
N
P
P
a
r
i
b
a
s
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
J
P
M
o
r
g
a
n
C
h
a
s
e
S
o
c
i
é
t
é
G
é
n
é
r
a
l
e
B
a
n
k
o
f
A
m
e
r
i
c
a
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
B
a
n
k
o
f
C
h
i
n
a
L
i
m
i
t
e
d
I
n
d
&
C
o
m
m
.
B
a
n
k
o
f
C
h
i
n
a
A
g
r
i
c
u
l
t
u
r
a
l
B
a
n
k
o
f
C
h
i
n
a
69% 67%
62%
62%
53% 52%
36%
29% 28%
14% 12%
5% 3% 1%
Bank assets abroad Banks' domestic assets
95%
60%
56%
74%
46%
64% 65%
36%
22%
50%
21%
0% 2% 3% 1%
Banks's net revenues abroad Banks' domestic revenues
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
89
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The opposite would be the case if banks allocate their assets to countries with relatively
high ?nancial and economic risk with the expectation of receiving better investment
returns.
We next examine whether bank asset diversi?cation across borders can reduce
spillover effects of home sovereign risk to banks, and vice versa. This will enable us to
determine whether the bene?ts of diversi?cation can shield banks from home sovereign
shocks. Since distinguishing between strong diversi?cation fromweak diversi?cation is
dif?cult, we simply compare the impact of foreign-asset diversi?cation on the
bank-sovereign risk relationship during a tranquil period to the impact during the crisis
period.
The relationship between bank-asset diversi?cation and the bank-sovereign risk
linkage is shown in Figure 11. The sample is divided into three sub-periods: pre-crisis
(2005-2006), crisis (2007-2008) and European sovereign debt crisis (2009-2010).
A negative relationship indicates foreign asset diversi?cation provides bene?ts by
reducing the bank-sovereign risk linkage.
In the pre-crisis and crisis periods, bank risk is found to be negatively correlated with
sovereign risk for more globally diversi?ed banks, suggesting the bene?cial effect of
cross-border banking. The relationship turns positive during the European sovereign
debt crisis period, however. A possible explanation for this is that market participants
may perceive bank and sovereign risk as being systemic – the risk of individual banks
in Europe cannot be easily distinguished fromone another. Likewise, sovereignrisk may
not be easily separable formthe riskiness of the domestic banking system. Figure 11 also
shows that the bene?ts of foreign diversi?cation in reducing spillover effects are lesser
during the crisis period (the middle chart) than the pre-crisis period (the left chart)[3].
Figure 12 shows the third chart of Figure 11 but with more bank-speci?c detail for
the period 2008-2010. It is worth noting that banks in the upper-right corner are all
European banks that are large and known to be global. These globally diversi?ed
banks are likely to be considered risky because of the recent uncertainty over the
?nancial condition of the European banking system, and also because investors ?nd it
dif?cult to differentiate the extent to which the majority of the banks overseas assets
Figure 11.
Impact of bank asset
diversi?cation on bank
and sovereign risk
correlations
Deutsche Bank
Bank of China
y = –0.0083x + 0.3844
R
2
= 0.2228
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.8 –0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 20 40 60 80 100
Correlation Correlation Correlation
Foreign Asset/Total Asset
0 20 40 60 80 100
Foreign Asset/Total Asset
0 20 40 60 80 100
Foreign Asset/Total Asset
Pre-crisis period
y = 0.0019x + 0.5151
R
2
= 0.0355
European sovereign debt crisis period
Pre-Crisis
Financial Crisis,
Bailout
Post Bailout,
European
Sovereign
Crisis
Citigroup
y = –0.0092x + 0.8107
R
2
= 0.2807
Crisis/bailout period
Notes: The pre-crisis period is from 2005 to 2006, crisis period is from 2007 to 2008, and
European sovereign debt crisis is from 2009 to 2010; the y-axis is the correlation between
sovereign and bank CDS spreads; the correlation is computed annually (i.e. a correlation
coefficient for 2005 refers to CDS data points during January 1, 2005-December 31, 2005)
to be comparable with the bank diversification data which are available each year
Source: Bloomberg; Milken Institute
Banco Santander
JFEP
4,1
90
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
are exposed to the bank and sovereign risk of troubled European countries versus
other parts of the world. As shown in the ?gure, Chinese and US banks still exhibit a
negative relationship between foreign diversi?cation and the correlation between bank
and sovereign risk.
