© 2015 International Monetary Fund
IMF Country Report No. 15/98
CENTRAL AND EASTERN EUROPE:
NEW MEMBER STATES (NMS) POLICY FORUM, 2014
SELECTED ISSUES
This Selected Issues Paper on Central and Eastern Europe: New Member States (NMS)
Policy Forum, 2014 was prepared by a staff team of the International Monetary Fund. It is
based on the information available at the time it was completed on February 26, 2015.
Copies of this report are available to the public from
International Monetary Fund Publication Services
PO Box 92780 Washington, D.C. 20090
Telephone: (202) 623-7430 Fax: (202) 623-7201
E-mail: [email protected]
Web: http://www.imf.org
Price: $18.00 per printed copy
International Monetary Fund
Washington, D.C.
April 2015
CENTRAL AND EASTERN EUROPE:
NEW MEMBER STATES (NMS) POLICY FORUM, 2014
SELECTED ISSUES
Approved By
The European
Department
Prepared by: John Bluedorn, Greetje Everaert, Nan Geng,
Anna Ilyina, Plamen Iossifov, Jiri Podpiera, Jesmin Rahman,
John Ralyea, Ara Stepanyan, Yan Sun, Johannes Wiegand,
Jiae Yoo, Jessie Yang and Li Zeng. Min Song provided
research assistance.
EURO ADOPTION—MACROECONOMIC BENEFITS AND CHALLENGES _______________ 4
A. European Integration and Euro Adoption ______________________________________________ 5
B. Advantages from Adopting the Euro ___________________________________________________ 7
C. Advantages from Preserving Monetary Policy Autonomy and Exchange Rate
Flexibility ________________________________________________________________________________ 11
D. The Impact of Euro Adoption on Macroeconomic Policy Frameworks ________________ 15
E. Conclusions ___________________________________________________________________________18
BOXES
1. The 2004 IMF Study, and New Evidence on its Main Findings _________________________ 19
2. New Member States, Euro Adoption, and the Theory of Optimal Currency Areas _____ 20
3. Economic and Price Convergence in Emerging Europe ________________________________ 21
FIGURES
1. Euro Premium, 2001-13 ________________________________________________________________ 9
2. New Member States: Monetary Policy, 2003-07 _______________________________________ 14
3. New Member States: Monetary Policy, 2008-14 _______________________________________ 16
APPENDICES
I. Estimating the Euro Premium __________________________________________________________ 28
II. The Index of Monetary Conditions ____________________________________________________ 33
III. Model-Based Inflation Variance Decomposition ______________________________________ 34
REFERENCES ____________________________________________________________________________22
CONTENTS
February 26, 2015
CEE NEW MEMBER STATES POLICY FORUM
2 INTERNATIONAL MONETARY FUND
OPTING INTO THE BANKING UNION BEFORE EURO ADOPTION ____________________36
A. Why Did Europe Need a Banking Union? _____________________________________________ 37
- ________________________ 39
C. Banking Union Opt-In: Pros and Cons for Non-Euro EU Countries ____________________ 44
D. Conclusions ___________________________________________________________________________ 53
BOXES
1. Key Elements of the Euro Area Banking Union ________________________________________ 54
2. The SSM Modalities ___________________________________________________________________ 55
3. The SRM Modalities ___________________________________________________________________ 56
4. Macroprudential Policy Space for BU members _______________________________________ 57
5. Theoretical Considerations in Designing an Optiomal Banking Union _________________ 58
6. Cross-Country Differences in Policymakers Relative Preference for Promoting
Domestic Banks__________________________________________________________________________ 59
TABLE
1. Benefits and Costs of Joining Banking Union for Non-Euro Area Countries ___________ 47
APPENDIX
I. Largeset Banks in NMS-6 and their Ultimate Owners __________________________________ 60
REFERENCES ____________________________________________________________________________61
THE EU FISCAL FRAMEWORK AND PENSION REFORM _______________________________64
A. Public Pension Systems in New Member States: The Broad Picture ___________________ 65
B. Pillar II Pension Schemes: History, Rationale, Performance ____________________________ 68
C. The EU Fiscal Framework and Pillar II Reversals _______________________________________ 71
D. Conclusions ___________________________________________________________________________ 77
FIGURES
1. Selected European Countries: Key Demographic Data ________________________________ 65
2. Second Pension Pillars and National Savings __________________________________________ 70
3. Pillar II Pension Funds: Returns and Fees ______________________________________________ 72
4. Pillar II Pension Reversals _____________________________________________________________ 75
TABLES
1. Pension Spending Projections _________________________________________________________ 66
2. Net Present Value of Pension Deficits _________________________________________________ 67
3. Treatment of Net Cost of Systemic Pension Reforms in the EU Fiscal Framework (the
Stability and Growth Pact (SGP)) _________________________________________________________73
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 3
APPENDICES
I. Pension Systems in NMS-6 and Recent Pension Reforms ______________________________ 81
II. Evolution of Pillar II Systems and Contribution Rates __________________________________ 84
REFERENCES ____________________________________________________________________________79
MAKING THE MOST OF THE EU SINGLE MARKET ____________________________________85
A. Introduction ___________________________________________________________________________ 86
B. Evolution of Exports to EU: Relative Success and its Determinants ____________________ 87
C. Export Quality in NMS-6: Room for Growth ___________________________________________ 93
D. Services Exports: Scope for Further Increase __________________________________________ 97
E. Policy Implications ____________________________________________________________________ 99
BOX
1. Czech Republic and Hungary: What can be learnt from Korea? _____________________ 101
FIGURES
1. NMS: Gross and Value Added Exports of Goods and Services _________________________ 87
2. Structural Factors: Relative Importance for Exports to the EU _________________________ 91
3. NMS-6: Contribution of Structural Factors to Relative Export Performance in the EU
Market ___________________________________________________________________________________ 92
4. NMS-6: Export Quality and Room for Improvement ___________________________________ 95
5. NMS-6: Services Exports and Services Directive ______________________________________ 98
TABLES
1. Determinants of Value-Added Exports of Goods and Services to EU: NMS-10,
200311 _________________________________________________________________________________ 90
2. RCA: Exports on Professional and Technical Services __________________________________ 99
ANNEX
Data Appendix and Robustness Check for Regression Analysis for Export Integration _ 102
REFERENCES __________________________________________________________________________ 103
CEE NEW MEMBER STATES POLICY FORUM
4 INTERNATIONAL MONETARY FUND
EURO ADOPTION—MACROECONOMIC BENEFITS
AND CHALLENGES
1
1
Prepared by Jiri Podpiera, Johannes Wiegand and Jiae Yoo. Jessie Yang provided excellent research
assistance. The authors are grateful to Csaba Balogh (Hungarian National Bank), Kalin Hrivstov (Bulgarian
National Bank), Paul Kutos (European Commission), Andrzej Raczko (Polish National Bank), and Vedran Sosic
(Croatian National Bank) who participated as discussants in the session on euro adoption at the New
Member States (NMS) Policy Forum in Warsaw on December 12, 2014, and to the NMS-6, EC and ECB
representatives who provided comments during bilateral discussions in November 2014. Ernesto Crivelli,
Anna Ilyina, Plamen Iossifov, Murad Omoev, Andrea Schächter, Michelle Shannon, (other) members of NMS-
6 country teams, and participants at an IMF seminar provided helpful discussions and comments.
2
NMS-6 includes Bulgaria, Croatia, Czech Republic, Hungary, Poland and Romania. NMS includes these
countries and NMS-EA consisting of Estonia, Latvia, Lithuania, Slovakia and Slovenia.
Summary
New Member States (NMS)
2
have considerable leeway over the timing of euro adoption. Even
though the NMS have committed to eventually joining the euro area in their accession treaties, key
steps to initiate adoptionsuch as harmonizing the legal framework with euro area standards, or
applying for ERM2 entryremain under sovereign control. Conversely, the euro area institutions have
substantial discretion in admitting countries to ERM2a pre-condition to euro adoption.
The NMS-6 maintain different monetary regimes. Bulgaria and Croatia have tied their currencies to
the euro, while the Czech Republic, Hungary, Poland, and Romania target inflation and allow exchange
rates to float.
The euro area crisis has reduced some attractions associated with joining the euro. While countries
credit risk, this premium has mostly vanished with the euro crisis. This said, euro adoption continues to
hold advantages for highly euroized economies.
For NMS that have maintained monetary autonomy, this has been helpful in containing macro-
economic imbalances, suggesting that, for them, ceding autonomy could be costly. Monetary
tightening and exchange rate appreciation helped contain credit booms in the mid-2000s. After the
outbreak of the 2008-09 financial crisis, monetary easing supported domestic demand. More recently, it
has helped offset imported disinflationary pressures.
The trade-offs associated with euro adoption present themselves differently for the NMS-6,
depending on their monetary regimes. For floaters, key issues are the extent to which monetary
autonomy can be replaced by instruments such as macro-prudential tools and fiscal policy, and the
scope for internal adjustment in the euro area. For peggers, the question is to what extent they can
continue to use macro-prudential and other regulatory tools after adopting the euro.
For many NMS-6, uncertainty about the euro area’s evolving institutional framework provides a
rationale to wait for final outcomes before taking an irreversible adoption decision.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 5
A. European Integration and Euro Adoption
1. Euro adoption forms the endpoint of monetary integration in the EU.
In their EU accession treaties, the NMS-6 committed to adopting the euro “once the
necessary conditions are fulfilled.
3
Parallel biannual reports by the European
Commission (EC) and the European Central Bank (ECB) assess the readiness of non-euro
area EU member states to join. In the latest reports from June 2014, no NMS-6 fulfills all
adoption criteria, hence, none is assessed as ready to join the euro at this stage.
4
The NMS-6 have considerable leeway over the timing of euro adoption. Especially
two adoption criteriaharmonization of the legal framework with euro area standards
and joining the European Exchange Rate Mechanism (ERM2)require a sovereign
decision. At the same time, the euro area institutions have also substantial discretion in
the euro adoption process, especially as regards admitting countries to ERM2.
2. Since 2004—when the first Central and Eastern European countries joined the
EU—five NMS have adopted the euro: Slovenia, the Slovak Republic, and the Baltic
countries. While in the Baltics, euro adoption followed many years of unilateral hard pegs to
the euro, Slovenia and the Slovak Republic maintained monetary autonomy until shortly
before euro adoption. At the same time, new countries have joined the EU in the past
decadeBulgaria, Romania, Croatiaand now face the issue of euro adoption.
3. The NMS-6 that have not yet adopted the euro maintain fairly different
monetary regimes and strategies.
Bulgaria and Croatia have tied their currencies to the euro: the Bulgarian lev by means
of a currency board, the Croatian kuna in the form of a tightly managed quasi-peg. In
both countries, the exchange rate anchor was introduced in the mid-1990s to combat
hyper-inflation (in Croatia in the context of the dissolution of former Yugoslavia). Thus,
3
This distinguishes the NMS-6 from the United Kingdom or Denmark, both of which negotiated an opt-out.
Sweden did not negotiate an opt-out and is subject to the same assessment procedures as the NMS-6.
4
The criteria are fixed in the Treaty on European Union and defined as follows: 1/ Legal = includes the
statutes of the national central bank; 2/ Fiscal = a government budgetary position without a deficit and debt
level that are determined excessive; 3/ Price stability = a rate of inflation which is close to that of, at most,
the three best performing Member States in terms of price stability; 4/ Exchange rate stability = the
observance of the normal fluctuation margins provided for by the exchange-rate mechanism of the
European Monetary System, for at least two years, without devaluing against the euro; 5/ Interest rate
stability = observed over a period of one year before examination, a Member state has had an average
nominal long-term interest rate that does not exceed by more than 2ppt that of, at most, the three best-
performing Member States in terms of price stability.
CEE NEW MEMBER STATES POLICY FORUM
6 INTERNATIONAL MONETARY FUND
By contrast, central banks in the Czech Republic, Hungary, Poland, and Romania target
inflation and, correspondingly, have in general allowed exchange rates to floata policy
targeting in 1997. In Hungary and Romania, central banks have at times carried out
significant interventions in foreign exchange markets, arguably to prevent the revaluation
of foreign currency denominated loans in case of excessive exchange rate volatility. As a
the Polish zloty is classified a
unconventional monetary policy at the zero interest bound. Until 2013, howeveri.e., in
the period covered by the empirical analysis of this paperthe Czech koruna was
4. This paper’s objective is to illustrate economic benefits and costs from euro
adoption by reviewing the main arguments and empirical evidence. The last time Fund
staff analyzed this issue systematically was in 2004 (published as Schadler et al, 2005; see
Box 1 for a summary of the main findings), in a study informed by the literature on optimal
currency areas (Box 2). This paper follows and builds on the conceptual framework of 2004,
by both reviewing new results from the literature and contributing analysis in areas less
covered by existing research. We also discuss issues that were less on the radar screen in
2004 but have since come to the forefront, especially in the wake of the global financial crisis.
5. Broadly speaking, the trade-off presents itself as follows:
On the positive sidei.e., in favor of euro adoptionthe 2004 study identified: (i) trade
generation that could translate into higher growth; and (ii) improved country risk
perception from deeper integration with the euro area that could, inter alia, lead to lower
funding cost.
EU Currency Exchange rate regime classification Earlier regimes
Accession (IMF, 2014)
Bulgaria 2007 Lev Hard peg - currency board Float until 1997. Currency board adopted in response to hyperinflation.
Croatia 2013 Kuna Soft peg - crawl-like arrangement Dinar until 1994, hyperinflation. Soft peg to deutsche mark until euro introduction.
Czech Republic 2004 Koruna Other managed arrangement Peg to dollar/currency basked until 1997. Inflation targeting; free float until 2013.
Hungary 2004 Forint Floating - inflation targeting Various pegs and crawling pegs until 2008 with increasingly wide bands (15% 2001-08).
Poland 2004 Złoty Free floating - inflation targeting Various pegs and crawling pegs until 2000.
Romania 2007 Leu Floating - inflation targeting Monetary based targeting with managed float until 2005. Abandoned amid high inflation.
