7/31/2019 95418609 Risks and Repercussions of a Greek Exit
1/43
CROSS ASSET RESEARCH 31 May 20
PLEASE SEE ANALYST CERTIFICATIONS AND IMPORTANT DISCLOSURES STARTING AFTER PAGE 39
* This research report has been prepared in whole or in part by equity research analysts basedoutside the US who are not registered/qualified as research analysts with FINRA.
EURO THEMES
Risks and repercussions of a Greek exit
A cost/benefit analysis. We think rationality will prevail and Greece will remain inthe euro area in the near term. The costs of an immediate Greek exit are still too
high for either Greece or the euro area. A disorderly exit would lead to a massive run
on bank deposits, a meltdown of the Greek banking system, and further aggravation
of Greeces large economic downturn. For the euro area, the main cost would be
contagion (see Will Greece abandon the euro?).
A Greek exit can still be avoided. Our base case is that a Greek exit can be avoidedeven if the left-wing Syriza party wins the elections on 17 June and forms a new
government. Its leader has indicated that he would prefer to stay in the eurozone,provided he can negotiate a radical overhaul of the EU-IMF programme. We believe
a programme could be agreed with some concessions by troika (EU, IMF, ECB) on a
smoother fiscal consolidation path and some delays to the speed of public sector
reform (see Will Greece abandon the euro?).
However, a disorderly exit cannot be ruled out. An "accident" after the June electionscannot be ruled out, leading to a disorderly exit for Greece and elevated stress for
the euro area as a result of contagion. The direct costs of a Greek default and exit
appear manageable (euro area government exposure to Greek public debt is
estimated at EUR290bn), but contagion risk could complicate the already precarious
financial positions of countries like Portugal and Ireland, and the systemically more
significant economies of Spain and Italy. The overall costs of a Greek event would
amount to multiples of direct costs (see If Greece exits, can contagion be contained?).
The euro area policy reaction to a Greek exit would likely entail: (a) stepping upthe Securities Markets Programme (SMP), potentially through a leveraged European
Financial Stability Facility (EFSF) or through the European Stability Mechanism
(ESM); (b) aggressive monetary policy action by the ECB; (c) moves towards a pan-
European deposit insurance scheme, although implementation would be lengthy;
(d) if all else fails, acceleration of EU economic and financial integration, including
support for the eventual adoption of Eurobonds.
Implications for European banks. Greek banks have lost almost a quarter of theirdeposit base over the past two years, despite having a deposit guarantee scheme,
suggesting that such schemes are not sufficient to deal with either systemic risks or
redenomination risks. We argue that even a European-wide deposit guarantee
scheme would not be sufficient to manage contagion risks (see Deposit risks).
Implications for rates and FX markets.In the event of a Greek exit, money marketswould suffer a renewed bout of stress, with Italy and Spain bonds likely selling off
aggressively, despite any policy responses that would be introduced (see A Greece
exit is not fully priced into rate markets). Global currencies would follow the pattern
since the Greek election on 6 May, only it would be much more severe: the EUR
would depreciate against the USD, JPY and GBP in particular (see EUR/USD likely to
fall independent of the Greek election result).
Contents
Macro View:
Will Greece abandon the euro? Contagion Costs:If Greece exits, can contagion be contained?
Implications for European Banks:Deposit risks
Implications for Interest Rates Markets:A Greece exit is not fully priced into ratemarkets
Implications for FX Markets:EUR/USD likely to fall independent of the Greekelection result
Economics
Antonio Garcia Pascual
+44 (0)20 3134 6225
Piero Ghezzi
+44 (0)20 3134 2190
Thomas Harjes
+49 69 716 11825
Fabrice Montagne
+33 (0) 1 4458 3236
Asset Allocation StrategyMichael Gavin
+1 212 412 5915
Credit Research
Jonathan Glionna
+44 (0)20 3555 1992
Equity Research
Simon Samuels*
+44 (0)20 3134 3364
Barclays, London
Rates Strategy
Laurent Fransolet
+44 (0)20 7773 8385
Foreign Exchange Strategy
Paul Robinson
+44 (0)20 7773 0903
www.barcap.com
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]7/31/2019 95418609 Risks and Repercussions of a Greek Exit
2/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 2
MACRO VIEW
Will Greece abandon the euro?
The costs of Greece exiting the euro are too high for either Greece or the euro area. A
disorderly exit would lead to a massive run on bank deposits, a meltdown of the Greekbanking system, and further aggravation of Greeces large economic downturn. For the
euro area, the main cost would be contagion.
Greece New government may reject the EU-IMF programme
The national elections in Greece on 3 May marked a turning point in the European political
landscape. For the first time, and in stark contrast to earlier national elections in Ireland and
Portugal, Greek political parties committed to policies agreed under an EU-IMF programme
failed to achieve a majority in Parliament. Support for the two main parties, New Democracy
(centre right) and Pasok (centre left), which have governed Greece for decades and
negotiated the recent EU-IMF programme, fell below even modest expectations. At the
same time, the left-wing Syriza party and more radical (left and right) parties that reject the
policies under the EU-IMF programme gained a significant share of the votes (see, Election
round-up: A vote for policy change in Europe). After failed attempts by the leaders of New
Democracy (18.9% of the vote), Syriza (16.8%), and Pasok (13.2%) to negotiate a coalition,
and by the Greek President to form another technocrat government with broad parliamentary
support, Greece is heading for another round of elections scheduled on 17 June.
It is difficult to predict the outcome of these elections, but the latest opinion polls suggest
that New Democracy is currently leading. However, Syriza, led by the young and charismatic
Alexis Tsipras, is not far behind New Democracy and could become the largest party. If this
were to happen, it would automatically get an extra 50 seats in the Greek Parliament, and
may amass enough support to form a government. Mr Tsipras has been adamant that he
would cancel austerity policies in Greece, reverse plans to reduce public employment, and
cancel interest payments and debt redemptions. (He has also said that he would prefer that
Greece remain in the eurozone, but only if the EU-IMF programme is radically overhauled.)
Other parties, including ND and Pasok, have stepped up their rethoric and are seeking to
cast the next elections as a vote on the future of Greece in the euro area. Polls still suggest
broad support amongst the Greek population for remaining in the euro area. However, the
elections on 3 May have shown that frustration and anger about past ND and Pasok policies
prevailed when Greeks cast their votes: they delivered a protest vote that may also be
interpreted as a popular rejection of the EU-IMF adjustment programme that is associated,
rightly or wrongly, with the miserable economic and social conditions that now prevail.
Developments within Greece are now in the hands of the Greek electorate; the countrys
continued membership in the European Monetary System hinges upon decisions that will betaken in the 17 June elections, and the response of other European governments to that
political outcome. But these political processes are being driven in large part by economic
developments, as they have been interpreted by politicians and electorates. Without
underplaying the importance of the elections, fundamentals suggest that Greece may remain
a concern for financial markets, regardless of who wins. The main difference the election
could make is whether Greeces day of reckoning happens very soon or somewhat later.
Antonio Garcia Pascual
+44 (0)20 3134 [email protected]
Michael Gavin
+1 212 412 5915
Piero Ghezzi
+44 (0)20 3134 2190
Thomas Harjes
+49 69 716 11825
Fabrice Montagne
+33 (0) 1 4458 3236
https://live.barcap.com/go/publications/content?contentPubID=FC1819189https://live.barcap.com/go/publications/content?contentPubID=FC1819189mailto:[email protected]:[email protected]:[email protected]:[email protected]://live.barcap.com/go/publications/content?contentPubID=FC1819189https://live.barcap.com/go/publications/content?contentPubID=FC18191897/31/2019 95418609 Risks and Repercussions of a Greek Exit
3/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 3
A grim economic outlook within the euro zone
2011 marked the fourth year of Greeces recession, in which real GDP has already fallen
nearly 15%. Unfortunately, the wrenching external and fiscal adjustments that are putting
intense pressure on the economy are only partly accomplished, and the end of the
downturn does not appear to be imminent.
Public finances remain far from equilibriumGreece ended 2011 with a primary (non-interest) deficit of 2.4% of GDP, and an overall
budget deficit of 9.1%. Even after the 2011 restructuring of privately-held public debt, the
debt is expected to reach nearly 164% of GDP by year-end, and is likely to continue rising in
2013 due to the shrinking economy and a primary deficit. (The still-high amount of debt is
due, in part, to the high cost of recapitalizing the Greek banks, which will add over 25% of
GDP to public debt in 2012 as a result of the large losses of Greek banks on their holdings of
government debt.) Under the existing programme, Greece will need to engineer a large
swing from deficit into a primary surplus of over 4% of GDP (a swing of roughly 6
percentage points of GDP), in order to stabilize its high level of public debt.