4.3 Bank exposure to the PIIGS sovereign debt and the bank-sovereign risk relationship
The recent spike in sovereign risk in the peripheral eurozone countries (i.e. Greece,
Ireland, Italy, Portugal and Spain; or PIIGS) generates concerns about the riskiness of
banks in other countries that hold sovereign bonds issued by the PIIGS. Cross-country
sovereign exposure occurs insofar as banks headquartered in one country ?nd it
pro?table to hold sovereign debt issued by other countries due to its higher yield, and
this effect is re-enforced to the extent that banks are not required to hold capital against
such debt (see Section 2).
Data for European sovereign debt exposure by European banks are available from
the EBA stress test[4], as the test required the 91 participating European banks to
release their holdings of sovereign debt in 21 European countries. The exposure data
reveal that national banks tend to hold the bonds of their own governments. In other
words, the majority of the PIIGS sovereign debt is held by these countries’ domestic
banks. Even though the negative impact of sovereign defaults would fall primarily on
the domestic banking sector, there could be devastating effects for some individual
banks in other countries of the European Union (Barth et al., forthcoming).
Figure 12.
Bank and sovereign risk
are highly correlated after
the bailout period
regardless of foreign asset
diversi?cation, 2008-2010
BNP Paribas
Deutsche Bank
HSBC
Barclays
Bank of America
JP Morgan
ICBC
Banco Santander
Bank of China
UBS
BNP Paribas
Deutsche Bank
HSBC
Barclays
Banco Santander
UBS
Bank of America
JP Morgan
Bank of China
ICBC
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 10 30 40 50 6 0 70 80 90 100
Correlation
Foreign Asset/Total Asset
2010
2009
2008
Source: Bloomberg; Milken Institute
20
The eurozone
?nancial crisis
91
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
According to EBA (2011), the amount of Greek sovereign debt outstanding is
EUR82.7 billion, of which59per cent of this amount is heldbyGreekbanks as of December
2010 (Table IV). The amount outstanding for Ireland and Portugal is EUR15.8 and
37.6 billion, of which 64 and 50 per cent is held domestically, respectively. (Comparable
data for the individual banks in our sample are provided in Appendix Table AIII).
We make use of these newly released exposure data on a gross basis to assess the
extent to which bank riskiness correlates with Irish, Greek and Portuguese sovereign
risk. As shown in Table IV, although the majority of the PIIGS sovereign debt is held
by domestic banks, the proportion that is held by banks in other European countries is
not a negligible amount, particularly for banks in Germany and France. The positive
relationship shown in Figure 13 shows that bank risk becomes more highly correlated
with the sovereign risk for those banks with greater sovereign exposure to these three
countries, demonstrating market perceptions of greater risk of banks that are more
exposed to this sovereign debt[5].