Estonia 2004 Euro (since 2011) … Pegged to deutsche mark/euro (currency board).
Latvia 2004 Euro (since 2014) … Pegged to SDR 1994-2004, pegged to euro 2004-2014.
Lithuania 2004 Euro (since 2015) … Pegged to dollar 1994-2002, pegged to euro 2002-2014 (currency board).
Slovak Republic 2004 Euro (since 2009) … Float until 2004. Float with a 15% band during ERM II (2005-08), repeated parity adjustments.
Slovenia 2004 Euro (since 2007) … Float until 2004. Pegged to euro during ERM II 2004-06.
New Member States - Currency Arrangements
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 7
On (i) there is a solid body of evidence now that trade generation has remained far below
original expectationssee Box 1and we do not delve into this issue further. As for
(ii) we reassess the impact of euro adoption on investor perception of country risk in the
light of the evidence of the past 10 years.
We also analyze a benefit of euro adoption not covered in the 2004 paper: the
elimination of currency mismatches resulting from the prevalence of foreign currency,
mainly euro
illustrated, currency mismatches can translate into large vulnerabilities in times of
financial strain.
As the main factor on the negative sidei.e., cautioning against rapid euro adoption
the 2004 study identified the loss of monetary autonomy; i.e., ceding the ability to adapt
hange rate as shock
absorber. We re-assess the value of monetary autonomy in light of the experience of the
past ten years that were characterized by far higher macro-economic volatility than
expected (Box 1).
Finally, we discuss how euro adoption changes
frameworks beyond monetary policy, and how recent euro area wide reforms in these
areasfiscal compact, banking union
6. Importantly, weighing the pros and cons of euro adoption is ultimately an issue
of preference that can differ between countries. Further, the decision on adoption clearly
goes beyond purely economic aspects and includes political economy and broader political
considerations. These are beyond the scope of this paper. As a result, we refrain from
recommendations on whether a country should adopt the euro or not, but instead focus on
key macro-economic tradeoffs.
7. The remainder of this paper is organized as follows. Section B reviews key
economic advantages of euro adoption as sketched above, Section C the advantages from
maintaining monetary policy autonomy. Section D discusses the impact of euro adoption on
policy frameworks. Section E concludes.
B. Advantages from Adopting the Euro
8. As outlines above, this section analyzes two possible advantages of euro
adoption:
Improved country risk perception
institutional framework that may translate, inter alia, into lower funding costs, and
the elimination of currency mismatches between euros and domestic currencies.
CEE NEW MEMBER STATES POLICY FORUM
8 INTERNATIONAL MONETARY FUND
Euro Area Membership and Country Risk Perception
9. Adopting the euro can reduce perceived risks through the elimination of exchange
rate risk and access to lender-of-last resort facilities in a global reserve currency. For
countries with weak institutions, euro adoption can also strengthen the credibility of the
monetary anchor.
5
Conversely, euro adoption may undermine a country risk perception,
especially when it reduces the ability to handle country-specific shocks (as discussed in
Section III).
10. To analyze the effect of euro adoption on country risk perception, we estimate
an econometric model (Figure 1).
Country risk perception indices. We use two different indices: the Institutional
sovereign credit ratings (S&P). The IIR index is based on anonymous inputs from
economists and risk analysts at banks, money market funds, and securities firms.
directly measures investor attitudes toward country risk. The S&P rating is based on a
formal assessment methodology, complemented by judgment. It is used both for
regulatory purposes and to inform asset allocations of institutional investors, but has at
times been criticized for (alleged) biases (see e.g. Vernazza et al, 2014). The two indices
display similar patterns, and their relationship is broadly linear, with 45 points on the IIR
scale corresponding to one rating notch with S&P (Figure 1).
Analysis. We estimate the relationship between perceived credibility and country
characteristics for 34 countries during 200113. One characteristic is euro area
membership. A positive coefficient suggests that investors put a on
membership.
6
The premium is estimated for each year separately (see Appendix I).
Results. The reputational value of euro area membership has declined. Through most of
the 2000s, membership provided a substantial country risk premium of 1015 rating
points on the IIR scale, and of about two rating notches with S&P. This premium has
mostly vanishedentirely for the S&P index, and to a somewhat lesser degree for the IIR.
The timing of the declinestarting in 2010suggests that the euro area crisis triggered
a reassessment among investors of the relative benefits and drawbacks of euro
membership for country risk.
5
These advantages were clearly on the mind of the early euro adopters. For example, in 2003 the Slovenian
providing a more stable environment for the whole
economy”, and the adoption of the single currency will
represent the completion of the integration process”.
6
Strictly speaking, the premium measures the impact of euro membership on the perceived distance to
default.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 9
Figure 1. Euro Premium, 2001-13
Sources: EBRD; International Investor Ratings; Standard and Poors; and IMF Staff calculations.
-9
-6
-3
0
3
6
9
12
15
18
2001
2003
2005
2007
2009
2011
2013
International Investor Ratings Premium for Euro
Membership (Rating points)
-2
-1
0
1
2
3
4
2001
2003
2005
2007
2009
2011
2013
Standard and Poors Rating Premium for Euro
Membership (Rating notches)
y = -0.2258x + 21.957
R² = 0.9296
1
3
5
7
9
11
13
15
17
0 20 40 60 80 100
The S&P and IIR Credit Ratings
IIR Rating
S & P Rating (linearized)
50
55
60
65
70
75
80
85
90
95
2001
2003
2005
2007
2009
2011
2013
Euro area
EU, not euro
Outside EU
Evolution of Ratings per Country Group
IIR (lhs) S&P (rhs)
AAA
A
AA+
AA
AA-
A+
A-
BBB
BBB-
BBB+
CEE NEW MEMBER STATES POLICY FORUM
10 INTERNATIONAL MONETARY FUND
11. As a result, the NMS-6 today face a different situation than the early euro
adopters in the 2000s. While Slovenia or the Slovak Republicthat adopted the euro in
2007 and 2009, respectivelycould expect to benefit from a sizeable reputational boost
upon joining the euro, this benefit has vanished with the euro area crisis.
Eliminating Currency Mismatches
12. Beyond the impact of the euro on general risk perception, euro adoption can
benefit a country if it reduces specific vulnerabilities—notably foreign currency
mismatches. Such mismatches arise when households and corporations are indebted in
foreign currency (FX), while their assets and income streams are in domestic currency. In this
case, currency depreciation can trigger an upward revaluation of debt that can harm financial
stability and economic activity. While countries can self-insure against depreciation risk
through accumulation of FX reserves and regulations forcing financial institutions to hold
extra buffersthis is costly. FX indebtedness is widespread in Central and Eastern Europe,
oftenbut not alwaysintermediated by subsidiaries of banks located in the euro area (see,
e.g., Brown and de Haas 2012, or Rosenberg and Tirpak 2009). Among the NMS-6, balance
sheet euroizationproxied here by the share of bank loans to the private sector
Investor Perception and Balance Sheet Euroization, 2006-14
Balance Sheet Euroization Euro Premium and Balance Sheet Euroization
(FX loans to the private sector as share in total) (Fixed effects, 2006-12)
Sources: EBRD; International Investor Ratings; Standard and Poors; and IMF Staff calculations.
0 20 40 60 80 100
HRV
ROU
BGR
HUN
POL
CZE
LTU
SVN
LVA
EST
SVK
2006
2010
2014 (Nov)
Balance Sheet Euroization
(FX loans to the private sector as share in total)
CZE
POL
HUN
ROU
BGR
LTU
HRV
LVA
R² = 0.390
0
10
20
30
40
50
60
70
80
90
-15 -10 -5 0 5 10
Credibility Premium and Balance Sheet
Euroization (Fixed effects, 2006-12)
Average premium (in IIR points)
FX loans to the private sector as share of total
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 11
denominated in FX (see figure)is considerable in all economies except Poland and the
Czech Republic.
7
13. As most FX debt in the NMS is denominated in euro, euro adoption can
eliminate sizeable currency mismatches,
8
as it did in the Baltic states and in Slovenia upon
joining the euro (see figure). Similarly, euro adoption grants access to euro lender-of-last-
resort facilities for banks with a high share of euro denominated assets and, correspondingly,
high euro funding needs.
9
In contrast to the general euro premium discussed above, these
FX-specific benefits are disproportionately to the advantage of economies with a high share
of FX loans prior to euro adoption.
14. Our results suggest that for highly euroized economies, eliminating mismatches
through euro adoption is indeed beneficial. This is shown by correlating the country-
specific premiumcaptured by country dummieswith the share of FX bank loans. The
results show that for nearly fully euroized economies, euro adoption can eliminate a ratings
malus of more than 10 IIR points. Further statistical analysis points to a non-linear impact, i.e.,
the malus increases disproportionately with higher levels of euroization (see Appendix I).
15. In contrast to the general euro premium discussed above, FX-specific benefits
from euro adoption have not vanished. This provides an economic rationale why the Baltic
countriesall of them highly euroizedsought euro area membership in the early 2010s,
even though at the time the general euro premium was already waning.
C. Advantages from Preserving Monetary Policy Autonomy and
Exchange Rate Flexibility
16. We now turn to advantages from maintaining monetary autonomy, i.e., the
exchange rate to operate as a shock absorber. If successful, monetary policy autonomy helps
stabilize domestic demandespecially when economies are exposed to shockscontain
inflation volatility and credit developments. To assess the value of monetary autonomy for
the NMS, we look at three distinct episodes in the past 1012 years:
7
This proxy correlates closely with more comprehensive metrics of balance sheet euroization, such as the
currency mismatch index in Ranciere et al. (2010). Data requirements for these indices are larger than for FX
bank loans, however, which would have restricted our sample significantly.
8
In some economiesnotably Poland and Croatia a significant part of FX loans is denominated in Swiss
francs rather than euro (in Hungary, most CHF loans were converted into domestic currency, with a de-factor
conversion date of November 2014). Thus, some currency mismatch would remain even after euro adoption.
9
The value of ECB access became apparent during the financial crisis of 2008/09, when cross-border
currency markets became impaired. The problem was especially severe for banks with sizeable FX assets
and therefore high FX refinancing needsbut without a euro area based parent bank that could have
accessed ECB facilities on their behalf.
CEE NEW MEMBER STATES POLICY FORUM
12 INTERNATIONAL MONETARY FUND
economic convergence, i.e. the period of about 200307, when the NMS outgrew the
old EU member states (EU 15) by about three percentage points per year;
the financial crisis of 2008/09 and its aftermath that brought the convergence process
in most NMS to a halt; and
the most recent period of 2012–14
when the NMS were affected by
disinflationary pressures from global
commodity markets and the euro area.
For each episode, we map a standard
monetary conditions index
10
against domestic
demand volatility and related outcomes, such
as real credit growth and inflation, in (i) NMS
that have maintained monetary policy
autonomy and (ii) NMS that have used
external monetary anchorseither by
adopting the euro, or by tying their currencies
to the euro or other currencies/currency
baskets. While there is considerable
heterogeneity of outcomes within both
groupsreflecting
monetary policysome broad patterns emerge.
17. Importantly, this section is not about whether fixed or flexible exchange rates
are generally preferable. Many reasons that go beyond short-to medium-term demand
management as discussed here may call for one regime or the othere.g., in the case of a
fixed exchange rate regime, the need for a credible monetary anchor. What choice is
appropriate will typically depend on country-specific circumstances. A widespread finding in
the literature is that long-term growth and inflation outcomes tend to be broadly similar
under both regimes, while fluctuations tend to be larger with fixed exchange rates (see, e.g.
IMF, 2005 and 2013b).
Managing Convergence
18. In 2003–07, the NMS grew at an average annual real rate of about 6 percent,
3 percentage points faster than the old members of the European Union (EU 15). As a
10
The monetary conditions index is calculated as the weighted average of real interest rate and real
exchange rate, with weights representing the impact of each component on domestic demand (see
Appendix II).
0
10
20
30
40
50
60
70
80
90
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
BGR
HRV
CZE
HUN
POL
ROU
LTU
EST
LVA
SVK
SVN
GDP Per Capita in Purchasing Power Standard
(Percent of the euro area)
-4
-2
0
2
4
6
8
POL
CZE
HUN
SVN
SVK
Growth differential w/ Euro Area
Monetary Conditions differential w/ Euro Area
Growth and Monetary Conditions
(Percent; average 2003-07; peggers labeled red)
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 13
result, the average income gap per capita (PPP) against the EU 15 fell by about 10 percent,
until convergence stalled with the outbreak of the global financial crisis.
19. Higher growth rates for an extended period require a tighter monetary policy
stance. Less wealthy countries have typically a smaller capital stock and therefore higher
returns of capital. This attracts capital inflows, boosts economic growth, and gradually
increases the capital share until economies converge. Unless the higher real return on capital
is matched by tighter monetary conditions, however, there is a risk that capital inflows trigger
credit and housing booms, inflation, and external imbalances (see Lipschitz et al., 2002).
11
As
countries with fixed exchange rate regimes tend to adopt the monetary policy stance of the
economy to which their currency is tied, there is a risk that during convergence, monetary
conditions are too loose. This risk increases with l income gap with the euro
area, and therefore with its potential for rapid growth.
20. The data show a pattern of tighter monetary conditions for economies that
maintained monetary policy autonomy during convergence (Figure 3). This is especially
evident for the Czech Republic and Poland, both of which let their currencies float during this
period, and to a lesser degree for Hungary (that held the forint within a wide band against
the euro). Importantly, the tightening in monetary conditions was achieved mostly by means
of nominal exchange rate appreciation rather than higher central bank interest rates (Box 3).
21. By contrast, monetary conditions in countries with fixed exchange rates tended
to be, overall, fairly accommodative, often even more accommodative than in the euro
area. Correspondingly, such countries tended to experience stronger overshooting in
domestic demand, larger credit and asset booms, and higher and more volatile inflation (for
a detailed discussion, see Bakker and Gulde, 2010).