Adjustment of external payments is far from complete
Greeces current account deficit has been shrinking since 2008, but remains untenably highat roughly 10% of GDP. With no capacity to promote expenditure switching via
devaluation, the current account adjustment has been associated with a painful
compression of public and private spending. The adjustment would of course have been far
more abrupt if eurozone governments had not stepped in to finance the external payments
deficits that market participants are no longer willing to finance. The result has been
intensified dependence on official financing of external payments deficits. By now, the lions
share of the countrys international liabilities is official funding (c.EUR313bn), including
the EU (EUR126bn), the IMF (EUR22bn), and the Eurosystem (EUR165bn).
Banks remain under severe stress
As a result of the sharp and sustained drop in economic activity, non-performing bank loans
have risen to about 15%. More importantly, the large losses imposed by the February
restructuring of their government bond holdings have left banks with a large capital
deficiency. The new programme has set aside an additional EUR50bn for bank
recapitalization, which will (as noted above) be added to the public debt.
Greek debt is expected to reach
nearly 164% of GDP by year-end
Figure 1: Current account adjustment A work in progress Figure 2: Deposit outflows have intensified funding squeeze
-40
-35
-30
-25-20
-15
-10
-5
0
Dec-97 Dec-99 Dec-01 Dec-03 Dec-05 Dec-07 Dec-09 Dec-11
Current account balance (12-mo total, bn euros)
0
50
100
150
200
250
300
Jan-01 Jan-03 Jan-05 Jan-07 Jan-09 Jan-11
Domestic residents deposits and repos (bn euros)
Source: Haver Analytics Source: Haver Analytics
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
4/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 4
Moreover, the loss of capital markets access has left the central bank of Greece (through
the Eurosystem) as practically the only funding channel for Greek credit institutions. Most
of the outstanding funding is now provided through the Emergency Liquidity Assistance
facility of the Eurosystem (c.EUR100bn of the total c.EUR130bn provided by the
Eurosystem). The Greek banks funding squeeze has been intensified by a persistent outflow
of bank deposits since the crisis started in late 2009. From the end of 2009 to March 2012,
the banking system lost nearly a third of its domestic deposits. Anecdotal evidence indicatesthat the outflow accelerated in May, especially after the election results, and it may continue
given the elevated uncertainty and downside risks that confront domestic depositors.
A grim outlook
With large fiscal and external imbalances that have yet to be fully unwound, the immediate
outlook for the Greek economy remains grim. The economy is likely to shrink by more than
5% in 2012, the 5th consecutive year of contraction, while the rate of unemployment is
approaching 20%. Because the fiscal and external adjustment will likely last into 2013, we
(and the IMF) are forecasting yet another year of recession in 2013. Moreover, the public debt overhang is unlikely to be resolved in the immediate future. Under
the February 2012 programme, public debt is projected to peak at 167% of GDP in 2013,
but then rapidly decline to reach 116% of GDP by 2020 and 88% of GDP by 2030. However,
this downward trajectory is explained by programme targets that include a primary surplus
of 4.5% of GDP by 2014, real growth in the range of 2.5-3% over the medium term, and
privatization revenues of EUR45bn, which are spread over several years. We consider these
assumptions too optimistic. Under a more realistic baseline macroeconomic scenario, the
Greek public debt would stabilize, but at an extremely high level (Figure 3).
Figure 3: Debt sustainability not addressed by the February PSI (gross public debt/GDP)
0
20
40
60
80
100
120
140
160
180
200
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Base Debt relief (zero interest on all official debt for 20y)
Source: Barclays Research
Stabilization of the debt at such a high level satisfies one narrow and theoretical definition of
sustainability, but it would leave the Greek government in a precarious position; highly
exposed to fiscal shocks, and reliant upon official funding for decades to come. Figure 3
also illustrates how one form of debt relief, a forgiveness of all interest on public debt for 20
years, could create a more robustly sustainable trajectory for the public debt. But because
the public debt remains very high for a long period of time (above 125% of GDP at least
until 2020), even this form of debt relief may not be sufficient to restore confidence in
public financial stability over a reasonable period of time.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
5/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 5
Box 1: Debt sustainability analysis: medium-term assumptions
Our medium term baseline scenario includes a less aggressive fiscal consolidation path than
the February 2012 programme. Realistically, Greece could achieve a fiscal consolidation of
about 1% of GDP per year over the next five years, to achieve a primary surplus of about
2.5% of GDP by 2016. This is in stark contrast to the medium-term assumptions in the
programme, which expects Greece to achieve a primary surplus of 4.5% of GDP by 2014and sustain it at around that level through 2020.
Our fiscal baseline assumes that the moderate parties (New Democracy and PASOK) form a
government and an agreement on the revised programme can be reached. In any event, the
new government is highly likely to renegotiate a less aggressive fiscal consolidation targets.
Obviously, if the radical left parties were to form a government after the 17 June election, it
is likely than the pace of fiscal consolidation would be even less aggressive. In that scenario,
the fiscal path assumed in the baseline would have to be revised downwards and,
consequently, the debt dynamics would look even more challenging.
Our growth assumptions also differ from the EU-IMF programme, as we expect real GDP to
shrink by 6% instead of 4.8% in the programme. Macroeconomic conditions have
deteriorated, mainly as a result of the uncertainty about a potential exit. Private
consumption and investment will contract further than originally envisaged, as well as
exports. We expect economic activity to recover only very gradually over the next 5 years to
reach a real GDP growth of 1.5% instead of 3% assumed under the programme, which also
assumes that real GDP will remain above 2% through 2020 (while we assume a growth of
1.5% on average).
The debt dynamics would look very different if Greece were to restructure or default on the
outstanding debt, but over half of that debt are official loans from the EU and IMF (see
Figure 4), and much of the remainder is held in the domestic banking system and by the
ECB. As we have seen, debt held by the domestic banking system cannot be written down
without breaking the banks and requiring a government-financed bank recapitalization. And
official creditors including the ECB have so far resisted a restructuring of the Greekobligations that they hold.
Figure 4: Holdings of Greek public debt
Before PSI Post PSI, end 2012
IMF loans 20 28
EU loan package 1 53 74
EU loan package 2 0 88
ECB SMP + Investment portfolio 55 46
T-bills 15 15
Sub total 1 143 251
Banks (o/w Greek banks c.EUR40bn) 70 (40) 22 (13)
Insurance 10 3
Central banks/official institutions 38 12
Other investors (real money, etc) 70 22
Greek Social Security fund 18 6
Sub total 2 206 0
New exchange bonds from PSI 0 65
Total Greek debt 349 316
Source: Barclays Research
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
6/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 6
In short, the economic outlook over the next two years is cloudy. No immediate end of the
recession is in sight, fiscal adjustment is far from complete, and the unresolved debt
overhang means that Greece may ultimately need to obtain official sector debt relief (in NPV
terms) or leave the eurozone.
A Greek exit would be very costly for Greece
The only possible justification for a policy framework that delivers a prognosis of this sort is
that outcomes would be even worse under alternative policies. If the alternative policy is
rejecting the EU-IMF programme and exiting the eurozone, a strong case can be made that
the costs would be higher than the benefits, at least in the short run.
In the short term, costs of exit would likely be very high compared with benefits
A Greek exit in the near term would be disorderly almost by definition, as this would likely
be triggered by a failure of a Greek government to reach a compromise with the rest of the
euro area on a programme. The EU and the IMF would then stop funding Greece, and this
would (most likely) imply that the ECB could not continue the provision of emergency
liquidity support to Greek banks. While this decision might be taken by the Governing
Council of the Eurosystem (by at least a 2/3 majority), given the serious implications, it willbe essential to have the political backing of the European Council and the IMF. Indeed, once
Greece is locked out of the Eurosystem without an EU-IMF programme, it would quickly run
out of cash and be forced to exit the EMU. Expulsion from the EU does not appear to be a
legally viable option and Greece could decide to remain within the EU.
Nevertheless, Greece would be going it alone from a position of economic and financial
weakness, not strength. The curtailment of official financing of the government and the
balance of payments would enforce a more abrupt adjustment of both imbalances. Even if it
is accompanied by a cessation of public debt service, the pace of fiscal adjustment in such a
scenario would need to be even larger than contemplated by the existing programme
(though it might be facilitated by a burst of inflation over the medium term). The sudden
cut in government expenditure would likely aggravate social unrest, which could lead to a
government crisis. If history serves as any guidance for the Greek case, few governments
have survived traumatic exchange rate devaluations.