V. Conclusion
Based upon our empirical results, allowing banks to invest in sovereign debt without
requiring them to hold any capital against the “true” risk of such debt increases
the likelihood of insolvency. This means that interdependencies between bank
Sovereign exposure by counterparty country (as of
December 2010), percentage of total debt outstanding
Greece Ireland Italy Portugal Spain
Participating banks in
Austria 0.56 0.33 0.42 0.36 0.08
Belgium 4.72 1.71 7.19 5.55 1.08
Cyprus 7.03 2.29 0.01 0.00 0.02
Denmark 0.09 2.47 0.15 0.16 0.05
Finland 0.00 0.25 0.00 0.00 0.00
France 11.57 8.77 14.35 10.88 3.51
Germany 9.24 5.99 11.48 9.18 6.45
Greece 58.50 0.11 0.04 0.00 0.00
Ireland 0.05 64.34 0.30 0.65 0.13
Italy 1.66 1.09 55.53 0.95 1.12
Luxembourg 0.10 0.00 0.83 0.48 0.06
Malta 0.01 0.04 0.00 0.01 0.00
The Netherlands 1.41 2.75 2.87 2.06 0.78
Portugal 1.71 3.28 0.36 50.37 0.10
Slovenia 0.02 0.09 0.03 0.04 0.01
Spain 0.54 0.50 2.31 14.44 84.07
Sweden 0.15 0.01 0.13 0.35 0.06
UK 2.62 5.97 4.00 4.53 2.49
Total (91 European banks, billion euros) 82.7 15.8 286.2 37.6 264.4
Notes: Data are net direct positions (gross exposures (long) net of cash short position of sovereign
debt to other counterparties only where there is maturity matching) of 91 European banks (summed up
by country) to PIIGS sovereign debt; the italicized numbers indicates sovereign debt held by domestic
banks
Source: European Banking Authority; Milken Institute
Table IV.
PIIGS sovereign exposure
by counterparty country
JFEP
4,1
92
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
and sovereign risk are extremely important when setting regulatory capital
requirements and considering whether action is needed to limit any increase in the
likelihood of contagion.
Notes
1. As Bolton and Jeanne (2011, p. 10) point out:
The European Central Bank modi?ed its rules and declared that Greek bonds would remain
eligible as collateral even with junk status, before announcing a policy of supporting the
price of certain government debts through open market purchases.
2. In addition, a relatively weak relationship is likely to be due to our bank sample selection
which includes only large banks in each country.
3. We also performed a similar analysis for foreign revenue diversi?cation.
The conclusions are very similar to those for foreign asset diversi?cation described in
this subsection.
4. We use the data from the second stress test which became available to the public in July 2011
(the data are as of December 2010). The ?rst stress test was completed in the summer of
2010.
5. Although this relationship is not surprising, it must be interpreted with caution as this
positive link could be driven by relatively the few banks which have sovereign debt holdings
of these three countries in their portfolios. The data for bank exposure to European
sovereign debt are, unfortunately, limited for non-European banks. We assume that
non-European banks’ direct exposure to the peripheral European debts is relatively minor.
Since the majority of European cross-border banking is concentrated within the European
region, this is a reasonable assumption.
Figure 13.
Impact of bank exposure
to Irish, Greek and
Portuguese sovereign debt
on the bank-sovereign risk
correlation
y = 0.0981x – 0.0244
R
2
= 0.4848
–1
–0.8
–0.6
–0.4
–0.2
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10 12
Correlations
Log (Exposure)
Notes: The vertical axis is the correlation between domestic bank
CDS spreads and Greek (Irish or Portuguese) sovereign CDS spreads;
the correlation is computed annually from October 2008-August 2011
Source: Bloomberg; Milken Institute
The eurozone
?nancial crisis
93
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
References
Acharya, V., Drechsler, I. and Schnabl, P. (2010), “A pyrrhic victory? Bank bailouts and
sovereign credit”, Working Paper No. 17136, National Bureau of Economic Research,
Cambridge, MA.
Allen, F., Beck, T., Carletti, E., Lane, P., Schoenmaker, D. and Wagner, W. (2011), Cross-border
Banking in Europe: Implications for Financial Stability and Macroeconomic Policies, Centre
for Economic Policy Research, London.
Barth, J., Li, T. and Prabhavivadhana, A. (2011), “Greece’s ‘unpleasant arithmetic’: containing the
threat to the global economy”, Global Economy Journal, Vol. 11 No. 4, pp. 1-13.
Bolton, P. and Jeanne, O. (2011), “Sovereign default risk and bank fragility in ?nancially
integrated economies”, Working Paper No. 16899, National Bureau of Economic Research,
Cambridge, MA.
CGFS-BIS (2011), “The impact of sovereign credit risk on bank funding conditions”, Committee
on the Global Financial System Papers No. 43, Bank for International Settlements, Basel.
EBA (2011), EU-wide Stress Test Aggregate Report, European Banking Authority, London.