Managing Downturns
22. In the global financial crisis of 2008/09 and its aftermath, the NMS were hit by a
severe negative demand shock that stalled or even reversed convergence. Monetary
autonomy allowed central banks to respond by cutting policy rates andmore importantly
allow nominal exchange rates to depreciate and act as shock absorbers (Figure 4). Nominal
depreciations were especially large in Poland and Romania, where domestic demand held up
better than elsewhere.
12
By contrast, in countries with fixed exchange rates, monetary
11
This thought goes back as far as Wicksell (1898). It is also reflected in a simple Taylor rule that specifies the
cyclical behavior of policy rates around a neutral rate. Policy rates that are systematically above/below the
neutral rate will result in a monetary policy stance that is systematically too tight/too lose.
12
In Hungary, the scope for exchange rate depreciation was arguably constrained by the high share of
FX loans, see Bakker and Gulde (2010).
CEE NEW MEMBER STATES POLICY FORUM
14 INTERNATIONAL MONETARY FUND
Figure 2. New Member States: Monetary Policy, 2003-07
Sources: Haver Analytics; and IMF Staff calculations.
POL
CZE
HUN
HRV
SVN
BGR
LTU
EST
SVK
ROU
LVA
y = -0.998x + 6.8999
R² = 0.4554
0
2
4
6
8
10
12
14
-6 -4 -2 0 2 4 6
Domestic Demand Growth and Monetary Conditions
(2003-07 average; percent; peggers labeled red)
Monetary conditions (+tighter)
Domestic demand growth
-2
0
2
4
6
8
10
12
SVK
CZE
POL
HUN
SVN
Inflation differential
Nominal exchange rate appreciation
Inflation volatility (std)
Real Exchange Rate Appreciation Against Euro
(Contributions in percentage points; average 2003-07)
LVA
ROU
BGR
EST
HRV
LTU
BGR
HRV
CZE
HUN
POL
ROU
SVK
SVN
LVA
LTU
EST
y = -4.1588x + 29.01
R² = 0.556
0
5
10
15
20
25
30
35
40
45
50
-3 -1 1 3 5
Real Credit Growth and Monetary Conditions
(Percent; average 2003-07; peggers labeled red)
Monetary conditions (+tighter)
Real credit growth
-4
-2
0
2
4
6
8
POL
CZE
HUN
SVN
SVK
Growth differential w/ Euro Area
Monetary Conditions differential w/ Euro Area
Growth and Monetary Conditions
(Percent; average 2003-07; peggers labeled red)
LVA
ROU
BGR
EST
HRV
LTU
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 15
conditions tended to be tighter driven in part by the need to defend the exchange rate
arrangement.
13
As a consequence, the demand contraction in crisis tended to be larger, even
though there are exceptions.
14
Offsetting Disinflationary Spillovers
23. In recent years, the NMS have been affected by deflationary shocks from falling
global commodity prices and euro area disinflation. The impact on NMS has been
asymmetric: while in economies with flexible exchange rates, disinflation has mostly affected
headline inflation, in exchange rate pegging economies deflationary pressures have tended
to creep into core inflation.
15
24. Econometric analysis suggests that the link between disinflation spillovers and
the exchange rate regime is systematic. Iossifov and Podpiera (2014) decompose headline
inflation in the NMS-6 into country-specific, global, and euro area factors using an open-
economy New Keynesian Phillips curve within 200414 panel data framework (see
Appendix III). The share of inflation variance explained by euro area developments is largest
in Bulgaria and Croatia, both of which have tied their currencies to the euro . By
contrast, inflation targeting central banks have thus far been able to largely offset the
deflationary impulse from the euro area with their monetary policy stance.
16
D. The Impact of Euro Adoption on Macroeconomic Policy
Frameworks
25. Euro adoption would change the NMS-6’s policy frameworks. Ceding monetary
autonomy is only onealthough the most importantcomponent. In addition, the stability
and growth pact foresees stricter enforcement mechanisms for euro area members
13
In this regard, there is an important difference between countries with currencies tied to the euro and
those already in the euro area (Slovak Republic and Slovenia), as the latter did not have to tighten monetary
policy in order to defend a peg.
14
Bulgaria, for example, had a relatively muted demand contraction, arguably reflecting in part the use of
fiscal buffers
15
This result holds even after controlling for the size of import exposure to the euro area.
16
The value of monetary autonomy is also being recognized by rating agencies, as the following quote from
Unlike in the Baltic countries, euro adoption would not immediately
lead us to raise the ratings on Poland. This is because the Baltics, unlike Poland, already had currency pegs to
the euro in place, which limited their monetary flexibility. Poland, on the other hand, has a flexible exchange
rate, and the current ratings clearly take into account its monetary flexibility and effective monetary
policymaking as a ratings strength. On the one hand, euro membership would give Poland access to a reserve
currency and the ability to issue debt in it, which would be positive for the ratings. On the other hand,
membership of a monetary union that has a monetary policy not necessarily geared toward the needs of
individual countries could be ratings-negative, especially if we saw a strong asynchrony between European
Central Bank (ECB) policies and a monetary policy stance needed for Poland's purposes.”
CEE NEW MEMBER STATES POLICY FORUM
16 INTERNATIONAL MONETARY FUND
Figure 3. New Member States: Monetary Policy, 2008-14
Sources: Haver Analytics; and IMF Staff calculations.
0
20
40
60
80
100
HRV
BGR
CZE
HUN
POL
ROU
World food and energy prices, incl. taxes and admin. prices
Taxes and administered prices
Unemployment gap
Nominal effective exchange rate
Euro Area core inflation
Headline Inflation Variance Decomposition
(Contributions in percent; 2008-14)
LTU
BGR
HRV
SVK
SVN
EST
LVA
y = -0.4535x - 1.3689
R² = 0.6605
-4
-3
-2
-1
0
1
2
3
-6 -4 -2 0 2 4 6
Real Monetary Conditions and Domestic Demand Growth
(2008-14 average; percent; peggers labeled red)
Domestic demand growth
Monetary conditions (+tighter)
-6
-4
-2
0
2
4
6
LTU
BGR
HRV
SVK
SVN
EST
LVA
Real Effective Exchange Rate
Real Interest Rate
Monetary Conditions Index
Contributions to Monetary Conditions
(2008-14 average; percent; peggers labeled red)
POL
CZE
HUN
ROU
-2
-1
0
1
2
3
4
5
6
7
2008Q4
2010Q4
2012Q4
2014Q4
NMS-IT
NMS-ET
World core inflation (excl. NMS and
Euro Area)
Euro Area core inflation
Core Inflation
(y-o-y, percent)
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 17
than for countries outside the euro area, and euro area members are obliged to join the
Banking Union, which comes with the transfer of most authority for micro-prudential
supervision and some aspects of macro-prudential policy to the ECB.
26. The challenge to adapt to the euro area policy framework presents itself
differently for countries with flexible and with fixed exchange rates.
17
For NMS-6 that let exchange rates float and target inflation, monetary policy
autonomy would have to be replaced with other instruments, such as counter-cyclical
fiscal policy, macro-prudential tools, and internal adjustment through wages.
For the NMS-6 that have pegged their currencies to the euro, an important issue is to
what extent they could use macro-prudential policies also within the euro areaas
Bulgaria and Croatia have been among the heaviest users of macro-prudential
instruments in Europe (see Lim et al., 2011, and , 2014).
18
27. At this juncture, key euro area reforms that will affect both fiscal and financial
sector governance are ongoing or have only just been completed. The final shape and
icy space in the euro
area, as well as the support they can expect from common euro area institutions in
combating economic and financial shocks.
28. Given the large uncertainty surrounding these reforms—regarding both their
final shape and how they will work in practice—there is a rationale to waiting for final
outcomes before taking an irreversible euro adoption decision. In an uncertain
environment, the option to wait until benefits and cost become clearer has value by itself.
The standard economic applicati
decisions (see McDonald and Siegel, 1986, or Dixit and Pindyck, 1994). However, the rationale
also applies to irreversible policy choices, such as euro adoption.
17
It is worth noting that in case the NMS-6 were to adopt the euro, this would not only affect the NMS-6
but also the existing euro area members. The NMS-6 would account for about 8 percent of
GDP (14 percent on a PPP basis), and a similar share of euro area consumption. In case convergence would
resume at the same pace as in the early 2000s, staff estimates that NMS-6 adoption could increase average
euro inflation by up to 0.3 percentage points, which would require a tighter monetary policy stance by the
ECB to preserve its current inflation target. This, in turn, would further diminish space for nominal adjustment
in case euro area members suffer asymmetric shocks.
18
The jury is still out on the effectiveness of these measures. A recent paper by Cerutti et al. (2015), for
example, finds some effectiveness of macro-prudential policies in managing credit cycles, but also severe
issues of avoidance through greater cross-border borrowing.
CEE NEW MEMBER STATES POLICY FORUM
18 INTERNATIONAL MONETARY FUND
E. Conclusions
29. The parameters of the euro adoption debate have shifted. While countries joining
the euro area in the 2000s could expect to benefit from a significant country risk premium,
this premium has mostly vanished with the euro crisis. When or whether it will return is
uncertain and will depend in part
institutional framework. At the same time, for countries with sizeable shares of foreign
currency loans, there are still important financial stability benefits from adopting the euro.
30. The NMS that have maintained exchange rate flexibility and monetary policy
autonomy have, in general, made good use of it. During convergence, nominal currency
appreciation supported more balanced growth and restrained credit and asset price booms.
In the crisis-induced downturn of 200809, depreciation and monetary loosening helped
stabilizing demand and, more recently, prevented external deflationary pressure from spilling
into domestic core inflation.
31. It is an open question whether the macroeconomic volatility of the past decade
will recur. If divergent growth patterns and volatility were to repeat, euro adoption would
constrain macro-policy options, especially for economies with large income gaps and a-
synchronized business cycles vis-à-vis the euro area. Thus, a large burden would be placed
on other policy instruments to safeguard balanced growth, notably counter-cyclical fiscal
policywhich, in turn, requires fiscal spaceand macro-prudential policies. Structural
reforms to boost growth potential and facilitate internal adjustment would also be key.
32. For countries that peg their currencies to the euro, the balance of the argument
is somewhat different, as they have already traded monetary autonomy and exchange rate
flexibility for benefitting from the euro as monetary anchor. Thus, for the most part, their
policy frameworks would not change materially upon adopting the euro. An important
remaining issue though is to what extent they could employ macro-prudential and other
regulatory instruments also within the euro area.
33. The scope for using fiscal policy and macro-prudential instruments is currently
being re-defined in the context of ongoing reforms of the European fiscal and financial
architecture. Depending on the final shape of these reforms, this may or may not constrain
policy space for euro area members. The uncertainty about the outcomes of these reform
effortsand the limited means of the NMS-6 to affect themprovides a rationale to wait for
and analyze final outcomes before taking an irreversible adoption decision.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 19
Box 1. The 2004 IMF Study, and New Evidence on its Main Findings
The 2004 IMF study “Adopting the Euro in Central Europe—Challenges of the Next Step in European
Integration” was cautiously optimistic about the net economic benefits of euro adoption.
As key factor in favor of euro adoption, the study singled out trade generation and, as a result, higher
growth. Based on estimates from the literature (e.g., Frankel and Rose, 1998 and Rose, 2000), staff
considered real GDP gains of 10 to 25 percent over 20 years plausible. Staff noted though that the
mechanism generating such large effects was not entirely clear, as the reduction in transaction cost and
the elimination of exchange rate uncertainty would explain only a minor portion. The study also
anticipated that adoption would reduce perceived credit risk which could, inter alia, reduce funding
costs for NMS in the euro.
As for possible risks from euro adoption, the study acknowledged that coping with asymmetric demand
shocks would become more difficult without an independent monetary policy. However, the study noted
that the NMS had experienced low growth volatility in the preceding years, suggesting that shocks may
be manageable also without monetary autonomyprovided countries had fiscal space for
countercyclical policies, and wages and prices were sufficiently flexible. Further, economic integration in
the wake of euro adoption was expected to lead to more synchronized business cycles. As another
potential risk from adopting the euro, the study identified large and volatile capital inflows and lending
booms, as well as higher inflation due to the Balassa/Samuelson effect. Again, fiscal space and wage
flexibility were seen as key to manage these phenomena, together with structural reforms to boost
competitiveness, and strong financial
supervision.
1
Subsequent developments confirmed
parts of the 2004 assessment, but in other
parts yielded new insights.
Risk spread compression happened as
anticipated in 2004, at least until the
warnings of excessive asset and credit
booms were well placed.
Recent estimates of trade and growth
effects are much smaller than assumed
at the time, however, and range from nil
(Havranek, 2010) to 2-3 percent of GDP
(Baldwin, 2006). Further, recent studies suggest that trade generation is more related to EU entry than to
euro adoption.
Demand volatility has been significantly larger than anticipated in 2004, especially in the wake of the
global financial crisis of 2008/09 and the ensuing euro crisis. Endogenous business cycle
synchronization with euro adoption has not materializedrather, there have been signs of divergence,
such as growing external imbalances (Gayer 2007; Holinski et al. (2012), Enders et al. 2013, Lehwald,
2013; and Degiannakis et al., 2014).
_____________________________
1/
The 2004 study also discussed ERM2-related strategies. The study recommended completing all necessary structural and fiscal
reforms prior to entry into the ERM2 to ensure a smooth transition. Countries with autonomous monetary regimesmostly
inflation targetingshould maintain these until ERM2 entry.
0
1
2
3
4
1963
1968
1973
1978
1983
1988
1993
1998
2003
2008
2013
Europe
Emerging Economies
Growth Volatility
(5-year rolling standard deviation of per-capital real GDP growth,)
CEE NEW MEMBER STATES POLICY FORUM
20 INTERNATIONAL MONETARY FUND
Box 2. New Member States, Euro Adoption, and the Theory of Optimal Currency Areas
The academic literature on currency unions often casts the issue in terms of whether member
countries form an “optimum currency area, OCAMundell, 1961; McKinnon, 1963; and Kenen, 1969).