Potentially even more disruptive for economic activity could be a massive run on bank
deposits: depositors would run on the banks as they would fear a forced conversion in a
reintroduced drachma. In the absence of Eurosystems support and without bank and
capital controls to limit the outflows, the exit would lead to a meltdown of the Greek
banking system, further aggravating the large economic downturn, quite possibly forcing
massive government intervention in the banking system.
The reintroduced drachma would likely depreciate significantly and hence many local
companies (clearly those in the non-tradable sector) and households would need to default on
their foreign currency debt, now including euro-denominated liabilities. Many of the domesticcontracts that are now denominated in euros would also become unviable and need to be
restructured. Non-performing loans would surge because of: 1) the negative balance sheet
effects for firms and households; and 2) the local currency needed to pay euro debts would
increase with the devaluation, exceeding the increase in local currency revenue. Likewise, the
government would also be forced to default on its euro-denominated liabilities.
Redenomination away from the euro will also cause massive transfers between agents,
adding to the above-mentioned transfers between debtors and creditors. Households with
local accounts and savings will suffer substantial losses while cash-rich agents with
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
7/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 7
accounts abroad will be the big winners and could take advantage of the chaos to seize
capital and production capacities. Given the weak state of the government, these
redistributions will likely benefit to the already oversized unofficial sector.
In short, the existing contracting framework and financial infrastructure would be broken,
and would need to be rebuilt. Inflationary finance would likely be used, to some extent at
least, to replace the official finance that now supports Greece. Politically difficult fiscal andstructural reforms would still be required to make the country more competitive, and
promote economic growth.
Over the longer term, exit by Greece would have some offsetting benefits
While the costs of exit seem likely to be very high in the short run, there would of course be
some long-term benefits, as well as costs. Devaluation would provide a faster route to
international price competitiveness than the grinding internal devaluation that will be
required within the eurozone. The IMF estimates the need for a 15% improvement in cost
competitiveness on a unit-labor-cost basis. International evidence suggests that internal
devaluations, that is, improvements in competitiveness that are achieved by reductions in
domestic wages and prices rather than a devaluation of the nominal exchange rate, are hard
to achieve. In the Greek case, the difficulty is compounded by inflexible labor and productmarkets, by the public debt overhang (which is aggravated by slow growth or even deflation
in the nominal tax base), and because other stressed European economies are also trying to
engineer similar improvements in competitiveness, making it all the more difficult to achieve
the required improvement with neighbours who are important trading partners. However, if
higher inflation triggers higher wages, the positive effects of large devaluations on
competitiveness are very limited.
Like other participating countries, Greece entered the eurozone because it offered important
long-term advantages created by integration into a continental monetary and financial
system with commonly governed EU institutions. By leaving the eurozone, Greece would
not only suffer enormous short-term dislocation, but it would also forego the longer term
structural advantages of membership, especially if an exit from EU would follow.
Greek exit would also be very costly for the euro area
We consider the potential direct and indirect costs for Europe of a Greek exit in considerably
more detail in subsequent sections. Here we extract the elements of that discussion that are
relevant for the eurozones decision whether or not to take actions that might precipitate a
Greek departure from the monetary area.
Direct links appear manageable
The direct trade and financial links of Greece with the rest of the euro area are meaningful,
but the direct economic fallout from a Greek exit appears manageable given the size of the
Greek economy (only about 2% of eurozone GDP). Most of the worlds financial exposure to
Greece has been shifted to public balance sheets. The exposure of the euro area officialsector is somewhat above EUR290bn, and very sizable in absolute terms in France (about
EUR60bn) and Germany (about EUR80bn); losses would also be politically very sensitive.
Recovery rates on this exposure would probably be low, but even a total loss would be
manageable from a fiscal policy and debt sustainability perspective; in fact, most of the
existing sovereign exposure is already included in gross debt figures. Assuming final
recovery rates in the range of 20-50%, the loss, or involuntary transfer to Greece from other
member states and the ECB could roughly amount to EUR150-230bn, or 1-2.5% of euro
area GDP (see Greece: Euro area official sector exposures in excess of EUR290bn).
https://live.barcap.com/go/publications/content?contentPubID=FC1823228https://live.barcap.com/go/publications/content?contentPubID=FC18232287/31/2019 95418609 Risks and Repercussions of a Greek Exit
8/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 8
Liabilities of the Greek private sector to the rest of the world are considerably smaller, at
about EUR130bn, not all of which is owed to European creditors. In the event of a Greek exit
from the eurozone, many of these liabilities would probably need to be written down
dramatically. The main concern is not the overall size of the potential loss, but the possibility
that it might be concentrated enough to destabilize a systemically significant financial
institution. The French banking systems exposure is the largest, but even in this case the
systems exposure to the Greek private sector is less than 20% of core tier one capital.Accidents do happen, and it is possible that an isolated institution may experience enough
trouble to create some market reaction. But it seems unlikely that this exposure will create a
large economic problem for Europe.
The larger risk for Europe is financial contagion
We consider the risks and potential policy responses to contagion in much more detail
below. For now, we simply note that there are a number of channels through which a
disorderly Greek exit could compound the difficulties that already face other stressed
eurozone sovereigns such as Ireland, Portugal and the systemically far more important
economies of Italy and Spain, and ultimately the entire continent. Italy and Spain are already
facing very shaky confidence and a withdrawal of private international capital. The costs for
Europe of a Greek exit under these circumstances are quite literally incalculable, because
they depend in large part upon psychological responses to the exit. But they are potentially
enormous. This gives European policymakers a strong incentive to avoid a Greek
abandonment of the EU-IMF programme and exit from the eurozone.
From economics to politics
Greeces future is now being determined by very public political processes, in which
electorates (and their representatives) rather than technocrats will have the final word. This
is obvious in Greece, given the recent focus on the election, but it is increasingly true in the
broader European context. There are no easy or appealing alternatives for the main actors in
this drama. Greek voters face a choice between a painful future within the eurozone, or a
future outside the eurozone that is very likely much more painful in the short term, with anuncertain balance of costs and benefits down the road.
We could explore these costs and benefits in more detail, but it would probably take us
away from the underlying political dynamic. What matters is the assessment reached by the
political actors in Europe, and these have by now, we think, hardened into an essential
political reality. Within Greece, public opinion polls suggest that a large majority of the
Greek population strongly supports continued membership in the eurozone. However, the
May elections seemed also to signal that the population is reaching some limits of its
tolerance for austerity, and the economic misery with which this is rightly or wrongly
associated, but which the countrys partners in the EU insist is a necessary condition for
continued membership in the monetary union. One way to interpret these apparently
conflicting opinions is that the Greek population is committed to membership in theeurozone, but not under any conditions whatsoever.
Other countries in the eurozone appear to be edging toward a similar perspective. Given the
economic duress under which the Greek population is labouring, and the potentially high
cost for Europe of a Greek exit, we suspect that there is more room for negotiation about
the programme than some recent statements may suggest. However, that room is not
unlimited. Greeces official creditors have already put enormous quantities of their
taxpayers wealth at risk, and have no mandate from their voters to extend a blank check
for financial support in unlimited amounts over an indefinite horizon. Moreover, the EU-IMF
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
9/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 9
programme is organized around the view that the structural reforms and institution-
building elements of the reform agenda are necessary conditions to regain competitiveness
and financial stability, and thus to thrive within the monetary union over the longer term.
How far the EU and IMF might be willing to alter the existing economic programme is highly
speculative, and may depend upon political events in the coming weeks (such as the
upcoming French parliamentary elections). Our thoughts at the moment are as follows:
The EU has signalled that it is willing to introduce pro-growth initiatives in the existingadjustment programmes. This could include investments in infrastructure, some active
labour market policies (targeted at reducing youth unemployment), and possibly some
limited support for health services. The problem is that such initiatives are not easy to
come up with (if they were, what government would tolerate slow growth?), and
seldom come without a price tag. Our best guess is that pro-growth initiatives and
social policies may provide a Greek government with political ammunition to defend a
revised agreement that it finds otherwise acceptable, but will not alone transform Greek
attitudes toward the memorandum.
The EU and IMF may well be willing to offer some leeway in the pace of fiscalconsolidation. Tolerating a slower adjustment would not be costless, from theperspective of the EU. Even without such concessions, the existing programme is likely
to require additional EU financing given the likelihood of a worse-than-projected
economic growth outlook since February, as financial turmoil takes its likely toll on the
economy. Any delays in the fiscal consolidation will, of course, create a yet larger
financing gap that would need to be filled by taxpayers in other eurozone countries.