Gerlach, S., Schulz, A. and Wolff, G.B. (2010), “Banking and sovereign risk in the euro area”,
Discussion Paper No. 9, Deutsche Bundesbank, Frankfurt am Main.
Goldman Sachs Global Economics (2010), “Rising funding costs in the periphery”, European
Weekly Analyst Research Report No. 10/13, Goldman Sachs Global Economics,
New York, NY, November 4.
Mody, A. (2009), “From Bear Stearns to Anglo Irish: how eurozone sovereign spreads related to
?nancial sector vulnerability”, Working Paper No. 10, International Monetary Fund,
Washington, DC.
Reinhart, C. and Rogoff, K. (2008), “Banking crises: an equal opportunity menace”,
Working Paper No. 14587, National Bureau of Economic Research, Cambridge, MA.
Wall Street Journal (2011), “Raters fail to see defaults coming: history shows ?rms rarely
anticipate sovereign failures”, August 12 (online.wsj.com).
About the authors
James R. Barth, PhD, is the Senior Finance Fellow at the Milken Institute, as well as the Lowder
Eminent Scholar in Finance at Auburn University.
Apanard (Penny) Prabha, PhD, is an Economist at the Milken Institute.
Apanard (Penny) Prabha is the corresponding author and can be contacted at:
[email protected]
Greg Yun was a former intern at the Milken Institute and currently is an undergraduate
Economics major at the University of Chicago.
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
JFEP
4,1
94
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Appendix
Bank name Asset to GDP (%) Total assets ($US billions)
Australia
National Australia Bank 53.5 662
Westpac Banking Corporation 48.2 597
Commonwealth Bank of Australia 44.2 547
Australia and New Zealand Banking Group 41.5 514
China
Industrial & Commercial Bank of China 34.7 2,042
Bank of China 27.0 1,587
Agricultural Bank of China 26.7 1,569
France
BNP Paribas 104.2 2,671
Cre´dit Agricole 83.1 2,130
Socie´te´ Ge´ne´rale 59.0 1,513
Natixis 23.9 612
Germany
Deutsche Bank 77.5 2,547
Commerzbank 30.7 1,008
IKB Deutsche Industriebank 1.5 48
Greece
National Bank of Greece 52.8 161
EFG Eurobank Ergasias 38.2 117
Ireland
Bank of Ireland 52.8 161
Allied Irish Banks 38.2 117
Italy
UniCredit 60.5 1,242
Intesa Sanpaolo 42.8 881
Banca Monte dei Paschi di Siena 15.9 327
Banco Popolare 8.8 181
Japan
Bank of Tokyo-Mitsubishi UFJ 34.4 1,876
Mizuho Financial Group 30.6 1,672
Sumitomo Mitsui Banking Corp 24.1 1,318
Resona Bank 8.0 436
Portugal
Banco Comercial Portugueˆs 58.3 134
Banco Espirito Santo 48.8 112
Banco BPI 26.6 61
Spain
Banco Santander 115.4 1,627
Banco Bilbao Vizcaya Argentaria 52.4 739
Banco Popular Espanol 12.3 174
Banco de Sabadell 9.2 130
Switzerland
UBS 267.2 1,411
(continued)
Table AI.
Banks and countries
included in sample (data
as of year-end 2010)
The eurozone
?nancial crisis
95
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Bank name Asset to GDP (%) Total assets ($US billions)
Credit Suisse Group 209.3 1,105
UK
Barclays Bank 103.2 2,323
Royal Bank of Scotland Group 100.7 2,266
Lloyds Banking Group 68.7 1,546
HSBC Bank 55.6 1,250
Standard Chartered 23.0 517
USA
Bank of America Corporation 15.6 2,265
JP Morgan Chase & Co. 14.6 2,118
Citigroup 13.2 1,914
Wells Fargo & Company 8.7 1,258
Notes: UK’s largest traded bank is the HSBC holding companies, but the CDS data are available for
the HSBC bank. Japan’s largest traded bank is Mitsubishi UFJ Financial group, but CDS data are
available for Bank of Tokyo-Mitsubishi UFJ the core retail and commercial banking arm of the
Mitsubishi UFJ Financial Group
Source: Bloomberg; World Economic Outlook, International Monetary Fund; Milken Institute Table AI.