In principle, countries can benefit from a currency union through capital market integration, lower
terms-of-trade volatility and reduced exchange rate fluctuations. Countries with weaker institutions can
import monetary policy credibility (McKinnon, 2004; Tavlas, 1993).
There are also costs, mostly associated with forgoing exchange rate flexibility and monetary policy
autonomy for managing cyclical conditions. To minimize such costs, countries in currency unions would
best have synchronized business cycles and growth patterns, which is typically enhanced by intra-
industry trade (Frankel and Rose, 1998).
In currency unions with incomplete business cycle synchronization, flexible non-monetary
adjustment mechanism are needed: notably price and wage flexibility and cross-border labor mobility.
Risk sharing through integrated financial markets can also help, by diversifying income sources through
cross-country asset holdings (McKinnon, 2004; Mongelli, 2008).
Business cycle synchronization of NMS with the euro area is in general lower than within the euro
area, even though there are differences between countries (Fidrmuc and Korhonen, 2003; Artis et al.,
2004; Darvas and Szapáry, 2005; Van Arle et al. 2008). For Hungary, Poland, and the Czech Republic,
synchronization with the euro area is higher than for the other NMS; and also higher relative to Greece and
Portugal (Fidrmuc and Korhonen, 2006; Rinaldi-Larribe, 2013).
Wage flexibility in the NMS tends to be high while cross-border labor mobility has been increased
(Gruber, 2004; Dao et al. 2014), with substantial outward migration from some NMS in the wake of the
global financial crisis (OECD, 2013):
Wage flexibility. According to Gruber (2004) and Boeri and Garibaldi (2006) NMS tend to have lower
statutory minimum wages, union density rates and more decentralized wage bargaining structure than
the euro areaall pointing to wage flexibility.
Labor mobility. Outward migration in the post-crisis period of 2009-2011mostly was younger and
educated workers, typically finding employment abroad below their skill level (Anacka et al., 2011; Jauer
et al., 2014).
Cross-country risk sharing through financial market integration is low. While prior to the 2008/09
financial crisis, there was some evidence of increasing integration in equity and debt marketssuch as lower
interest rate dispersion and increasing effects of the euro area shocks on NMS equity markets (Cappiello et
al. 2006; Baltzer et al. 2008)these reversed in the wake of the crisis, with higher interest rate spreads
(Pungulescu, 2013) and funding market segmentation along national lines (van Rixtel and Gasperini, 2013).
In addition, cross-border asset holdings tend to be one-directional: euro area banks hold sizeable assets in
NMS, and also FDI and portfolio investment from euro area residents in NMS is much larger than vice versa.
As a result, domestic investment in the NMS is more sensitive to domestic saving than in EU15, and national
consumption is closely correlated with GDP (Pungulescu, 2013).
Overall, while there are significant structural differences and lack of integration with the euro area, for NMS
like the Czech Republic, Hungary, and Poland, this do not seem larger than heterogeneity within the euro
area itself.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 21
Box 3. Economic and Price Convergence in Emerging Europe
The political transformation in NMS in 1990s led to economic convergence through economic reforms,
including trade and price liberalization, privatization, and the adoption of the legal framework of advanced
Europe. Institutional change established the base for rapid growth, facilitated by external capital investment,
a supportive external financing environment, and an often skilled labor force and reasonable infrastructure.
As a result, productivity and income per capita increased by 20 percent relative to the Euro Area countries by
between 1995 and 2013 (see IMF, 2014).
Improvements in productivity go typically hand in hand with real exchange rate appreciation. There
are two, not mutually exclusive channels the Balassa-Samuelson and the Podpiera effect:
The Balassa-Samuelson effect reflects faster productivity growth for tradable than for non-tradable
goods during convergence. As the price for tradable goods is fixed in global goods markets, the price for
non-tradables has to rise. However, contrary to initial expectationsformed during early stages of
convergencethat such an effect could account for 1-2 percentage points of consumer price inflation per
year (Cipriani, 2000, Kovacs et al., 2002, and Mihaljek and Klau, 2003), more recent studies find much
smaller effects (Mihaljek and Klau, 2008).
The Podpiera effect is due to converging economies shifting production toward higher quality goods. As
quality improvements come with price increases and are typically underreported in domestic CPI indices,
these effects are reflected in real exchange rate appreciation. Cincibuch and Podpiera (2006) and
Fabrizio et al. (2007) document the rapid increase in prices in tradable sectors in NMS. Bruha and
Podpiera (2010 and 2011) devise and calibrate a model that explains the rapid (2-3 percent a year) CPI-
based real exchange rate appreciation during convergence. Sonora and Tica (2014) test jointly the
Balassa-Samuelson and Podpiera effects and find that the latter primarily explains the real exchange rate
appreciation in Emerging Europe.
Real exchange rate appreciation happened partly through nominal exchange rate appreciation in
countries with flexible exchange rate regimes. In exchange rate targeting countries, it happened solely
through higher inflation differentials (box charts and see figure on Investor Perception and Balance Sheet
Euroization).
y = 0.7597x + 0.2395
R² = 0.4716
0.8
0.9
1
1.1
1.2
1.3
0.85 0.95 1.05 1.15 1.25
Real and Nominal Convergence
(Indexes, 2005=1, 2003Q1-07Q4)
GDP growth differentials
Real exchange rate against euro
-5
-3
-1
1
3
5
7
9
11
13
15
SVK
CZE
POL
ROU
HRV
SVN
BGR
LVA
EST
LTU
HUN
Nominal Exchange Rate Against Euro
(Percentage change 2003Q1-07Q4)
Source: Haver and IMF Staff calculations.
CEE NEW MEMBER STATES POLICY FORUM
22 INTERNATIONAL MONETARY FUND
References
Anacka, Marta,
The 2011 SOPEMI Report.
Monte Carlo Evidence and an Review of
Economic Studies, Vol. 58/2, pp. 277-297.
CEPR Discussion Paper 4457.
Bakker, Bas B., and Anne-
IMF Working Paper No. 10/130. (Washington:
International Monetary Fund).
Working Paper 594, European Central
Bank, March.
Baltzer, Markus, Lorenzo Cappiello, Roberto A. De Santis and Simone Manganelli, 2008,
European Central Bank
Occasional Paper Series 81.
Boeri, Tito and Pietro Ga
Journal of Banking & Finance, Vol. 30, pp. 1393-1407.
Economic Policy, Vol. 27(69), pp. 57-98.
Economics of Transition, Vol. 18/3, pp. 599-628.
________________omic Convergence: The Role of Alternative
Journal of Economic Dynamics and Control, Vol. 35/7,
pp. 1032-1044.
Integration of New Member States,European Central Bank Working Paper Series 683.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 23
Cerutti, Eugenio, Stijn Claessens and Luc Laeven, 2015,
Macroprudential Policies: New Evidence
-Samuelson: Real Appreciation in
Economics of Transition, Vol. 14/3, pp. 547-573.
-
prepared by the CEC5 National Banks for the Basle meeting on March 2002.
Dao, Mai, Davide Furceri and Prakash Loungani, Market Adjustments in
the United States IMF Working Paper No. 14/26.
larged EU:
CEPR Discussion Paper No. 5179.
EU: A Time-Scottish Journal of Political Economy, Vol 61(4),
pp. 348-370.
Dixit, Avinash K. and Robert S. Pindyck, 1994, Investment under Uncertainty, Princeton
University Press.
- Is
th
Dubrovnik
Economic Conference.
Enders, Zeno, Philip Jung and Gernot J. Müller
European Economic Review, Vol. 59, pp 189-211. European Commission,
2014, Convergence Report, ISSN 1725-3217.
Fabrizio, Stefania, Deniz Igan, and
IMF Working Paper No. 07/97.
Economic Systems, Vol. 27(3), pp. 313-334.
Fidrmuc, Jarko and -analysis of the business cycle correlation
Journal of Comparative Economics, Vol. 34(3),
pp. 518-537.
Frankel, Jeffrey A. and Andrew K. Rose, 1
Economic Journal, Vol. 108(449), pp. 1009-1025.
CEE NEW MEMBER STATES POLICY FORUM
24 INTERNATIONAL MONETARY FUND
Freedman, C., 1994, The Use of Indicators and of the Monetary Conditions Index in Canada",
In: Balino, T. J. T. and Cottarelli, C. (eds.), Frameworks for Monetary Stability: Policy
Issues and Country Experiences, Chapter 18, pp. 458-476, International Monetary
Fund, Washington, D.C.
Journal of Monetary Economics, 44(2), October, pp. 195-222.
European Economy, Economics Paper 287.
Government of the Republic of Slovenia and Bank of Slovenia, 2003, The Euro is Coming,
OPOCE .
Focus 01/04, Osterreichische Nationalbank, pp. 96-121.
HavránekReview of World
Economics, June 2010, Vol 146(2), pp. 241-261.
the Role of
IMF Working Paper No.
14/17 (Washington: International Monetary Fund).
Federal Reserve Bank of St. Louis Review,
January/February 2012, Vol 94(1), pp. 1-20.
Ilzetzki, Ethan, Car
Economics, mimeo.
International Monetary Fund
Chapter IV, World Economic Outlook, September, (Washington).
International Monetary Fund, 2013a, Annual Report on Exchange Arrangements and Exchange
Restrictions, (Washington).
International Monetary Fund
BaWorld Economic Outlook,
September (Washington).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 25
International Monetary Fund-Communist Europe and the
, (Washington).
-Euro Area EU Countries Importing Low
IMF Working Paper No. 14/191.
Adjustment Mechan
OECD Social, Employment and Migration Working Papers No. 155.
Indicators: Methodology and Analytical IssWorld Bank Policy Research Working
Paper No. 5430.
In
R. Mundell and A. Swoboda eds, Monetary Problems of the International Economy,
The University of Chicago Press, pp. 41-60.
-Samuelson Effect in CEC5
on March 2002.
Synchronization? Evidence from
Empirica, Vol. 40 (4), pp. 655-684.
Lim, C., F. Columba, A. Costa, P. Kongsamut, A. Otani, M. Saiyid, T. Wetzel and X. Wu, 2011,
essons From
IMF Working Paper No. 11/238 (Washington: International
Monetary Fund).
Capital Flows to Transition
No. 02/11. (Washington:
International Monetary Fund).
Journal of International Money and Finance, Vol. 31/3,
pp. 657-672.
Matei, Iuliana and Angela Cheptea, 2012, "Sovereign Bond Spread Drivers in the EU Market in
the Aftermath of the Global Financial Crisis," in Essays in Honor of Jerry Hausman,
pp. 327-352.
CEE NEW MEMBER STATES POLICY FORUM
26 INTERNATIONAL MONETARY FUND
Quarterly
Journal of Economics, Vol. 101 (4), pp. 707-727.
American Economic Review, Vol. 53,
pp. 717-724.
Journal of Common Market Studies, Vol. 42(4), pp. 689-715.
-Samuelson effect in Central Europe: A
BIS Working Paper 143.
______________________________-up and inflation in transition economies: the
Balassa-BIS Working Paper 270.
Economic Paper 302,
Directorate General Economic and Monetary Affairs (DG ECFIN), European
Commission.
American Economic Review,
Vol. 51, pp. 657665.
D
Publishing.
Emerging Markets Review, Vol. 17, pp. 106-124
ex of Currency
IMF Working Paper No. 10/263 (Washington:
International Monetary Fund).
Rinaldi-Larribe, Marie-
The European Journal of Comparative Economics,
Vol. 5(2), pp. 133-154.
Economic Policy, Vol. 15(30), pp. 733.
ign Currency Borrowing
Journal of Economics and Finance, vol. 59(3),
pp. 216-228.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 27
Schadler, Susan, Paulo Drummond, Louis Kuijs, Zuzana Murgasova, and Rachel van Elkan,
Adopting the Euro in Central Europe Challenges of the Next Step in
IMF Occasional Paper 234.
Slovakian National Bank, 2003, S
BIATEC, Volume XI, 10/2003.
d Samuelson (Re)Visit Eastern
Cogent Economics & Finance (2014-2).
The World
Economy, Vol. 16(6), pp. 663-685.
Working paper B18. Center for European Integration Studies, ZEI,
Bonn.
Van Arle, Bas, Marcus Kappler, Andreas Sachs, Atilim Seymen, and Klaus Weyerstrass, 2008,
European Economic Research (ZEW) and Institute for Advanced Study (HIS).
Van Rixtel, Adrian and Ga
BIS Working Papers 406.
21.
European Journal of
Political Economy, Vol. 27(1), pp. 36-43.
Wicksell, Knut, 1898/1936, Interest and Prices (Geldzins und Güterpreise), Sentry Press,
New York.
CEE NEW MEMBER STATES POLICY FORUM
28 INTERNATIONAL MONETARY FUND
Appendix I. Estimating the Euro Premium
Euro Premium
The empirical analysis of the euro premium is based on panel data regression for
34 European countries with annual observations for the period 2001–2013. The time
period is mostly determined by data availability. We estimate the following equation:
, t=T
where
C is an expert assessment of country perception. We use two different measures, (i) the
a linearized
ings (S&P).
1
A possible alternative
measure would have been credit spreads as used, e.g., in Heinz and Sun (2014), Matei
and Cheptea (2012), Maltritz (2012), or von Hagen et al. (2011). The high volatility of
credit spreads, and the fact that spreads are affected by many other factors than
perceived credibilitysay, global liquidity conditions and/or cross-country spillovers
makes us select a metric that assesses investor perceptions directly.
X is a vector of country characteristics that may affect a coun
worthiness. X includes macro-variablessuch as fiscal and external balances, public debt,
per-capita-GDP, real GDP growth, the unemployment rate, inflation, the exchange rate
regime, national investment as share of GDP, and the c
position (NIIP)but also an indicator measuring institutional quality.
Most data are drawn from the IMF’s World Economic Outlook database. For
institutional quality, we use a simple average of the Kaufmann et al. (2010) governance
quality indices that is available, at this juncture, only for 200212. The exchange rate
regime is captured with a dummy variable for floating currencies (freely or managed)
drawn from Ilzetzki et al. (2011) until 2010 and extrapolated with IMF (2013a) for the
period thereafter. Public debt data is available for only about 90 percent of the
observations; we control for missing observations by including a corresponding dummy
variable.