It would be impossible to alter the long-run fiscal targets without addressing the publicdebt problem that dictates the long-run fiscal effort that is arithmetically required to
ensure eventual fiscal solvency. In fact, as explained above, we think there are reasons
to doubt that solvency can be re-established without substantial debt relief; any delays
in the pace of fiscal consolidation would only compound the already untenable
overhang of public debt. The subject of debt relief is likely to be a theme of discussionbetween the new government and the EU, IMF, and ECB. In a larger sense, it will have to
be related to the question of fiscal transfers within the EA.
If there is anything non-negotiable for the EU and IMF, we think it would be theproposed structural reforms. Without them, they believe that Greeces prospects inside
the monetary union would be bleak. Reforms are needed to regain competitiveness and
thus lay a firm foundation for economic growth over the long term, and to build a public
sector that is efficient and sustainable without continued injections of foreign resources.
However, these reforms touch upon politically sensitive areas of Greek society.
Can the minimum requirements for a programme be reconciled with the maximum policy
effort that a new Greek government can sustain? We think that is heavily dependent upon
the outcome of the 17 June election. We see three main potential outcomes:
The first and in some ways the worst possibility is that, as in the first election, no party orcoalition of parties manages to secure a mandate to govern. In this case, the political
deadlock would have to be resolved by yet another round of elections. Uncertainty would
persist, and the potentially destabilizing flight from the Greek banking system could
accelerate further. While the election process plays out, we think it would be difficult for
eurozone governments to interrupt the financial support for the Greek banking system that is
required to keep Greece in the eurozone, but the escalation of contingent financial liabilities
associated with this support would also be difficult for the ECB to countenance indefinitely.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
10/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 10
A second possibility is that a coalition of the radical left, led by Syriza, wins the electionand forms a government. In this case, negotiations between the troika and government
could be lengthy (possibly several months). Given the economic programme upon
which Syriza has been campaigning, we think it is entirely possible that discussions
could end without agreement on a programme. The most contentious themes would
likely be the delay in fiscal consolidation targets and cuts to the size of the public sector,
including dismissals of public sector employees. But our baseline scenario is that eventhen, a programme could be agreed with some concessions by troika on a smoother
fiscal consolidation path and some delays to the speed of public sector reform.
The third of the most plausible scenarios is that the center-right New Democracy partywins the election and forms a government in coalition with other traditional and centrist
parties. Given that these are the political parties who negotiated the existing
programme, it is reasonable to think that their election would facilitate negotiations over
another revision of the programme. The question would then become whether the
government, likely surviving with a razor-thin parliamentary majority and enjoying weak
support from the public, would be able to implement the agreed programme more
effectively than it was able to implement similar programmes in the past.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
11/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 11
CONTAGION COSTS
If Greece exits, can contagion be contained?
The direct costs of a Greek default and exit appear manageable, but contagion risk
could complicate the already precarious financial positions of countries like Portugaland Ireland, and the systemically more significant economies of Spain and Italy. The
overall costs of a Greek event could be orders of magnitude larger than the direct
costs.
Direct costs of a Greek exit are not big enough to destabilize therest of the eurozone
The direct exposure to Greece for euro area countries is non-negligible: about EUR290bn
of exposure to the Greek government, with roughly one third of this amount in exposure to
private-sector entities. However, it appears to be manageable even if Greece were to
default and recovery rates were low. In any event, some credit losses from a restructuring
of Greek liabilities to the rest of the eurozone will likely be realized even if Greece remains
part of the monetary union. The exposure to the Greek public sector that eurozone
governments have accumulated now amounts to roughly 3% of GDP.
Figure 1: Official exposure to Greece in EMU by country and type
Member States Eurosystem Total
Bilateral
loans
EFSF
garantees SMP Target 2 , Bn
% of
GDP
Nominal
GDP (2011,
, Bn)
Austria 1.6 2.2 1.0 3.9 8.7 2.9 301
Belgium 1.9 2.7 1.3 4.9 10.8 2.9 368
Cyprus 0.1 0.2 0.1 0.3 0.6 3.4 18
Estonia 0.0 0.2 0.1 0.4 0.7 4.5 16
Finland 1.0 1.4 0.7 2.5 5.6 2.9 192
France 11.4 15.9 7.6 28.3 63.3 3.2 2002
Germany 15.2 21.2 10.2 37.8 84.5 3.3 2571
Greece 0.0 0.0 0.0 0.0 0.0 0.0 215
Ireland 0.3 0.0 0.0 0.0 0.3 0.2 156
Italy 10.0 14.0 6.7 25.0 55.8 3.5 1580
Luxembourg 0.1 0.2 0.1 0.3 0.8 1.8 43
Malta 0.1 0.1 0.0 0.1 0.3 4.5 6
Netherlands 3.2 4.5 2.1 8.0 17.8 3.0 602
Portugal 1.1 0.0 0.0 0.0 1.1 0.6 171
Slovakia 0.0 0.8 0.4 1.4 2.6 3.8 69
Slovenia 0.2 0.3 0.2 0.6 1.3 3.7 36
Spain 6.7 9.3 4.4 16.5 36.9 3.4 1073
Total 52.9 73.0 35.0 130.0 291.1 3.1 9419
Note: * IR, PT and GR stepped out.Exposures are reported in nominal amounts. For SMP, EFSF and Target 2, national exposure have been allocatedaccording to capital keys.Source: European commission, EFSF, National central banks, Barclays Research
The eurozones exposure to the Greek private sector is substantially smaller than its
exposure to the Greek government. Much of this exposure is in the form of bank claims
(EUR70bn). But even in France, the most heavily exposed of the larger European banking
systems (EUR34bn), the exposure does not seem large enough to pose a systemic threat.
Antonio Garcia Pascual
+44 (0)20 3134 6225
Michael Gavin
+1 212 412 5915
Piero Ghezzi
+44 (0)20 3134 2190
Thomas Harjes
+49 69 716 11825
Fabrice Montagne
+33 (0) 1 4458 3236
mailto:[email protected]:[email protected]:[email protected]:[email protected]7/31/2019 95418609 Risks and Repercussions of a Greek Exit
12/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 12
Figure 2: Foreign Bank Claims on Greece
End-December 2011 European banks France Germany Italy Spain Switzerland UK US
Foreign claims 90,473 44,353 13,355 2,186 969 1,940 10,537 4,455
Public sector 21,929 6,502 6,749 773 302 168 1,759 725
Banks 3,167 223 759 146 39 476 1,005 672
Non-bank private sector 65,295 37,628 5,847 1,266 627 1,256 7,772 3,058
Other potential exposures 29,032 6,901 3,779 1,790 417 1,406 11,825 46,231
Derivatives contracts 6,548 1,545 470 462 172 505 2,953 907
Guarantees extended 15,835 3,692 2,818 280 24 155 8,199 44,831
Credit commitments 6,649 1,664 491 1,048 221 746 673 493
Source: Bank of International Settlements, valuations at EUR prices at end-December 2011
In short, the direct losses from a Greek exit seem unlikely to destabilize the rest of the
eurozone, even if they lead to widespread restructuring of claims on the Greek private
sector. That said, the losses may undermine political support for financial programmes in
other eurozone economies. In particular, ECB losses from its Greek exposure would further
undermine support for the SMP within the ECB. Political backlash could complicateattempts to mount a determined effort to support other stressed sovereigns in the months
following a Greek exit.
The question is contagion
The direct costs of a Greek default and exit appear manageable, but the bigger concern is
the potential for contagion, which could complicate the already precarious financial
positions of countries like Portugal and Ireland, and the systemically more significant
economies of Spain and Italy. If developments in Greece create even deeper anxiety about
the rest of the euro area, the overall costs of a Greek event could be orders of magnitude
larger than the direct costs.
Contagion is not inevitableIt is possible that this contagion will be more limited than many now fear. After all,
Greece restructured its debt in March, while markets rallied in Q1 2012. More than a
decade ago, Argentina defaulted and suffered a historic economic and financial
collapse, but contrary to anxiety about contagion, other emerging markets were, in the
end, little affected. It is not inevitable that a Greek exit from the eurozone will trigger
strong contagion to other stressed sovereigns.