All All PIIGS Others
Change bank CDS 0.157
* *
(0.011) 0.157
* *
(0.011) 0.179
*
(0.086) 0.090
* *
(0.013)
Crisis/pre-bailout 0.066 (0.224) 0.005 (0.809) 0.026 (0.159)
Bailout 2.074
* * *
(0.000) 2.968
* * *
(0.000) 1.478
* * *
(0.001)
Post-bailout 0.220
* *
(0.038) 0.267 (0.193) 0.025 (0.394)
Eurozone sovereign
dept in the second
half of 2010 0.467
* * *
(0.007) 0.915
* *
(0.028) 0.076 (0.108)
Eurozone sovereign
dept 2011 0.617
* *
(0.027) 1.275
*
(0.086) 0.114
* *
(0.015)
Constant 0.043
* * *
(0.000) 20.352
* *
(0.013) 0.934
* * *
(0.007) 20.051
* *
(0.037)
Country dummies Yes Yes Yes Yes
Bank dummies Yes Yes Yes Yes
No. of observation 60,436 60,436 24,529 35,907
R
2
0.044 0.045 0.047 0.060
Notes: Signi?cance at:
*
10,
* *
5 and
* * *
1 per cent levels. The numbers in parentheses are p-values,
based on the robust standard errors clustered at the country level. The dependent variable is change in
sovereign CDS spreads. The regressions are based on the sample of banks and countries listed in the
Appendix Table AI (All, all sample; PIIGS, Greece, Ireland, Italy, Portugal and Spain; others, the rest of
countries in the sample). Crisis/pre-bailout, bailout, etc. refer to a dummy which is assigned a value of
one in each speci?ed period (see note in Table III). The dummy of the pre-crisis period is excluded to
avoid the perfect multicollinearity
Table AII.
Impact of bank risk
on sovereign risk
JFEP
4,1
96
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
E
x
p
o
s
u
r
e
t
o
P
I
I
G
S
c
o
u
n
t
r
i
e
s
s
o
v
e
r
e
i
g
n
d
e
b
t
(
a
s
o
f
D
e
c
e
m
b
e
r
2
0
1
0
)
i
n
m
i
l
l
i
o
n
e
u
r
o
s
C
o
u
n
t
r
y
B
a
n
k
s
P
o
r
t
u
g
a
l
I
r
e
l
a
n
d
I
t
a
l
y
G
r
e
e
c
e
S
p
a
i
n
P
I
I
G
S
F
r
a
n
c
e
B
N
P
P
a
r
i
b
a
s
2
,
0
3
3
4
9
7
2
4
,
1
1
4
4
,
9
9
6
3
,
9
0
0
3
5
,
5
4
0
C
r
e ´
d
i
t
A
g
r
i
c
o
l
e
1
,
1
0
9
1
3
5
1
0
,
1
2
3
6
5
5
2
,
7
7
2
1
4
,
7
9
4
S
o
c
i
e ´
t
e ´
G
e ´
n
e ´
r
a
l
e
6
3
1
4
4
2
3
,
3
4
1
2
,
6
5
1
2
,
2
2
0
9
,
2
8
5
G
e
r
m
a
n
y
D
e
u
t
s
c
h
e
B
a
n
k
8
6
4
7
7
5
,
3
3
6
1
,
5
1
0
2
,
0
8
2
9
,
4
9
1
C
o
m
m
e
r
z
b
a
n
k
9
7
6
2
6
1
0
,
1