1
For the IIR, respondents grade each country on a scale of zero to 100, with 100 representing the least
for the S&P index, the linearization assigns a value of 1 to an AAA rating, of 2 to AA+, and so on. We use the
foreign currency sovereign credit rating, but for the vast majority of countries, the local and foreign currency
ratings are identical.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 29
-2
-1
0
1
2
3
4
-9
-6
-3
0
3
6
9
12
15
18
2001
2003
2005
2007
2009
2011
2013
EU Membership Premium: IIR (red, lhs) and S&P (blue,
rhs) (Rating points)
-6
-3
0
3
6
9
12
2001
2003
2005
2007
2009
2011
2013
IIR Premium for Euro Area Membership: Arellano/Bond
GMM Dynamic Panel Estimator (Rating points)
is typically the expected outcome of variable for year t at the time when the WEO
fall forecast is produced. We experimented with leads and lags, but found that
concurrent values have the greatest explanatory power.
δ and θ are a country/time-dummies capturing unobserved, time invariant country-
specific factors/common unobserved time-specific factors, respectively. The time
dummies control for common macroeconomic and financial developments.
Euro is a dummy variable that takes on value 1 if country i was member of the euro
area in year t, and 0 otherwise (or technically speaking, we interact a general euro area
dummy with a time dummy). The
thus measure the impact of euro area membership
on country perception, holding country characteristics—both observed and
unobserved—constant. We allow γ to vary over time, to verify whether the value of euro
membership has changed. Similarly, EU is a dummy variable capturing EU membership.
Table A1 displays the regression results.
Euro and EU premia. While estimates
for the IIR index (column 1) and the S&P
indices (column 2) are in general similar,
the IIR tends to value euro area
membership higher than S&P
throughout. Further, the decline in the
premium in the wake of the euro crisis is
steeper for the S&P index than for the
IIR.
2
The premium for EU membership is
generally insignificant (see chart), except
for 2001/02, when EU membership
displays modest positive significance with
the IIR index.
3
Among controls, unemployment, and
government debt have the strongest impact on perceived credibility, followed by
inflation, growth, and the NIIP. Note that the IIP puts most weight on stock variables,
while for the S&P assessment flow variables also matter (budget deficit, current account
balance). Governance quality shows up as a strong determinant of credibility (column 3).
2
This is consistent with claims that rating agencies downgraded euro area countries excessively in crisis—
see, for example, Vernazza et al. (2014). The finding of a vanishing euro premium is consistent with the
results in Heinz and Sun (2014) for credit spreads.
3
The pattern in the early 2000s reflects arguably more an improvement in investor perception of non-EU
countries than a deterioration in the perception of EU member countries—see also Figure 1 in the main text.
CEE NEW MEMBER STATES POLICY FORUM
30 INTERNATIONAL MONETARY FUND
At the same time, its inclusion (or
omission) does not materially affect the
pattern of the euro premium over time.
To check for robustness we estimated
various other specifications, both by
including additional control variables—
such covariates capturing global
macro/liquidity conditions, e.g. the VIX
index—and by estimating dynamic
panels—notably the Arellano-Bond
(1991) GMM estimator. The basic
pattern—a steep decline in the euro area
premium from 2010—persists across all
specifications.
4
4
With Arellano-Bond, the estimated size of the euro premium is on average smaller by about one-third.
-2
-1
0
1
2
3
4
-9
-6
-3
0
3
6
9
12
15
18
2001
2003
2005
2007
2009
2011
2013
EU Membership Premium: IIR (red, lhs) and S&P (blue,
rhs) (Rating points)
-6
-3
0
3
6
9
12
2001
2003
2005
2007
2009
2011
2013
IIR Premium for Euro Area Membership: Arellano/Bond
GMM Dynamic Panel Estimator (Rating points)
(1) (2) (3)
Coeff. t-value Coeff. t-value Coeff. t-value
Controls
Per-capita GDP -1.8^(-5) -0.51 0.1^(-5) 0.58 -1.7^(-5) -0.39
Real growth 0.40 3.63*** 0.12 4.64*** 0.37 3.41***
Investment/GDP 0.13 1.36 0.04 1.74* 0.11 1.09
Unemployment rate -1.03 -9.56*** -0.31 -12.16*** -0.84 -7.13***
CPI inflation -0.26 -3.92*** -0.68 -5.88*** -0.20 -2.46**
Gen. gov. balance -0.10 -0.95 -0.07 -2.69*** -0.00 -0.01
Gen. gov. debt -0.17 -9.01*** -0.03 -6.62*** -0.14 -7.16***
External balance 0.06 0.82 0.04 2.42** 0.08 0.99
NIIP 0.03 5.15*** 0.01 4.10*** 0.02 3.60***
Floating (dummy) 0.16 0.14 -0.50 -1.82* 0.72 0.56
Governance quality … … … … 18.08 5.63***
Country fixed effects, time dummies, euro and EU membership dummies, dummy for missing gov.
debt observations (not reported)
Observations 442 442 374
Adjusted R
2
0.952 0.952 0.956
Significance of country FE F(33,359)= F(33,359)= F(33,292)=
37.63 (0.000) 45.30 (0.000) 22.10 (0.000)
1/
Linearized and inverted
Significance at the 1 (***), 5 (**), and 10 (*) percent level.
Table A1.Euro Premium - Panel Regressions
2001-2013
2002-2012
IIR Rating
S&P Rating
1/
IIR Rating
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 31
Euroization
The analysis of the impact of balance-sheet euroization on perceived credibility is
constrained by limited data availability. We have information for only 11 new EU members
states from 20062012 (in a few cases going back to 2004). This prevents including
euroization as a covariate into the regression equation above. As an alternative strategy, we
country fixed effects
and (ii)
residual credibility premium
with the share of FX loans in total private sector loans.
The estimated correlation (i) is unbiased as long as the share of FX loans is invariant over
time (which for most countries holds at least approximately), but the estimate is based on
very few observations. By adding a time dimension, (ii) greatly enlarges the sample (from i to
observations), but makes an orthogonality assumption about the relationship between
euroization and the other covariates included in the main regression equation above. This
assumption can, in general, not be assumed to hold, giving rise to potential estimation bias
in a priori unknown direction.
Figure A1 shows key results, using the
and
coming out of model (3).
5
While
correlation with euroization is marginally weaker for the residual credibility premium than for
fixed effects, the difference is so small that potential bias is unlikely to materially affect the
analysis. Further, the richer sample obtained from using the residual premium allows
extracting a non-linear relationship from the data: euroization becomes disproportionately
more detrimental for country perception the larger FX balance sheet mismatches are. This
relationship appears broadly stable over timethus, in contrast to the general euro
credibility premium, the FX-specific credibility premium has not vanished with the euro crisis.
5
Euroization correlates strongly with governance quality, thus inclusion of governance quality in X is
necessary lest to overstate the link between euroization and perceived credibility.
CEE NEW MEMBER STATES POLICY FORUM
32 INTERNATIONAL MONETARY FUND
Figure A1. Ratings Premia and Euroization, Detailed Results
Source: IMF Staff calculations.
0
10
20
30
40
50
60
70
80
90
100
-15 -10 -5 0 5 10 15
Fixed Effects (blue/black) vs. Residual Premium (red)
(2006-12)
Residual premium (IIR points)
FX loans to the private sector as share of total
0
10
20
30
40
50
60
70
80
90
100
-15 -10 -5 0 5 10 15
FX loans to the private sector as share of total
Residual Premium: 2006-09 and 2010-12
2010-12
2006-09
Residual premium (IIR points)
CZE
POL
HUN
ROU
BGR
LTU
HRV
LVA
R² = 0.390
0
10
20
30
40
50
60
70
80
90
-15 -10 -5 0 5 10 15
Investor Perception and Balance Sheet Euroization
(Fixed effects, 2006-12)
Average premium (in IIR points)
FX loans to the private sector as share of total
0
10
20
30
40
50
60
70
80
90
100
-15 -10 -5 0 5 10 15
Residual Premium: Non-Linear Specification
(2006-12)
Residual premium (IIR points)
FX loans to the private sector as share of total
R
2
=0.390
R
2
= 0.346
R
2
=0.394
R
2
=0.433
R
2
=0.430
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 33
Appendix II. The Index of Monetary Conditions
Following Freedman (1994) and subsequent applications by several central banks and the IMF, the
index of monetary conditions (MCI) has been calculated as the weighted average of the indexes of
the real interest rate (RIR) and the real effective exchange rate (REER):
MCI = α RIR + (1-α) REER,
in the Czech Republic, Poland, and Slovenia, through 0.8 in Bulgaria, Estonia, Latvia, Lithuania, and
Slovakia, to 0.9 in Croatia and Hungary. For comparison, the European Commission
1
its MCI calculations for the euro area. All indexes were set to 1 in 2003.
1
http://ec.europa.eu/economy_finance/db_indicators/conditions/index_en.htm
CEE NEW MEMBER STATES POLICY FORUM
34 INTERNATIONAL MONETARY FUND
Appendix III. Model-Based Inflation Variance Decomposition
Iossifov and Podpiera (2014) analyze inflationary developments in ten non-euro area EU
countries within the 2004–14 period, using a panel framework of an open-economy New
Keynesian Phillips curve (Galí and Gertler, 1999). They assume inflation to be both forward-looking
with some degree of inertia and driven by demand and supply-side shocks. The regression
specification can be described, for a country i, as:
Headline inflation is calculated using the Harmonized Indices of Consumer Prices (HICP)
published by Eurostat.
Euro Area price pressures is the euro area HICP core inflation, which is stripped of
direct, first-round effects of commodity price changes.
Expected inflation proxies expectations of future inflation by the mean forecasts of
average annual inflation two-years ahead published by Consensus Economics.
it
u
Unemployment rate gap is the cyclical unemployment rate extracted with the Baxter-King
bandpass filter using data from Haver and national sources.
1
vector of country-specific supply-side shocks:
Contribution of taxes and administered prices are captured by their combined contribution to
headline inflation, calculated with HICP data published by Eurostat.
Exchange rate appreciation/depreciation is calculated using the nominal effective exchange
rates published by the IMF.
vector of common external supply-side shocks:
World commodity price inflation—world oil and food price indices in US dollars from the
ic Outlook (WEO) are used to capture commodity price changes. They
are interacted with the weights of energy and food in consumer baskets to allow for
differentiated impact across countries.
1
The Baxter-King filter decomposes, in the frequency domain, the analyzed series into trend, cyclical, and irregular
components, which are additive. For all countries, the Baxter-King filter is based on an 11-quarter centered moving
average and a widely used definition of the business cyclemovements in economic series that occur with
periodicity of between 6 quarters and eight years (32 quarters). In order to obtain estimates for the whole sample
period, we augment the dataset with Fund staff forecasts through end-2016.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 35
Exchange rate regime dummies—based on the classification of exchange rate regimes in the
IMF, 2013a).
Share of foreign value added in domestic demand—calculated using OECD-
Value Added dataset as an average of the 2005 and 2009 values. Data for Croatia is not
available in the OECD-WTO database. We approximate the share of foreign value added in
Croatian domestic demand by the average of its readings in Poland and Romania, as the
ratio of imports to GDP of these three countries are very similar.
The estimation results suggest that food and energy prices account for a large share of the
variance of headline inflation, but exchange rate regime and trade linkages to the Euro Area
are also important. According to the inflation variance decomposition (Figure A2), world food and
energy price changes together with changes in administered prices and taxes account for about half
of the variability of headline inflation across non-euro area EU countries. However, disinflation
spillovers from the euro area have been an important factor for NMS that peg their currency to the
euro and inflation targeting countries with high foreign value added in domestic demand:
Countries with more rigid exchange rate arrangements tend to import more inflation from the
euro area.
Inflation spillovers from the euro area are also larger, the higher the share of domestically
consumed foreign value-added (e.g., in the Czech Republic and Hungary).
Figure A2. Model-based Inflation Variance Decomposition
CEE NEW MEMBER STATES POLICY FORUM
36 INTERNATIONAL MONETARY FUND
OPTING INTO THE BANKING UNION BEFORE EURO
ADOPTION
1
Summary
The main motivation for establishing the Banking Union (BU) was the need to reverse financial
fragmentation that crippled monetary transmission within the common currency area in the wake of
the euro crisis. By design, the BU would raise the credibility of the euro area bank supervision, eliminate
distinction between home and host supervisors, and sever the link between banks and sovereigns. This would,
in turn, lead to lower bank compliance costs, lower barriers to cross-border activity, and lower funding costs
for banks under the BU. As host countries of euro area banks, the NMS-6 would benefit from improved
resilience of the euro area financial system and lower funding costs for parent banks.
The full benefits of the BU will be realized once all its elements are in place, which is not yet the case.
Most notably an effective common backstop is still needed to break the sovereign-bank links. Furthermore,
there is no equal treatment of the eurozone and non-eurozone members of the BU with regards to their
role in the Single Supervisory Mechanism (SSM), access to common (ECB) liquidity support or to a common
fiscal backstop (with ESM currently acting as de facto common fiscal backstop for euro area banks).
When would opting into the BU make sense for the NMS-6? For those new member states (NMS) that
have set a target date for euro adoption (Romania), this amounts to choosing to frontload the phase-in of
some of the necessary institutional changes. For others, the BU opt-in decision requires a careful consideration
of country characteristimodalities and implementation:
BU design: the lack of equal (or fully equivalent) treatment of euro area and non-euro area members of
the BU tilts the NMS-s decision against early BU opt-in and in favor of waiting until euro adoption.
BU modalities: the lack of clarity on and experience with the BU operational modalities which may
affect macroprudential and possibly monetary policy space of the NMS-6 may be another factor in favor
of waiting. This, in particular, applies to coordination between the BU and local supervisors, as well as to
coordination between prudential policies at the national and BU-levels and national monetary policies.