However, we would caution against complacency. The Greek debt restructuring and a
potential exit from the eurozone are very different shocks to the economic, financial, and
political systems. First, abandoning the euro and re-denominating contracts into a currency
that does not yet exist is much more complex than a public-sector debt restructuring, for
which precedents abound. It is not only more complex but also more unpredictable in itsconsequences, many of which will be indirect. A Greek exit will likely result in private, as well
as public defaults, and the corresponding losses may be very difficult to predict. It also has
potentially greater negative value as a precedent. Government debt restructurings happen
all the time, but even partial breakup of a monetary union is very rare. If Greek becomes the
first departure from the eurozone, it will be hard to avoid the question whos next?
Second, we think that the main reason markets did not react negatively to Greeces
restructuring in March was because the default was fully priced when it happened the issue
of Greek default contagion affected peripheral debt markets throughout the second half of
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
13/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 13
2011 and hence had minimal residual effect in 2012. The same cannot be said of a Greek exit
from the eurozone: it is not priced in yet, especially not in key markets. Third, the Greek default
occurred in a context in which the ECB had surprised markets by effectively launching the 3-
year LTROs.
Stressed eurozone sovereigns are already suffering a confidence crisis
Whether financial instability that may result from a Greek exit is transmitted to other eurozoneeconomies depends, ultimately, upon market participants reaction to events; whether, for
example, depositors in Spain view turmoil in Greece as a risk scenario for themselves that they
had not adequately considered, or as further evidence that the situation in Greece and in Spain
are fundamentally different. While it is not possible to predict shifts in sentiment of this nature,
the risk of an adverse market reaction is heightened by the fact that stressed sovereigns in
Europe are already suffering a form of confidence crisis, concentrated to some degree in
foreign owners of public debt, but which could certainly become more generalized and intense
in the future. While there is no guarantee that a disorderly Greek exit would trigger a broader
and more intense run against vulnerable eurozone financial systems, it would be dangerously
complacent to rule this out.
As the situation in Europe has become more complicated, international investors inEuropean government bond (EGB) markets have either sold their exposures or failed to roll
over investments as they come due. In Spain, for example, we estimate that non-resident
investors (other than the ECB) had by April of this year cut their exposure by 50% from early
2010 levels (Spain: Banks still buying, foreigners still selling (but less), 30 April 2012). In
recent months, non-resident selling of Spanish and Italian assets has accelerated, and
seems to have broadened somewhat (Figures 5 and 6).
Thus far, the capital outflows have been accommodated without crisis in Spain and Italy,
thanks to the ability of the banking systems in those countries to tap the liquidity assistance
of the eurosystem. The SMP has provided direct support for government bond markets,
while the LTRO operations provided long-term funding to reduce liquidity pressures
confronting banks, and to permit banks in Spain and Italy to act as government bond buyersof last resort.
Figure 3: Capital outflows have been replaced by eurosystemfunding in Spain
Figure 4: and in Italy
-150
-100
-50
0
50
100
150
200
Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Jan-10 Jan-12
BOP, financial account, with Banco de Espana
ex BdE
-150
-100
-50
0
50
100
150
200
Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Jan-10 Jan-12
BOP, financial account, with Banca d'Italia
Ex-Banca d'Italia
Source: Haver Analytics Source: Haver Analytics
https://live.barcap.com/go/publications/content?contentPubID=FC1816302https://live.barcap.com/go/publications/content?contentPubID=FC18163027/31/2019 95418609 Risks and Repercussions of a Greek Exit
14/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 14
Figure 5: EGB investor are leaving the periphery (holdings by non residents, in %)
10%
20%
30%
40%
50%
60%
70%
80%
Sep-07 Jul-08 Jun-09 May-10 Apr-11 Mar-12
Greece
Portugal
Ireland
Italy
Spain
Source: Barclays Research
If foreign selling were confined to government debt, the remaining exposure would be finite
and, somewhat manageable. For example, if foreign bondholders sold 25% of their remaining
holdings of Italian, Irish, Portuguese and Spanish government debt, other (presumably official)
sources would need to be tapped in the amount of roughly EUR250bn.
This is a big number, but one that lies within the plausible limits of the eurozones capacity.
The numbers become much larger if the run extends beyond government bond markets.
Problems would be compounded dramatically if domestic investors also lose confidence
and sell domestic assets. In particular, if a Greek exit creates fears of loss among bank
depositors in other eurozone economies, a run against bank deposits could be triggered.
We have not observed bank deposit runs in the periphery thus far (other than Greece). In
fact the evidence shows that deposits are rather sticky even in Greece, where nearly two-thirds of the bank deposits have (astonishingly) not left despite the longstanding risk of a
potential disorderly exit. Deposits left Ireland in the months leading up to the EU-IMF
programme of November 2010, but have stabilized since then. We have not seen evidence
of any material depositors outflows in Portugal. In Spain, there is evidence of flight to
quality away from weak cajas but they have moved into stronger domestic banks. And in
Italy there is no material evidence of deposits outflows other than large multinationals that
had to diversify away from countries/banks with low ratings.
The fact that there has so far been no run against bank deposits is not, of course, proof that
it cannot happen in the future.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
15/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 15
Figure 6: Deposit liabilities of Euro Area MFIs
Germany France Italy Spain Greece Ireland Portugal
2009Dec 2,875,710 1,606,738 1,209,695 1,686,023 244,237 219,180 209,649
2010Mar 2,874,507 1,612,590 1,212,938 1,651,301 233,395 220,266 209,149
2010Jun 2,923,056 1,653,307 1,401,379 1,665,951 223,555 217,320 206,744
2010Sep 2,924,448 1,685,502 1,381,604 1,701,127 219,503 211,656 213,366
2010Dec 2,972,771 1,733,500 1,424,330 1,730,187 215,639 201,067 226,165
2011Mar 2,968,338 1,766,883 1,386,424 1,727,934 205,510 195,858 227,349
2011Jun 3,023,207 1,792,549 1,389,411 1,741,354 195,069 199,895 230,695
2011Sep 3,071,737 1,848,503 1,396,599 1,716,762 189,375 198,933 231,078
2011Dec 3,090,965 1,870,912 1,386,613 1,679,558 180,085 196,298 232,805
2012Mar 3,094,532 1,906,210 1,414,413 1,655,018 171,088 195,529 229,054
Source: ECB
If it does, there is much more at stake (even) than in the government bond markets. In Italy
and Spain alone, bank deposits total more than EUR3trn. If those banking systems were to
lose a quarter of their deposits (for the sake of illustration), alternative financing in the
amount of more than EUR750bn would need to be found.Policy responses
The most immediate policy response would in all likelihood make use of emergency liquidity
tools by the ECB, including:
A modified SMP programme, which would buy EGBs across different maturities andcountries. The SMP could be modified to explicitly give up its seniority status. Alternatively
(and perhaps politically more acceptable in Germany), the EFSF/ESM could be granted a
banking license to have access to the ECB and hence leverage its resources. The EFSF/ESM
would then play a similar role to that of the ECB. The intervention by the ECB or the
EFSF/ESM would have to be signalled as unlimited (in quantity and time) for it to be effective.
The ECB is expected to extend its full-allotment policy in the June meeting (currentlyexpiring in July), but if a crisis were to unfold in June/July, the ECB could consider a
further extension of full-allotment into 2013 and offer another round (or rounds) of very
long-term refinancing operations.
Cuts to the ECBs main policy rate to near the zero bound and a clear indication that thepolicy rate will stay at (near) zero through 2013.
A declaration of the European Council concerning a roadmap to fiscal union includingthe first steps (crisis related) as well as the longer-term issues (transfer of sovereignty,
Eurobonds, etc).
Other near-term crisis management tools that could be deployed relatively quickly include
and could complement ECB liquidity injection:
Changes to the status of the EFSF/ESM to allow it to directly inject equity intoundercapitalized European banks. Access to EFSF/ESM could put a ceiling to the debt
(and future losses) that the government would have to take. Above that, the ESM could
assume any further losses. The funds should have attached an EC-ECB-IMF programme
with conditionality exclusively focused on bank restructuring and recapitalization (ie to
satisfy Germany request for conditionality on EU help). By resolving weak European
banks and the potential downside that they put on some European sovereigns, including
Spain (and other countries), it would reduce an important source of uncertainty.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
16/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 16
Additional powers to the European Financial Stability Facility (EFSF)/European StabilityMechamism (ESM) as a precursor of a pan-European deposit insurance scheme.
However, this is far more complex and less likely in our view than the previous policy
response. Ideally, the common deposit insurance scheme could also be endowed with
common supervisory and bank resolution powers. This is technically complex and may
have fiscal implications that would be resisted by several countries. It would require
treaty changes and would need to be voted by all parliaments (just like the current ESM,which is yet to be approved by most countries).