3
2
3
,
0
4
3
3
,
1
6
5
1
7
,
3
4
2
G
r
e
e
c
e
E
F
G
E
u
r
o
b
a
n
k
E
r
g
a
s
i
a
s
0
0
1
0
0
8
,
7
4
0
0
8
,
8
4
0
N
a
t
i
o
n
a
l
B
a
n
k
o
f
G
r
e
e
c
e
0
1
8
0
1
2
,
8
8
3
0
1
2
,
9
0
1
I
r
e
l
a
n
d
A
l
l
i
e
d
I
r
i
s
h
B
a
n
k
s
2
4
3
5
,
0
4
3
8
1
6
4
0
3
3
5
6
,
4
7
7
B
a
n
k
o
f
I
r
e
l
a
n
d
0
3
,
2
6
9
3
0
0
0
3
,
2
9
9
I
t
a
l
y
I
n
t
e
s
a
S
a
n
p
a
o
l
o
7
0
1
1
4
5
7
,
6
2
2
6
1
9
7
6
4
5
9
,
1
8
9
U
n
i
C
r
e
d
i
t
8
4
5
8
4
7
,
4
4
5
6
3
7
1
,
9
1
0
5
0
,
1
3
4
B
a
n
c
a
M
o
n
t
e
d
e
i
P
a
s
c
h
i
d
i
S
i
e
n
a
2
0
2
0
3
2
,
0
1
8
8
2
8
4
3
2
,
5
1
2
B
a
n
c
o
P
o
p
o
l
a
r
e
0
0
1
1
,
7
5
9
8
7
0
1
1
,
8
4
5
P
o
r
t
u
g
a
l
B
a
n
c
o
C
o
m
e
r
c
i
a
l
P
o
r
t
u
g
u
e ˆ
s
5
,
8
2
9
2
1
3
5
0
7
2
7
0
6
,
8
1
8
B
a
n
c
o
E
s
p
i
r
i
t
o
S
a
n
t
o
2
,
6
8
3
0
0
3
0
9
5
5
3
,
0
4
8
B
a
n
c
o
B
P
I
3
,
8
9
6
2
8
3
9
7
2
3
2
5
0
5
,
4
7
5
S
p
a
i
n
B
a
n
c
o
S
a
n
t
a
n
d
e
r
3
,
6
3
2
0
2
6
1
1
7
7
4
1
,
8
0
7
4
5
,
8
7
7
B
a
n
c
o
B
i
l
b
a
o
V
i
z
c
a
y
a
A
r
g
e
n
t
a
r
i
a
6
4
7
0
3
,
8
9
7
1
2
7
5
3
,
4
5
2
5
8
,
1
2
3
B
a
n
c
o
P
o
p
u
l
a
r
E
s
p
a
n
o
l
6
4
3
0
2
1
0
0
8
,
8
7
4
9
,
7
2
7
B
a
n
c
o
d
e
S
a
b
a
d
e
l
l
9
1
3
8
0
0
7
,
2
9
6
7
,
4
2
5
U
K
R
o
y
a
l
B
a
n
k
o
f
S
c
o
t
l
a
n
d
G
r
o
u
p
2
0
8
4
0
2
4
,
6
5
4
1
,
1
5
5
3
7
9
6
,
7
9
8
H
S
B
C
B
a
n
k
3
2
0
1
3
4
3
,
8
5
6
9
1
9
6
3
6
5
,
8
6
5
B
a
r
c
l
a
y
s
B
a
n
k
1
,
1
7
4
4
0
7
2
,
9
1
5
9
3
5
,
4
9
6
1
0
,
0
8
5
L
l
o
y
d
s
B
a
n
k
i
n
g
G
r
o
u
p
0
0
3
2
0
6
2
9
4
S
o
u
r
c
e
:
E
u
r
o
p
e
a
n
B
a
n
k
i
n
g
A
u
t
h
o
r
i
t
y
Table AIII.
PIIGS sovereign exposure
by European bank
The eurozone
?nancial crisis
97
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Maria Cantero-Saiz, Sergio Sanfilippo-Azofra, Begoña Torre-Olmo, Carlos López-Gutiérrez. 2014.
Sovereign Risk And The Bank Lending Channel In Europe. Journal of International Money and Finance
. [CrossRef]
2. Markus K. Brunnermeier, Martin OehmkeBubbles, Financial Crises, and Systemic Risk 1221-1288.
[CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
3
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_922225459.pdf