Some may still opt into the BU because for them the BU participation may be a way to address specific
challenges, which outweigh other considerations, including BU shortcomings. Notably, some may see BU
as a way to enhance quality and credibility of bank supervision or to gain access to larger industry-funded
common backstop.
The BU opt-in would be more attractive for the NMS-6 if mechanisms were in place to ensure that the
NMS-6s concerns stemming from unequal treatment of opt-ins and euro area members are fully addressed.
Furthermore, greater clarity on the BU operational modalities that would shape the opt-y space and
the support they could expect from common euro area institutions is needed in order for the NMS-6 to
make a more informed decision on the BU opt-in.
1
Prepared by John Bluedorn, Anna Ilyina and Plamen Iossifov. Min Song and Jessie Yang provided research
assistance. The authors are grateful to Mr. Hampl (the Czech National Bank), Mr. Voinea (National Bank of Romania),
Ms. Szombati (Hungarian National Bank) and Ms. Field (ECB) who participated as discussants in the session on opting
into the Banking Union before euro adoption at the New Member States (NMS) Policy Forum in Warsaw on
December 12, 2014, and to the NMS-6, EC and ECB representatives who provided comments during bilateral
discussions in November 2014. The authors are grateful to Giovanni Dell'Ariccia (IMF) for helpful discussions.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 37
A. Why Did Europe Need a Banking Union?
“With a European supervisor, borders will not matter. Issues such as protecting national champions or
supervisory ring-fencing of liquidity will not be relevant.” (M. Draghi)
1. The global financial crisis exposed weaknesses in the EU financial architecture, arising
from misalignments between national mandates for financial sector oversight and the EU-
wide operations of many market participants:
Negative externalities: The pursuit of domestic financial stability and competitiveness
objectives, as well as resident taxpayer interests can create negative externalities for other EU
members, resulting in a sub-optimal Union-wide outcome. One example is the failure of home
supervisors of banks with subsidiaries in Central and Eastern Europe to rein in credit expansion
in the region, which fueled unsustainable domestic demand booms prior to 2008. Host
from loc et al, 2005). Another example is
the bailout of companies from the financial conglomerate Fortis Group according to their
country of incorporation, instead of restructuring on a consolidated basis (BIS, 2010).
Financial fragmentation: The national nature of deposit insurance schemes and public
backstops for financial institutions led to a post-crisis fragmentation of the European market for
financial services, as the funding costs of financial intermediaries and ultimately the cost of
borrowing became linked to sovereign creditworthiness (ECB, 2012). As a result, a number of
countries became caught in a negative feedback loop between bank solvency and sovereign
default risks, posing a major challenge for euro area countries which do not have country-
specific monetary autonomy (IMF, 2013a).
2. In the aftermath of the crisis, the EU embarked on ambitious financial sector reforms
aimed at improving the transparency and health of the financial system, strengthening and
harmonizing bank supervision and resolution, reducing market fragmentation, and minimizing the
cost to taxpayers of future bail-outs. Given the special challenges faced by euro area members, the
reform strategy has proceeded along two parallel tracks:
Harmonization of the regulatory and supervisory regimes for all participants in the single
market for financial services. To this end, the European System of Financial Supervision (ESFS)
was put in place in 2011 and the Single Rulebook was developed to harmonize prudential norms
for all EU banks (EC, 2013):
CEE NEW MEMBER STATES POLICY FORUM
38 INTERNATIONAL MONETARY FUND
o The European System of Financial Supervision comprises the European Banking Authority
(EBA), the European Systemic Risk Board (ESRB)
2
, European Securities and Markets Authority
(ESMA), European Insurance and Occupational Pensions Authority (EIOPA), the Joint
Committee of the European Supervisory Authorities (ESAs), and national supervisory
agencies.
o The core of the Single Rulebook is now in place, although some elements are to be phased in
gradually over time. The Capital Requirements Directive (CRD IV) and the Capital
Requirements Regulation (CRR)which harmonize capital definitions and implement Basel
IIIwere adopted in mid-2013. Work is ongoing on binding technical standards for
implementation of CRD IV/CRR, as well as on other chapters of the rulebook, including
further harmonization and strengthening of deposit guarantee schemes.
o The Bank Recovery and Resolution Directive (BRRD) and Deposit Guarantee Scheme Directive
(DGSD) were adopted in mid-2014. BRRD establishes baseline bank restructuring and
-
liabilities (conversion of liabilities to equity) as a means of reducing the contingent liability
for taxpayers (bailouts) in cases of resolution. Similarly, DGSD sets out minimal requirements
for the operation of national deposit guarantee schemes, ensuring they are funded ex ante
and pay out in a timely manner when needed.
o The next step is the development of a Single Supervisory Handbook and further
harmonization of supervisory practices to ensure the uniform implementation of the Single
Rulebook.
Building upon the EU-wide reforms above, euro area countries established common bank
supervision and resolution regimes—Banking Union (BU)—which is open to non-euro area
EU member states. The architecture of the BU includes the Single Supervisory Mechanism
(SSM) and the Single Resolution Mechanism (SRM), which centralize bank supervision and
resolution powers. Importantly, the other key elements of the BU a truly common fiscal
backstop and a common deposit guarantee scheme are not yet in place.
The rest of this chapter focuses on the pros and cons of participating in the BU prior to euro
adoption for the NMS-6 that are members of the EU, but are not yet part of the euro area.
2
The ESRB is responsible for the macro-prudential oversight of the financial system within the Union in order to contribute to the
prevention or mitigation of systemic risks to financial stability arising from developments within the financial system (ESRB
Regulation). The ESRB is also tasked with assessing national macroprudential frameworks and ensuring effective coordination and
internalization of cross-border spillovers. However, it does not have enforcement powers.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 39
Common bank supervision
Common bank resolution
Common fiscal backstop
Common deposit guarantee scheme
B. Banking Union Modalities and What an Early “Opt-In” Entails
“We have to consider that opt-in countries, as opposed to their partners from the euro area, don’t have
equal rights in the Banking Union”. (M. Belka)
3. Despite significant progress, not all
elements of the BU are yet in place. While both
SSM and SRM are now operational, an effective
common fiscal backstop is still needed to break the
sovereign-bank links (though ESM is currently
acting as de facto common fiscal backstop for euro
area banks see Box 1 for discussion of the BU
modalities). Other key elements include allowing
the Single Resolution Fund (SRF) (which will be fully
funded and mutualized only by 2024)
to borrow
against future industry levies and working towards
a pan-European deposit guarantee scheme (DGS)
3
.
What does “opting into the BU” entail?
4. The BU membership refers to participation in the Single Supervisory Mechanism (SSM)
and the Single Resolution Mechanism (SRM). For non-euro area countries
would entail entering into a close cooperation with the ECB (Article 7 of the SSM Regulation) and
passing any required national legislation to enable national authorities to work with the ECB and the
Single Resolution Board (SRB) under their supranational frameworks for supervision and resolution:
Entry into close cooperation with the ECB.
into national laws of the relevant EU legislation, and can request additional information for the
purposes of the envisioned comprehensive assessment of banks. Whereas the outcome of the
application is not conditional on the results from the comprehensive assessment, the ECB can, to
a certain extent, use its powers to request further information and carry out its own
comprehensive assessment, as levers to steer the process.
Exit option for non-euro area members: Unlike members of the euro area, non-euro area
participating member states have the option to suspend or terminate the close cooperation with
the ECB, and thus their participation in the BU. At the same time, the ECB also has the option to
suspend or terminate close cooperation with a non-euro area participating member state, if it is
3
The proper functioning of the SRM will depend on whether there is adequate backstop. Yet the SRF is not yet fully in place. All
banks in the BU countries will contribute to the SRF
banking union also presupposes a common deposit insurance scheme, which does not (yet) exist. So far, there is only a voluntary
mechanism of mutual borrowing between deposit guarantee schemes from different EU member states.
Not Yet a Perfect Banking Union
CEE NEW MEMBER STATES POLICY FORUM
40 INTERNATIONAL MONETARY FUND
determined that the member is not fulfilling their obligations. Resolution actions undertaken
prior to becoming a participating member state would not be covered by the SRM.
5. Acting as a de facto common fiscal backstop, ESM bank recapitalization will not be available
for any non-euro area BU participants, since the ESM Treaty is only open to currency union
members. The lack of an effective common fiscal backstop means that sovereign and bank risks can
become intertwined, especially in times of stress.
Participation in the SSM and SRM
6. After opting into the BU, non-eurozone members would have representation on the
Supervisory Board (SB) (on par with the euro area member states)
4
, but the modalities of their
participation in the decision-making process would differ from the eurozone members. Within
the SSM, the SB will manage oversight and make draft decisions. The draft decisions will be referred
to the ECB Governing Council, as the overarching authority, that can either automatically adopt the
decision under non-objection procedureor object to it (see Box 2 for details on the SSM
modalities). The non-euro area member states of the BU who do not have representation on the
ECB Governing Council – would be invited to send representatives to the ECB Governing Council, if
the ECB contemplates an objection to an SB draft decision or if the non-eurozone members
disagree with a draft decision of the SB. If no satisfactory compromise can be found in the
subsequent reconciliation process, the non-euro area member state can notify the ECB that it will
reasoned disagreement
can result in the eventual suspension or termination of the member sration with the ECB
in the SSM (per Article 7, SSM Regulation).
7. After opting into the BU, non-eurozone members would also have representation on
the Single Resolution Board (SRB) and would contribute to and have access to the Single
Resolution Fund (SRF) (see Box 3 for details on the SRM modalities).
Emergency Liquidity Assistance (ELA)
8. Central bank provision of ELA is not affected by the BRRD or by the SRM, as it was and
remains a national prerogative. ELA is extended to solvent institutions (that is, those without a
capital shortfall identified by the supervisor), subject, inter alia, to systemic importance and
interconnectedness considerations. These rules are applicable to all EU members.
5
However, for BU-
participating states, it will be the ECB in its supervisory capacity under the SSM that will determine
whether a bank is solvent or not, and thus its eligibility for ELA. Importantly, unlike eurozone
members, non-euro area BU participating members would not be entitled to supplementary access to
the ECB’s liquidity facilities. At present, any liquidity provision by the ECB to non-euro area members
4
The SSB includes one representative from each member state plus 5 ECB representatives (in their personal capacity) see Box 2.
5
Communication 2013/C 216/01) for further details.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 41
via repo or swap lines is granted on a country-by-country basis and subjugated to monetary policy
considerations.
Key takeaways: the modalities of the non-eurozone members’ participation in the BU are
notably different from those of the euro-area members: (i) role in the SSM: non-euro
decisions drafted by the Supervisory Board; (ii) fiscal backstop: non-euro area opt-ins are not
eligible for direct bank recapitalization from the ESM (acting as de facto common fiscal
backstop); and (iii) liquidity support: non-euro area opt-ins would not automatically have access
to the ECB liquidity facilities. That said, as a compensation for this unequal treatment, the BU
offers some safeguards for the non-euro area opt-ins, such as the possibility to present “reasoned
disagreement” and to exit the BU.
9. Certain features of their banking systems suggest that the NMS-6 would be
particularly sensitive to the lack of equal (or fully equivalent) treatment within the BU:
Ability to influence decisions related to parent banks is critical for the NMS-6 because most of
their banking systems are dominated by euro area bank subsidiaries, which tend to be more
important for local economies than for the parent banking groups (see text charts). If under the
BU all/most barriers to cross-border transfers of capital and liquidity are indeed removed, this
also means that local authorities would have less power to ring-fence. The latter raises the
importance of being able to influence the activities of parent banks through other means,
including via participation in the SSM decision-making process. Given that any decision
regarding cross-border banks will have to weigh prudential considerations of host and home
countries, any weakness in the ability to influence such decisions raises concerns that these
decisions may be tilted in favor of larger financial systems/institutions which have a greater
bearing on the financial stability of the BU as a whole.
Access to common liquidity and fiscal backstops is important for the NMS-6, because: (i) many still
have large external liabilities, though many NMS subsidiaries are now less reliant on foreign
parent bank funding than before the crisis (see Box 1 in the staff report); and (ii) banks in NMS-6
typically have less bail-inable funding (other than uninsured deposits) than eurozone banking
groups operating in the region. NMS-6 are, therefore, more likely to benefit from the risk-
sharing aspect of the SRF or other common backstop (see chart below).
10. At this stage, it is not clear how effective the safeguards for the non-euro area opt-ns
would be. Given that the SSM has been set up very recently and there is yet no experience with the
process, there are different views on whether the modalities of the BU opt-
SSM decision-making process can adequately protect their interests.
6
The exit clause may not be an
6
For example, Tröger (2013) argues is important
a significant difference, if a representative of the affected Member
deliberations or if the Member State has to rely dégagé on the benevolent consideration of a position articulated ex ante). In
contrast, Darvas and Wolff (2013) consider that this mechanism gives the non-euro area participating member a sufficient voice in
(continued)
CEE NEW MEMBER STATES POLICY FORUM
42 INTERNATIONAL MONETARY FUND
effective safeguard, if it is not used in practice due to significant negative reputational effects. The
question of whether, on balance, the opt-ins would gain or lose influence on decisions regarding
parent banks is complex and is discussed in more details below.
Three largest banks by assets, 2013
(Percent of GDP)
Assets of largest foreign-owned banks in NMS-6
(Individual bank assets)
Sources: Bankscope; and IMF staff estimates.
Note: Top 3 banks would be expected to come under SSM.
Sources: Bankscope; and IMF staff estimates.
Note: In some cases, the source data are consolidated for the financial
group, in which the bank is part of.
Funding Structure of Top-5 Banks in NMS-6 and Select Parent Bank Groups, 2013
(Percent of total liabilities)
Sources: Bankscope; and IMF staff calculations.
decision making, since it contains a special opt-out clause in the event of a disagreement with the ECB in its supervisory
role. The NMS-6 survey (see Box 5 in the main report) suggests that this is source of discomfort for many NMS-6.