However, all these measures may not be enough in the event of a real loss of confidence
owing to impending euro break-up risk. At that point, the only policy response that could
aid ECB emergency measures to stem the crisis is accelerating EU economic and financial
integration, including an explicit and credible support by all EMU leaders of the future
adoption of Eurobonds.
Yet any form of Eurobonds (joint liability) will also face an uphill legal and political battle
and are unlikely to be approved in the near term. Germany has repeatedly indicated that
more joint-liability requires common European institutions with decision-making authority
for fiscal policy in member states, but France and other EMU member states have so far
resisted the latter. However, if the crisis intensifies to the point of becoming a real threat to
the EMU itself, euro area leaders would be forced to agree and credibly communicate to the
markets what the end-game is for the European integration project.
Conclusion
Greece, and the euro area, are at a cross roads again. The outcome of the 17 June elections
remain uncertain and the rhetoric by some of the left parties suggest that, if elected, they
would aim for negotiations to secure a deep revision of the current EU-IMF programme. If
no agreement is found, the government would lose access to EU-IMF funds and without
them, Greek banks would quickly become insolvent; this would in turn lock out their access
to eurosystems liquidity. Liquidity would dry up very quickly, and a meltdown of the
banking system and, very likely, a disorderly euro area exit would follow.
Even if an accident of this sort is avoided in the near term, Greece's future in the EMU is not
cast in stone. While the Greek government has achieved a substantial fiscal consolidation,
the EU-IMF programme is evidently not delivering results that the Greek population
considers acceptable. In fact, even after the February debt restructuring, the public debt
dynamics remain unsustainable and further debt relief will be required. Poor
implementation of the programme has undoubtedly played a role, but this does not change
the fact that disillusionment with the EU-IMF strategy, combined with continued political
resistance by those who stand to lose privileges under the structural reforms demanded by
the 'memorandum', means that even if a new programme is agreed with the new
government, there is a non-negligible chance that the programme would go off-track again.
In the event of a disorderly exit, the costs for the euro area could be very large: in fact the
interconnectedness of the euro area markets makes the potential costs literally incalculable.
Contagion is already at work, as the EGB markets attest. A Greek exit could leave the door
open for speculation that other EMU members may follow, and this perception would in all
likelihood condition the near and long-term investment decisions of domestic and foreign
investors. In the extreme, it could possibly trigger a (systemic) bank run in some of the
periphery countries, even if there is no evidence of the latter thus far.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
17/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 17
We think that such a disorderly outcome can be avoided. But solutions that only entail
liquidity provision, while necessary, may not be sufficient to address the core problem.
Instead, a credible commitment to greater political and fiscal integration may also be
needed. In particular, a credible commitment with steps towards greater integration that
end with the adoption of Eurobonds may be what is required. There have been specific
proposals even from core countries. Those have been rejected by Germany but may get
traction again if the crisis becomes more acute. The debt redemption fund, as proposed bythe German Council of Economic Experts, could compress sovereign debt spreads
significantly while keeping average debt yields at relatively low levels. That would be the
price that the core may need to pay if it wants to save the euro.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
18/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 18
IMPLICATIONS FOR EUROPEAN BANKS
Deposit risks
Greek banks have lost almost a quarter of their deposit base over the past two years,despite having a deposit guarantee scheme (DGS), suggesting that such schemes arenot sufficient to deal with either systemic risks or redenomination risks.
Facing a similar challenge a decade ago, deposits in Argentina fell 20%, eventuallyresulting in a freeze on withdrawals. Whilst to date there is little evidence of deposit
outflows in other periphery countries (Spain, Italy, Ireland, Portugal), talk of a possible
exit of Greece from the European monetary union has sparked fears of contagion.
We view Europes current deposit insurance system as inadequate, suffering fromtwo drawbacks; first, the insurance is currently provided at a national level by risky
local entities; second, it does not protect against currency redenomination. If
redenomination risk is fully removed via a Euro DGS, then any impact from
redenomination is simply transferred from depositors to the guarantee scheme.
Whilst this may be affordable vis--vis redenomination fears in Greece, Ireland and
Portugal, it would lack credibility in a worst case scenario of contagion spreading to
Spain and Italy as well. Furthermore, the unintended consequences are likely
substantial: raising the costs of providing the guarantee and potentially
undermining a recovery upon euro exit.
All of this suggests that whilst a Euro DGS may initially sound appealing, it isprobably an inappropriate tool to be used in isolation to manage down the contagion
risks from a Greek exit.
Whats happened in Greece?
Before considering the issue of wider contagion of a Greek exit to Europes deposit markets,it is worth examining the situation in Greece itself. Greek banks have been losing deposits
for two years as concerns escalate regarding Greeces economic and fiscal prospects, the
health of its banks, and its potential for long-term inclusion in the European Monetary
Union. As shown in Figures 1 and 2, Greek deposits have declined approximately 24% to
171bn in the past two years. This outflow has been largely offset by the provision of
exceptional liquidity from the European Central Bank and Bank of Greece. This liquidity
provision has allowed the banks to remain adequately funded, but Greek banks have been
experiencing a steady deposit run that could accelerate as the June elections approach.
Most developed banking systems have deposit insurance schemes to offset the inherent
mismatch of making long-term loans funded with deposits that must be repaid on demand.
These schemes allow depositors to place their funds at a bank with confidence that they will
be returned when demanded, regardless of the solvency of the bank. Deposit insurance
exists in Greece, but it has not stopped the outflow of funds from the banking system.
Jonathan Glionna
European Banks Credit Analyst+44 (0)203 555 1992
Miguel Angel Hernandez, CFA
European Banks Credit Analyst
+44 (0)20 7773 7241
Simon Samuels*
European Banks Equity Analyst
+44 (0)20 3134 3364
[email protected], London
Mike Harrison*
European Banks Equity Analyst
+44(0)20 3134 3056
Barclays, London
Nimish Rajkotia*
European Banks Equity Analyst
+44 (0)20 3134 3719
Barclays, London
* This research report has been
prepared in whole or in part by
equity research analysts based
outside the US who are not
registered/qualified as research
analysts with FINRA.
mailto:[email protected]:[email protected]:[email protected]:[email protected]7/31/2019 95418609 Risks and Repercussions of a Greek Exit
19/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 19
Figure 1: Greek customer deposits outstanding, bn Figure 2: Q/Q changes in Greek household deposits, mn
0
50
100
150
200
250
Jun-10 Oct-10 Feb-11 Jun-11 Oct-11 Feb-12
-24%
-12,000
-8,000
-4,000
0
4,000
8,000
12,000
16,000
20,000
Q1 09 Q3 09 Q1 10 Q3 10 Q1 11 Q3 11
Deposits in non-resident banks
Cash and sight deposits in domestic banks
Cash burn
Total household deposits
Source: ECB, Barclays Research Source: ECB, Barclays Research
Contagion: Little evidence to dateHowever, perhaps the key area of focus for investors is that of contagion risk as talk of a
possible exit of Greece from the European monetary union has sparked fears about deposit
outflows from other peripheral countries. However, these concerns are not new. The bank
funding pressures prevailing prior to the 3-year LTROs led investors to closely monitor
customer deposit flows in peripheral countries in the second half of 2011.
Customer deposit balances declined across most of the European periphery during this
period, although the declines were moderate outside of Greece (Figure 3). Irish deposits fell
just 1% in the last six months of 2011, having declined more than 10% since mid-2010.
Spanish and Italian deposits fell 3% and 2%, respectively, while Portuguese deposits
actually rose. In the first three months of 2012, customer deposits remained stable or even
increased (Italy), as investor confidence recovered following the 3-year LTROs. As shown inFigure 4, aggregate deposits in Spain, Italy, Portugal, and Ireland have increased in 2012. We
find little evidence of a trend of recent large-scale deposit outflows from peripheral
European countries, excluding Greece.
Figure 3: Change in customer deposits since June 2011 Figure 4: Total deposits in Spain, Italy, Portugal, and Ireland, bn
-5%
-4%
-3%
-2%
-1%0%
1%
2%
3%
4%
Jun-11 Aug-11 Oct-11 Dec-11 Feb-12
Spain Italy Ireland Portugal
2,780
2,800
2,820
2,840
2,860
2,880
2,900
2,920
2,940
2,960
Jun-10 Oct-10 Feb-11 Jun-11 Oct-11 Feb-12
Note for both charts: Excludes government deposits and securitization bonds. Source for both charts: classified as deposits. Source: ECB, Barclays Research
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
20/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 20
In addition, we believe that looking solely at customer deposits does not provide the full
picture. This is because banks in some peripheral countries, most notably Spain and Italy,
have focused their fund-gathering efforts on placing bonds, usually short-dated
instruments, with their retail customers. In the case of Spain, where this phenomenon is
relatively new, this is in response to the governments decision to increase banks
contributions to the countrys deposit guarantee scheme (the contribution is calculated as a
percentage of deposits outstanding). In Italy, retail bonds have historically been animportant source of funding for banks because of the tax advantage they entail for the end
customer relative to traditional deposit products. Figure 5 shows the change in customer
funding since June 2011, which we proxy by the aggregate of customer deposits and debt
outstanding with a maturity of less than two years. These data confirm that, despite some
volatility during the period, customer funding of peripheral banking systems, excluding
Greece, has been relatively stable since June 2011.