0%
10%
20%
30%
40%
50%
60%
70%
80%
CZE
HRV
HUN
BGR
POL
ROU
Domestically owned
Foreign owned
0%
5%
10%
15%
20%
25%
30%
35%
Percent of GDP
Percent of parent assets
BGR
HRV CZE HUN
POL ROU
0
20
40
60
80
100
United Bulgarian
UniCredit Bulbank
DSK
CCB
First Investment
Komercni
Unicredit, CZ&SK
Sporitelna
Obchodni
HAA
PBZ
Raiffeisenbank,AT
Zagrebacka
ESB
OTP
K&H
Erste, HU
MKB
Raiffeisen Bank
Zachodni
Pekao
PKO
mBank
ING Bank Slaski
BCR
Raiffeisen, RO
BRD
Tiriac
Transilvania
Erste Group
BCP
Intesa Sanpaolo
Santander
UniCredit
KBC Group
Commerzbank
Rabobank
BG
CZ
HR
HU
PL
RO
Foreign-parent bank
groups
Equity
Subordinated borrowing
Senior debt
Deposits
Other funding
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 43
Macroprudential Policy Space and Policy Coordination
11. The CRR/CRD IV legislative package defines a range of tools over which national
macro-prudential authorities may set stricter requirements (above the industry-wide, micro-
prudential minima) based on systemic risk considerations, macro-prudential concerns, or to address
risks at individual firm level. National macroprudential authorities retain full control over
macroprudential measures, not specified in Union law, such as the loan-to-value and debt-to-income
ratios, among others (see Box 4).
12. All EU member states are required to notify the ESRB of changes in their
macroprudential policy stance. The ESRB then provides opinions on the proposed policies
regarding their financial stability and growth implications (at both the national and EU levels).
explain measures, questioned by the ESRB, introduces an additional, albeit soft, check-and-balance.
13. For BU members, the SSM entails some additional constraints on macroprudential
policies. Under the SSM Regulation (Article 5), national competent authorities (NCAs) can still
deploy macroprudential measures as they deem appropriate, subject to a notification requirement
to the ESRB.
7
However, in the case of CRR/CRD IV measures (see Box 4), BU-participating states must
also notify the ECB of their intention 10 working days prior to issuance of their decision. If the ECB
objects, then it supplies a written explanation within 5 working days, which the national authority
must take into consideration. Furthermore, if the ECB wishes, it may apply stricter macroprudential
requirements on banks, irrespective of whether they are under direct SSM supervision or not, than
the national authorities (subject to similar notification and consideration timelines). In other words,
macroprudential policies the ECB may always strengthen macroprudential policies set out in
relevant Union law, but it cannot compel loosening. On their part, NCAs can also tighten those
norms, but cannot loosen them
Key takeaway: The extent to which joining the BU would limit macroprudential policy space,
unless fully offset by lower likelihood of shocks or additional support from the euro area
institutions, would need to be taken into account. At present, a full assessment is complicated by
the lack of clarity on and experience with relevant operational modalities of the BU, including
the mechanisms for policy coordination between supranational and national levels.
7
Proposed national measures are analyzed by an Assessment Team consisting of two representatives of the ESRB Secretariat,
representatives from nine European Union national central banks, one representative of the ECB, and one representative of the SSM.
See the 2013 Annual Report of the ESRB (ESRB, July 2014).
CEE NEW MEMBER STATES POLICY FORUM
44 INTERNATIONAL MONETARY FUND
C. Banking Union Opt-In: Pros and Cons for Non-Euro EU Countries
14. The main motivation for establishing the BU was the need to reverse financial
fragmentation that crippled monetary transmission within the common currency area in the
wake of the euro crisis. The establishment of the SSM (supported by the SRM) would raise the
credibility of the euro area bank supervision, eliminate distinction between home and host
supervisors for cross-border banks, and sever the link between banks and sovereigns. This is
expected to lead to lower bank compliance costs, the removal of any barriers to cross-border
activity which may be in place to protect national interests,
8
and lower funding costs for banks under
the SSM supervision. That said, the full benefits of the BU will only be realized once all the BU
elements are in place (as discussed above).
15. Because of significant presence of euro area banks in all NMS-6 countries, a fully
established BU will have positive implications for the NMS-6. As host countries of euro area
banks, the NMS-6 would benefit from improved resilience of the euro area financial system and
lower funding costs for euro area banks:
Provided that euro area banks become safer and more conservative (under more consistent
supervision in the BU), the likelihood of negative spillovers for the NMS-6 from the euro area will
be lower.
Provided the BU will lead to greater fungibility of liquidity within the euro area, the funding costs
for cross-border banking groups will be lower,
9
which may help boost lending to faster-
growing/higher-return NMS.
Home-host interactions may become simpler with the euro area single supervisor. That said, the
key question for the NMS is whether the SSM itself, which will supervise most of the cross-
border banking groups with operations in the NMS, will treat bank exposures to counterparties
inside and outside the BU differently.
16. At present, direct participation in the BU prior to euro adoption is generally less
attractive for the NMS-6 than for euro area countries, given that the BU remains incomplete and
non-euro area members do not enjoy the same treatment as the euro area members. An imperfect
BU could work in practice for euro area countries, given that they have access to the ECB liquidity
facilities and to the ESM, but for non-euro area NMS that do not have either joining the BU
before adopting the euro is less attractive. This is because the amount of decision-making power
8
Some analysts note, however, that it yet remains to be seen whether the ECB will be able and willing to change the currently
reported ring-fencing of banking activities.
9
Removal of barriers to cross-border transfer of capital and liquidity would reduce the required capital and liquidity buffers at the
subsidiary level (see Cerutti and others (2010)).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 45
that they cede to the supranational level would be similar to that of the euro area countries
10
, but, in
return, the NMS-6 would receive less support from common liquidity and fiscal backstops than the
euro area countries.
17. In addition to the BU modalities (discussed above), whether an NMS-6 country would still
be better off in the BU than outside would depend on country characteristics, policy
preferences and BU’s operational modalities (see Box 5 for discussion of insights from theoretical
literature). Some of the key considerations are as follows:
The NMS-6 risk-sharing preferences depend on the country-specific characteristics that affect the
types of shocks they are likely to face (e.g., economies that are less integrated with the euro area
and hence, more likely to face asymmetric shocks, may derive greater benefits from having
access to common backstop).
The NMS-6 policy preferences may influence the desired stringency of prudential standards.
While lower incidence of financial instability is growth-enhancing, at any given time, policy
stringency required to contain systemic risks (e.g., tighter standards reduce the risk and cost of
financial instability, but also dampen credit growth and lower bank profitability).
11
The NMS-6
s.
The policy space configuration of the BU opt-ins is yet to be fully defined. The NMS-6 need to
consider that joining the BU may affect their macroprudential and monetary policy space.
18. The rest of this section focuses on the interaction between various country
characteristics and BU modalities and provides examples of key trade-offs for different types
of countries. Table 1 presents a taxonomy of country characteristics (top row) and policy objectives
(first column) and whether joining the BU could help or hinder the achievement of these objectives
(column showing potential benefits and costs). The cells in the matrix indicate which of the country
characteristics are likely to be associated with relative benefits or costs. The country characteristics in
Table 1 are the ones that are most relevant for the decision to join the common currency area or the
common regulatory area based on the literature:
The degree of real or financial integration with the euro area (columns 1 and 3) determines the
relative likelihood of common versus asymmetric shocks and hence, risk-sharing preferences.
The degree of economic flexibility (column 2) reflects the ability of the economy to absorb
shocks; less flexibility makes it more likely that negative shocks could trigger financial instability.
10
All banks that come under the SSM supervision have to satisfy the same criteria on systemic significance (Box 1).
11
The relative weights that national authorities place on different considerations may vary across countries, depending on the
institutional setup of financial sector oversight (its independence and accountability), the type of financial system (bank versus
market-based), ownership of the banking sector , the degree of market concentration. Weights may change through the cycle.
CEE NEW MEMBER STATES POLICY FORUM
46 INTERNATIONAL MONETARY FUND
The share of local bank assets owned by the euro area banks (column 4) indicates the importance
of intra-group cross-border flows of euro area banks for domestic financial stability.
The supervisory standards (column 5) refer to the stringency of rules and quality of supervisory
processes at the local level.
Local backstops for the financial system include local DGS (column 6) and fiscal policy space
(column 8) and refer to the national capacity to absorb shocks. Their adequacy is inversely
potential exposure to contingent liabilities, as measured by the ratio of
insured deposits to GDP, and the size of public debt relative to GDP.
Policy space indicates the availability and effectiveness of monetary and fiscal policies (columns
7 and 8), as tools for demand management. Policy space can be proxied by the ratio of public
debt to GDP, whereas the availability of monetary policies depends on the nominal anchor
(exchange rate versus inflation) chosen by the central bank.
Would joining the BU reduce financial stability (FS) risks for the new members?
YES, if joining the BU
Improves the overall quality/stringency of supervision. To the extent that supervision under
the SSM will be stricter than current national supervision, banks would be safer and financial
stability risks would be lower. This would be the case, if the SSM: (i) sets micro-prudential
standards for local banks that are at least as strict as the current standards in force in the new
members (see Box 5); and (ii) succeeds in distancing supervision from the influence of local
-owned banks (see Box 6). In order
for these benefits to accrue, it is critical for the SSM to establish early a strong track record. That
said, differences in legal and accounting standards across members would complicate
harmonized supervision in the BU. New members with less stringent supervisory standards and
those with weaker local backstops would benefit more (Table 1, Columns 5, 6, and 8, Ranks: Low).
Limits negative externalities stemming from the actions of current BU member banks. The
participation of the non-euro area countries in the BU could further reduce the scope for
regulatory arbitrage and leakages of macroprudential measures aimed at safeguarding financial
stability in member countries.
12
The possibilities for regulatory arbitrage have already been
reduced through the Single Rulebook, but the SSM would ensure compliance through
centralized supervision and greater harmonization of supervisory practices. New members with
strong financial links with the euro area, and a significant presence of BU member banks (Table 1,
Columns 3 and 4, Ranks: High), as well as those with less stringent supervisory standards and
weaker local backstops (Table 1, Columns 5, 6 and 8, Ranks: Low) would benefit more.
12
As discussed in the introduction, the macroprudential measures adopted by the local authorities to slow rapid
credit growth in CESEE countries during the pre-crisis boom were often not very effective because they were not
matched by similar measures by the home country supervisors of euro area banks operating in CESEE countries.
Table 1. Benefits and Costs of Joining Banking Union for Non-Euro Area Countries
Note: The table presents a simplified taxonomy of country characteristics (top row) and policy objectives (first column) and whether joining the BU could help or
hinder the achievement of these objectives (rows showing potential benefits and costs). The cells in the matrix indicate whetng on a given
country characteristic (in columns) has a material impact on the benefits or costs of joining. For example, the degree of real or financial integration with the euro
area (columns 1 and 3) affects the relative likelihood of common versus asymmetric shocks, with lower integration = higher likelihood of asymmetric shocks and
hence costs of giving up local policy space to respond to them. Types: for each characteristic listed in the top row, a country can be of two types: High at or
above the average across BU members; and Low -
f joining accrues independent of whether a country ranks low
or high on a particular country characteristic).
High Low High Low
High Low
High Low High Low High Low High Low High Low
1. Improve the overall quality of
supervision
+ + +
2. Limit negative externalities from
euro area banks
+ + + + +
3. Increased access to info and
improved home-host coordination
through SSM
+ +
4. Reduce ability to mitigate country
specific shocks
- -
- - -
5. Constrain ability to control cross-
border intra-group flows
- - -
6. Increase efficiency and lower cost
of cross-border bank resolution
+
7. Provide access to common,
industry-funded backstop (SRF)
+ +
+ + +
8. Loss of some local control over
resolution process in the absence of
fiscal backstops
- - - -
Growth
9. Reduce ability to smooth credit
cycles through prudential measures
- -
- - - -
Policy Space
(7) Monetary
(8) Fiscal
(5) Supervisory
standards
Financial stability
Likelihood of distress
Cost of distress
(6) Industry-funded
backstops (DGS)
Policy objective
Real Sector
Financial Sector
Supervision/Backstops
(1) Degree of real
convergence/
integration with
the euro area
(2) Degree of labor
& product markets
flexibility
(3) Degree of
financial
integration with
the euro area
(4) Banking system
structure (share of
banks owned by
euro area banks)
Benefit or Cost of Joining the
Banking Union
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND INTERNATIONAL MONETARY FUND 47
CEE NEW MEMBER STATES POLICY FORUM
CEE NEW MEMBER STATES POLICY FORUM
48 INTERNATIONAL MONETARY FUND
Better access to information and better home-host coordination through direct
participation in the SSM.
13
Joining the BU would provide non-eurozone members: (i) greater
access to supervisory information on cross-border banks operating in their jurisdictions (and
also in other jurisdictions);
14
and (ii) ability to directly participate in the SSM/SB decision making
process, though acting in their personal capacities for the good of the Union, rather than for national or
group interests. There is a range of views on whether this would ultimately give NMS greater
leverage over decisions regarding parent banks. On the one hand, as a member of the SB, the
NMS representative would be able to vote on all issues, including the ones that are currently
beyond the purview of local supervisors.
15
On the other hand, because of different treatment of
the euro area and non-euro area members of the SSM (discussed above), the ability of NMS to
influence decisions may be weaker than that of the euro area members. Another important
issue is that after opting into the BU, the new member would no longer have the final say on
certain matters that are of particular importance to them (e.g., local liquidity requirements –see
below). Hence, the net gain/loss of influence on the decisions regarding parent banks would
depend not only on the NMS’ role in the SSM, but also on how much control the new member
will de facto cede by joining the BU. New members with strong financial links with the euro area
and a significant presence of the BU member banks would benefit (Table 1, Columns: 3 and 4,
Rank: High).