Figure 5: Change in the aggregate of customer deposits and short-term debt outstanding(
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
21/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 21
coverage. To offset this, the remodelled directive (2009/14/EC) was published in 2009. This
sought to increase harmonisation and required member states to ensure a level of coverage
that was fixed at 100,000 by the end of 2010. Some of the additional features of the most
recent directive are summarised in Figure 6.
Figure 6: European deposit insurance directive harmonized basic terms
Right to compensation Scope of claim How much? When? Who funds the scheme?
All deposits held bynon-bankinginstitutions ie, individuals, smallcorporations,partnerships
All deposits-mainly demand,term and savingsdeposits, plusregistered savingscertificates
100,000
Payout no laterthan 21 daysafter funds arefrozen
Various. Some/all ofthe schemes funded bymember financialinstitutions throughannual proportionatecontributions
Source: European Commission, Barclays Research
In addition, individual countries also took action to solidify deposit insurance schemes
during the early stages of the crisis. In Ireland, for example, the statutory Deposit Guarantee
Scheme (DGS) was supplemented by the Eligible Liabilities Guarantee (ELG) scheme
initiated in October 2010. The first 100,000 of an individuals deposit is guaranteed by the
DGS, and the remaining balance over 100,000 (no limit) may be covered by the ELGScheme or Government Guarantee. The Irish ELG scheme is currently scheduled to expire
on June 30, 2012. However, the date can be extended further until December 31, 2012 with
EU state aid approval.
Despite the increased harmonization, a challenge exists in that the guarantees are still
provided locally, by governments and agencies that have credit risk, reducing the value of
the insurance. For example, in Greece deposit insurance is provided by the Hellenic Deposit
Guarantee and Investment Guarantee Fund (HDIGF). In Spain and Italy, the deposit
guarantee is provided by Fondos de Garantia de Depositos and Fonda Interbancaria di Titela
dei Depositi, respectively. A list of the deposit insurance provider in various countries is
shown in Figure 7. In most cases, the creditworthiness of these agencies is only as good as
the creditworthiness of the aggregate banking system or sovereign.
Figure 7: Deposit guarantees are still provided locally
Selected country Who is responsible for depositor insurance?
Belgium Protection Fund for Deposits and Financial Instruments
France Deposit Guarantee Fund (Fonds de Garantie des Depots-FGD)
Germany Compensation Scheme of German Banks(Entschdigungseinrichtung deutscher Banken- EdB) andadditional Deposit Protection Fund for Private Commercial Banks.Separate schemes exist for Savings and Cooperative Banks
Greece Hellenic Deposit Guarantee and Investment Guarantee Fund -HDIGF
Ireland Central Bank of Ireland
Italy Interbank Deposit Protection Fund (Fonda Interbancaria di Titeladei Depositi - FITD)
Portugal Deposit Guarantee Fund (Fundo de Garantia de Depositos- FGD)
Spain Deposit Guarantee Fund for Banking Institutions (Fondos deGarantia de Depositos - FGD)
UK Financial Services Compensation Scheme - FSCS
Source: European Commission and individual member state DGS websites
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
22/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 22
Most deposit insurance schemes are funded independently by the banking system,
although they often include a direct or implied guarantee from the government. Some of
these schemes are financed by regular contributions from the member banks annually to
build up the ex ante float in preparation for a future crisis. Others do not pre-finance at all
and require the member banks to fund the required payout ex post, after a bank has
failed. Greeces deposit insurance law states that if the funds in the Deposit Cover Scheme
were not sufficient to meet depositor compensation payments, the HDIGF may, at itsdiscretion, borrow from the credit institutions participating in the scheme for a determined
period of time. In 2009, this may have provided comfort that the fund could be replenished
following a bank failure, but it is less beneficial when the crisis involves the entire banking
system. Generally, deposit insurance funds are prohibited from borrowing from national
central banks or the ECB. In Figure 8, we outline the funding status of selected European
deposit insurance schemes. As shown, the available resources are moderate in relation to
the size of each countrys deposit base. For example, in Spain, a banking system with
approximately 1.6trn of deposits, the deposit insurance fund has total assets of just
7.9bn, of which 5.3bn has already been committed to the resolution of Caja de Ahorros
del Mediterraneo (CAM) banks.
Figure 8: Funding status of selected deposit insurance schemes
Country
Customer
Deposits
bn*
How is deposit insurance
funded? Resources within the fund Can it borrow?
Who provides for the
shortfall?
Does it have
preferred
creditor
status?
Spain 1,655 Credit institutions in theform of annualcontributions of 0.2-0.3% ofeligible deposits
Total assets of 7.9bn as ofOctober 2011, of which4.1bn are liquid; in Q4,5.3bn committed forresolution of CAM
Yes Credit institutions throughone-off contributions,borrowings in open market
No
Italy 1,414 Credit institutions, regularcontributions coveroperating expenses only;
fund interventions coveredby one-off contributions
Unfunded; maximumcontributions in 2012 canbe up to 1.8bn (0.4% of
eligible deposits)
Notcontemplated
Not contemplated No
Portugal 229 Credit institutions in theform of annualcontributions
1.4bn (of which 400mnare unfunded)
Yes Credit institutions throughone-off contributions; Bankof Portugal can also coverimmediate needs ifsystemic stability at risk
Yes
Ireland 196 Annual cash contribution of0.2% of deposit base
NA Yes Central Bank, reimbursedwithin 3mths by Treasury,who then seeks fundreplenishment from thecredit institutions
No
Greece 171 Credit institutions in theform of annual
contributions of 0.9-1.1% ofeligible deposits
1.4bn (end 2010) Yes Supplementarycontributions by credit
institutions capped at 3xregular annualcontribution. Also Fundcan borrow required levelfrom credit institutions
Yes- impliedby new Bank
Resolution plan
* As at May 2012Source: National regulators, Barclays Research
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
23/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 23
Problem 2: Redenomination risks
The second challenge facing any European deposit guarantee system is dealing with
redenomination risk and in particular the contagion risk that could follow a Greek exit from
the eurozone. We believe that the most likely transmission mechanism of contagion is that a
Greek exit leads to bank runs in other peripheral banking systems. The thinking here is that
depositors in the periphery (ex Greece) come to view a Greek exit as a realistic template forhow eurozone sovereign debt crises are resolved. As such, depositors may become
concerned that the end game for their own sovereign's debt crisis is that their deposits are
redenominated and, once in the new currency, depreciate heavily versus the euro. So they
begin to withdraw their deposits, potentially sparking bank runs.
Left unchecked, this could lead to widespread bank insolvencies, requiring national governments
to step in and shore up the banking system. But clearly, placing such an additional burden on
the finances of already-troubled sovereigns would risk exacerbating the current crisis further.
The conventional approach to prevent bank runs is a deposit guarantee scheme (DGS).
However, as noted, the current European system of deposit insurance has important limitations.
Moreover, the extant DGSs have been designed to handle individual rather than systemic bank
runs. Even in the US, it was the Fed/Treasury and not the Federal Deposit Insurance Corporation
(FDIC), which manages the US DGS, who provided the backstop to the systemic crisis in 2008-
09. As such, the current arrangements for protecting depositors against potentially widespread
currency redenomination concerns following a Greek exit are limited.
Argentina: 2001
Perhaps the most recent example of a country facing similar challenges around
denomination risk was Argentina in the period leading up to the abandonment of its
currency peg with the US dollar in 2001. Indeed, the Greek experience of deposit outflow is
not dissimilar. For example, Argentina underwent a period of deposit flight prior to
redenomination during the national economic crisis in the early 2000s, during which declining
output and high inflation strained the countrys economic and political infrastructure. Fearing arevaluation of the currency, Argentineans began to pull deposits out of peso-denominated
accounts, often reposting these funds into dollar-denominated accounts.
Deposit insurance, in place at the time, did little to stem this flow, since the provider of the
guarantee was in question, given the impending government default; the value of the
currency, not just the viability of the banks, was uncertain given the increasingly apparent
need for revaluation. As concerns escalated in late 2001, the pace of this outflow and
conversion increased. In three of the worst days of November 2001, 6% of bank deposits
were pulled from the system.1 Year-over-year, deposits fell 19.9% in 2001, to ARS66.0bn
from ARS82.5bn at the end of 2000, according to IMF data.
In response, the Argentinean government instituted a deposit freeze, or corralito, onDecember 3, 2001 to help stem the overwhelming outflow of deposits. Payments were
initially allowed between banks in the system (though additional restrictions were
subsequently applied), but depositors were not permitted to withdraw the majority of funds
from their accounts. Nevertheless, as the crisis continued, with the governments
declaration of default in late December 2001 and conversion of dollar-denominated
financial assets and liabilities into Argentinean pesos in February 2002, deposit withdrawals
continued at an estimated rate of ARS0.9-4.2bn per month in the first seven months of
1 IMF, Banks during the Argentine Crisis, Barajas et al, 2007.
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
24/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 24
2002.2 A combination of monetary tightening, government support for bank solvency, and
legal protection against deposit devaluation contributed to a reversal of these trends overthe course of 2002. Deposit withdrawals slowed in 3Q02 and eventually started to grow
again in late 2002 (Figures 9 and 10).
Part of the problem with bank runs is that they quickly become self-fulfilling. As Mervyn
King, Governor of the Bank of England, noted during the run on Northern Rock in 2007,
"once a bank run has started, it is rational to join in". In other words, it doesn't matter
whether a bank run starts for valid reasons or not. Once aggressive deposit flight is
underway and a banks solvency is threatened, it makes sense for the remaining depositors
to participate if they think that everyone else will withdraw their deposits as well. This holds
true regardless of whether all depositors share the concerns that initially started the run.
The idea behind a DGS is to break this vicious circle. If you know with confidence that you
(and everyone else) can access your deposits in full, there's no need to withdraw deposits inthe first place. So the key barometer of a successful DGS lies in its credibility. If a DSG can
cope with depositors worst case scenarios, bank runs will be prevented.
However, given the shortcomings of existing national DGSs, are there other ways to reduce
the risk of bank runs? One (limited) approach is as follows. If a bank's liquidity buffer
(including central bank reserves) is generally perceived to be greater than the proportion of
depositors who are concerned about redenomination, banks can cope with deposit
withdrawals without becoming insolvent. Whilst a Greek exit clearly increases the
proportion of depositors concerned about redenomination, any early-stage panic could in
theory be addressed by national authorities in the periphery signalling to the rest of the
depositor base that: (a) the group of 'concerned' depositors is too small to cause a run; and
(b) banks' liquidity buffers are sufficiently large. Measures to reassure depositors about the
size of liquidity buffers could potentially include further special measures from the ECB
(although this may lead to uncomfortable questions about the credit risk they assume) or,
alternatively, the introduction of depositor preference regimes across Europe (although this
would likely further undermine banks' wholesale funding models).
2 World Bank, Argentinas Banking System, Gutierrez and Montes-Negret, 2004.
Figure 9: Argentinean total deposits (bn) Figure 10: Argentinean annualized deposit growth (%)
0
20
40
60
80
100
120
140
160
1993 1995 1997 1999 2001 2003 2005
Deposits (Native Currency) Deposits (SDRs)
-80%
-60%
-40%
-20%
0%
20%
40%
60%
1994 1996 1998 2000 2002 2004
y/y Growth (Native Currency) y/y Growth (SDRs)
Source: IMF International Financial Statistics, Barclays Research Source: IMF International Financial Statistics, Barclays Research
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
25/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 25
Figure 11: Depositor preference laws by country
Country Depositor Preference Law?
United States Yes
United Kingdom No (proposed)
Switzerland Yes (indirect)
Germany Yes (indirect)Holland No
France No
Spain No
Italy No
China Yes
Japan No
Australia Yes
Argentina Yes
Brazil No
Source: FSB, Moodys, Barclays Research
But the short-term efficacy of such an approach faces at least two significant hurdles. First,
it's a complicated solution. In a world where the levels of financial sophistication vary
significantly across the population, it would be extremely challenging to quickly and
effectively communicate a complex idea like depositor preference. Similarly, further ECB
liquidity provisions may soothe financial markets and financially sophisticated corporates,
but they run the risk of bewildering many retail depositors. Would they understand at what
point 'enough' liquidity had been pumped into the systems?
Second, and even more problematic: how could authorities credibly assert that the
proportion of depositors concerned about redenomination is too small to precipitate a bank
run? How could they tell? Quite aside from the challenges presented in gathering such data,
bank runs can occur even if no depositors have redenomination as their base case. Provided
that the switching costs of moving deposits out of 'at risk' banks/geographies are lower
than the perceived probability-weighted impact of redenomination, then deposit
withdrawals can reach the critical mass needed for a bank run.
This suggests that any response that is either too complicated for retail depositors to
understand or requires authorities to have 'private knowledge' that they can't realistically
access will meet with limited success. Another way of putting this is to say that any policy
measure which is applied on a 'doing just enough' basis may well prove ineffective. Would a
more top-down, EU-wide approach to a DGS work any better?
An EU-wide DGS would be helpful in the sense of enhancing the credit-worthiness of the
guarantor but would be unlikely to offer adequate protection against currency
redenomination. This is because even if deposits are guaranteed in euros 'today', this offerslittle relief if redenomination occurs 'tomorrow'. The only way such a DGS could work is if
deposits were guaranteed in euros even after countries have left the euro with the risks of
devaluation being transferred from depositors to the DGS provider. This could result in the
bizarre situation of having (newly) non-eurozone countries having all their domestic
deposits denominated in euros. Would such a scheme be credible?
Given the limited pre-funding of DGSs across Europe, significant deposit outflows in the
periphery would have to be met either from the ECB or from non-periphery sovereign debt
issuance. To give a sense of scale to this contingent liability from a sovereign issuance
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
26/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 26
perspective Figure 12 shows the deposits/GDP across the 27 countries in the EU. We
calibrate the size of the contingent liability of an EU-wide DGS by comparing the size of the
European periphery deposit base (Spain, Greece, Italy, Ireland, Portugal, subsequently
referred to as SGIIP) relative to the GDP of various constituents of the DGS provider.
Figure 12: Customer deposits / GDP
Customer Deposits GDP Deposits / GDP
Luxembourg 220 44 503%
Cyprus 49 18 275%
Malta 11 6 167%
UK 2,777 1,775 156%
Spain 1,655 1,069 155%
Netherlands 860 599 143%
Portugal 229 169 136%
Belgium 480 371 129%
Ireland 196 155 126%
Germany 3,095 2,606 119%
Austria 322 305 106%
France 1,906 2,018 94%
Italy 1,414 1,575 90%
Greece 171 222 77%
Czech 111 157 71%
Denmark 160 241 66%
Finland 128 194 66%
Bulgaria 26 40 65%
Sweden 246 401 61%
Slovenia 22 36 60%
Slovakia 39 70 56%
Estonia 9 16 55%
Poland 187 355 53%
Hungary 46 96 48%
Latvia 8 21 38%
Lithuania 12 32 38%
Romania 44 165 26%
Total 14,419 12,755 113%
Total SGIIP 3,665 3,190 115%
Total Non-SGIIP 10,754 9,565 112%
SGIIP Deposits % Non SGIIP GDP 38%
SGIIP Deposits % AAA GDP 63%
SGIIP Deposits % German GDP 141%
Source: ECB, Bloomberg, Barclays Research
7/31/2019 95418609 Risks and Repercussions of a Greek Exit
27/43
Barclays | Risks and repercussions of a Greek exit
31 May 2012 27
SGIIP deposits are 38% of non-SGIIP GDP. This implies that the deposit/GDP of the EU-wide
DGS providers would rise to 151%. This is relatively high relative to the EU, but on a par with
the UK and Netherlands.
But the real credibility of such a scheme ultimately comes down to how much the
contingent liability would potentially cost the DGS providers and whether this is
affordable. However, this immediately gets us into murky waters. At what point would anEU-wide DGS pay out? Ordinarily, a DGS only makes payments once a bank has exhausted
its reserves and sold down its liquid assets to the point of insolvency. If taxpayer monies are
used to bail out depositors, these would usually be in the form of a loan (rather than a gift) -
with repayment occurring after the banks has been restructured or sold on.
However, it's unlikely that extending these mechanics to an entir