NOT necessarily, if joining the BU
Limits the ability to use prudential tools to address country specific shocks, to the extent
that the loss of powers in not compensated by a commensurate decline in the frequency or size
of such shocks. Under the Single Rulebook, local supervisors have significant flexibility to impose
additional macro- and microprudential requirements, early intervention powers and ability to set
conditions under which the local CB could provide liquidity assistance to troubled banks. After
joining the SSM, some of this flexibility (including “good” discretion) could be lost. For example,
in the event a NMS is hit by an asymmetric shock, SSM’s prudential requirements may end up
being stricter than might be warranted given country-specific circumstances, which could lead to
13
Prior to the BU, cross-border coordination of banking supervision of a banking group would occur via a college of supervisors,
involving supervisors from those jurisdictions spanned by the group. The college would provide a venue for interactions between
supervisors across countries to facilitate information sharing and coordination (particularly in emergencies or cases of restructuring
or resolution). A key innovation of the BU is the removal of this institutional layer for coordination between its members.
14
Being part of the supervisory college, non-euro area member can request any information about parent banks that it deems
relevant. Because there is a need to request information, access to information may not always be as timely as desired. In
comparison, being part of the SSM would automatically grant access to all info about the parent bank as well as other euro area
banks.
15
Currently, the extent to which local supervisor is able to influence any given decision depends on the specific issue under
consideration and who has competency over this issue. E.g., in the case of capital/liquidity requirements at the group level, if a
home supervisor decides to increase the requirements for the whole group, the host supervisor cannot block this decision; in the
case of capital/liquidity requirements at the subsidiary level, the host supervisor has the final say.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 49
higher (than optimal) incidence of bank closures or to lower recovery values on distressed assets
(less of “good forbearance”). This consideration is most relevant for countries that are relatively
less integrated with the euro area and hence more exposed to asymmetric shocks (Table 1,
Columns 1 to 4, Ranks: Low), as well as for supervisors with greater capacity to intervene (Table 1,
Column 5, Rank: High).
Leads to loss of full control over cross-border capital and liquidity flows, to the extent that
the loss of powers in not compensated by a commensurate reduction in the likelihood of
negative spillovers or in the absence of alternative mechanisms for dealing with such spillovers.
Ring-fencing of capital and liquidity of the euro area banks’ subsidiaries was used by national
supervisors during the crisis to prevent problems in foreign parent banks from spilling over to
the domestic banking systems. After joining the BU, local supervisors will lose control over the
liquidity requirements at the subsidiary level, though they will retain the ability to set large
exposure limits.
16
To the extent that BU would completely eliminate any negative externalities,
the NMS supervisor should not be concerned about losing the ability to ring-fence after joining
the BU. However, to the extent that some spillovers remain a possibility, national supervisors
may perceive a loss of control over cross-border intra-group flows as potentially increasing the
risk of financial instability. These considerations are most relevant for counties where the euro area
banks’ subsidiaries dominate in the local banking market (Table 1, Columns 3and 4, Ranks: High),
as well as for supervisors with greater capacity to intervene (Table 1, Column 5, Rank: High).
Would joining the BU reduce the cost of financial distress, once it occurs?
YES, if joining the BU
Increases efficiency and reduces the cost of bank resolution. The BRRD already goes some
way towards achieving this objective, but the SRM further ensures that the process of winding
down of large cross-border banks is orderly and “least cost” on a consolidated basis. This is a
positive factor for all, but especially for those countries that host subsidiaries of euro area banks
(Table 1, Column 4, Rank: High).
Provides access to common backstop (SRF). Joining the SRM allows the NMS banks to have
access to a larger backstop without adding to the fiscal burden of the sovereign. Having access
to a common backstop (SRF) would be relatively more attractive for countries that are more
likely to be hit by asymmetric shocks and those with weaker local backstops.
17
However, these
16
While in a BU it will be much harder for host supervisors to block intra-group cross-border transfers, there are still some powers
that are given to member states that could be viewed as safeguards. E.g., there is large exposure regime in the CRR and there are
two discretions: one given to supervisor and the one that allows member states to impose large exposure limits (Article 493). The
supervisory decision can never overrule the decision of a member state.
17
The logic is similar to that of the optimal currency area literature. In the Mundell II models (surveyed in Tavlas, 1993 and
McKinnon, 2004), differences in economic structure, lack of diversification, and high volatility of terms-of-trade increase the appeal
of a monetary union, as they increase the benefits of pooled foreign reserves and integrated capital markets.
CEE NEW MEMBER STATES POLICY FORUM
50 INTERNATIONAL MONETARY FUND
benefits are limited until the SRF is fully mutualized. The national contributions to the SRF will be
only gradually mutualized over the course of the next eight years, reducing the appeal of this
aspect of BU membership in the interim. Hence, less integrated countries (Table 1, Columns 1 to
3, Ranks: Low) and those with weaker local backstops (Table 1, Columns 6 and 8, Ranks: Low)
would derive the biggest benefit once the fully mutualized backstop is in place.
NOT necessarily, if joining the BU
Leads to some loss of local control over the resolution process, without commensurate
risk-sharing on supra-national level. Once a non-EA member joins the SRF, the decision on
whether or not to resolve a bank under SSM supervision will be taken at the BU level. Until the
SRF is fully mutualized, this raises the risk that the resolution decision may not fully take into
account available financing (for resolution purposes), as the latter would still largely consist of
local DGS and local fiscal backstop. In addition, there is a risk that the SSM will apply stricter
criteria (than might be warranted by local conditions) in determining whether a bank is solvent
or not, which would lead to higher incidence of resolution under the BU. This consideration is
most relevant for countries with strong supervision (Table 1, Column 5, Rank: High), those in which
subsidiaries of cross-border banks that would be resolved directly by the SRM have significant
market share (Table 1, Column 4, Rank: High), as well as countries with less adequate local
backstops (Table 1, Columns 6 and 8, Ranks: Low).
18
Would joining the BU facilitate or hinder achieving macroeconomic objectives?
Macroeconomic considerations. Joining the BU could reduce the national policy makers’ ability
to support access to credit through prudential measures, particularly when country specific
circumstances require more supportive financial regulation than in other BU members.
19
This is
partly an artifact of the asymmetry between the powers of the ECB and national supervisors to
tighten and loosen prudential norms: (i) national prudential norms can only be stricter than the
floor set by the ECB; and (ii) the ECB may always strengthen macroprudential policies, but it
cannot compel loosening. While in principle, the ECB does not have to set the same
macroprudential standards across all BU members, it is not clear how much heterogeneity it may
be prepared to accept given its objective of ensuring level playing field and preventing
regulatory arbitrage. This consideration is most relevant for less integrated economies that are
more likely to find themselves facing different cyclical conditions than the rest of the BU (Table 1,
18
In addition, initial conditions may matter as well. If asset quality, liquidity and profitability of local subsidiaries of euro area banks
are stronger than in the rest of the banking group, local stakeholders would be worse off if a banking group is resolved at BU-level
(on a consolidated basis) rather through the local resolution process. While this consideration is not relevant in a steady state, it
may provide a disincentive to joining the BU from a position of relative strength.
19
GFSR (2013) notes that during the latest crisis, a number of European countries used prudential measures to enhance credit
supply, including a reduction in risk weights for small and medium enterprise loans when calculating banks’ capital adequacy ratios,
forbearance of nonperforming loans, and countercyclical macroprudential regulations.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 51
Columns 1 to 4, Ranks: Low), as well as for supervisors with greater capacity to intervene (Table 1,
Column 5, Rank: High).
Does monetary policy autonomy make a difference?
19. All BU members, including those in the eurozone, retain some policy instruments (for
example, taxes and subsidies, housing policies, and so on) that could potentially be used to
offset the impact of measures adopted at the BU level. However, non-euro area members will
have an additional tool—they will retain sovereignty over monetary and exchange rate policies.
20
In
the BU, these national policies would need to be coordinated not only with prudential measures
21
taken at the national but also at the BU level. Independent monetary policy provides an additional
policy tool to manage the impact of shocks on the economy that could, in principle, allow a non-
euro area BU member to take advantage of the upsides offered by the BU, while mitigating potential
downsides.
22
Perspective BU members without independent monetary policy will, hence, be at a
disadvantage relative to their inflation-targeting peers (Table 1, Column 7, Rank: Low).
Key takeaways: Looking at purely economic considerations, some NMS-6 may want to opt in
because they see potential benefits from addressing specific challenges through participation in
the BU as far more important than all other considerations, including BU shortcomings. Notably,
some NMS may see BU as a way to enhance quality and credibility of bank supervision or
distance local supervision from local vested interests
23
or to gain access to larger common
(industry-funded) backstops. In general, the BU opt-in decision is complex, with many moving
parts and requires careful consideration of country characteristics, policy preferences and BU’s
modalities. Some trade-offs are illustrated below:
o Economies that are less integrated with the euro area and hence more likely to find
themselves facing different cyclical conditions than the rest of the BU (e.g., Bulgaria,
Croatia) face a trade-off between gaining access to a larger industry-funded common
backstop (SRF) and giving up some flexibility to deal with country specific shocks. While the
upside will fully materialize only once the SRF is fully mutualized, the downside can be
20
Monetary policy remains a national responsibility prior to euro adoption, but is subordinated to EU Treaty obligations. In
particular, its main objective should be price stability, with exchange rate policy being treated as a matter of common interest.
21
E.g., according to the IMF (2013c) monetary policy measures that were adopted during the recent crisis by a number of European
countries with the explicit objective of easing constraints on credit supply included direct and indirect credit easing as well as
widening of collateral eligibility for private sector assets.
22
In the more extreme case, the use of monetary policy for financial stability rather than price stability goals may undermine its
credibility. There are also other practical policy coordination challenges, such as interaction between multiple central banks (LoLR,
collateral requirements) and coordination of monetary and competition policies (see Scherf, 2014).
23
In Box 4, we infer the policymakers’ revealed preference between financial stability and support of domestic banks by the extent
to which they allow sizable gains in the market share of domestically-owned banks to occur alongside with these banks paying a
bigger premium in funding costs than foreign-owned banks.
CEE NEW MEMBER STATES POLICY FORUM
52 INTERNATIONAL MONETARY FUND
properly assessed only when there is more clarity on and experience with the relevant BU
operational modalities.
o Economies where the euro area banks dominate local banking systems (e.g., Czech
Republic, Croatia) face a trade-off between direct participation in the SSM deliberations
(which entails better access to information and ability to participate in the decision-making
on parent banks) and ceding full control over intra-group cross-border capital and liquidity
flows (ability to ring-fence). The big unknown here is the extent to which negative
externalities stemming from the activities of the euro area cross-border banks would indeed
be effectively eliminated under the BU, as this would determine the value of having control
over the intra-group cross-border flows for local authorities.
o Countries with monetary and exchange rate flexibility would need to better understand
how the centralization of micro- or macroprudential powers under the BU would affect their
ability to conduct monetary policy/lender-of-last-resort functions effectively. While the non-
eurozone BU opt-ins could, in principle, use their monetary policy/exchange rate flexibility to
offset tighter macroprudential requirements set at the BU level, in practice, this could lead to
tensions that would need to be resolved.
How would participation by one or more NMS change the “opt-in” calculus for others?
20. The opt-in decision by one NMS may have a bearing on the decision(s) of other
NMS(s), if it creates a competitive advantage for the NMS that joins but makes others worse off
compared to the status quo. This, however, requires an assumption that there is a limited pool of
capital and funding dedicated to the region by the euro area banks, which does not seem plausible.
Kisgergely and Szombati (2014) list other reasons why the opt-in decisions of NMS could be
interrelated, such as: (1) the possible use of BU-membership, by international capital market
participants, as a signaling device of lower systemic risk, with the potential of putting opt-
borrowers at a competitive disadvantage; and (2) possible dominance of BU-centric views in EBA
and ESRB operations that could marginalize NMS positions on issues.
21. Available evidence suggests that host country’ risk profile, growth potential and
banking system soundness are dominant factors in foreign banks’ assets allocation decisions:
The latest CESEE bank lending survey (charts below) shows that foreign banks have broadly
similar views on the Czech and Slovak banking systems, which have similar financial soundness
indicators (Box 2 in the main report), despite the fact that one is a member of the euro area/BU
and the other is not.
During the crisis, euro area banks scaled back their exposure to the region, while differentiating
across countries based on their risk profiles. Among the NMS-6, banking systems with relatively
higher asset quality and profitability, such as Poland and the Czech Republic, managed to attract
additional foreign funding, while countries with higher FX-currency mismatches, NPLs and lower
profitability experienced significant outflows (see Box 2 in the main report).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 53
D. Conclusions
22. For the NMS-6, the calculus of opting into the BU before euro adoption is complex
and has many moving parts. The opt-in decision has to take into account a range of economic and
political considerations, as well as evolving European financial architecture. For those NMS-6 that
already set the date for euro adoption (Romania), an early BU opt-in would simply be a matter of
phasing in the necessary institutional adjustments. For those that have not yet decided on the
timing of euro adoption, the BU opt-in decision requires a careful consideration of the
characteristics, policy preferences/space, as well as greater clarity on the BU implementation.
When would opting into the BU make sense for the NMS-6?
BU design: the lack of equal (or fully equivalent) treatment of euro area and non-euro area
members of the BU (in the areas of representation in the SSM, access to common liquidity and
fiscal backstops) tilts the decision against early BU opt-in and in favor of waiting until euro
adoption.
BU modalities: the lack of clarity on and experience with the BU operational modalities may be
another factor in favor of waiting. This, in particular, applies to coordination between the SSM
and local supervisors, as well as to coordination between prudential policies at the national and
BU-levels and national monetary policies.
Some may still opt in because for them the BU participation may be a way to address specific
challenges, which outweigh all other considerations, including BU shortcomings. Notably, some
may see BU as a way to enhance quality and credibility of bank supervision or to gain access to
larger industry-funded common backstop.
23. Opting into the BU before euro adoption would be more attractive for the NMS-6 if
mechanisms were in place to ensure that the NMS-s concerns stemming from unequal treatment
of opt-ins and euro area members are fully addressed. Furthermore, greater clarity on the BU
operational modalities that would shape the opt- policy space and the support they can expect
Foreign Banks’ Assessment of Market Positioning and Potential across CESEE
Subsidiary current positioning in the market
Assessed market potential
Sources: EIB, CESEE Bank Lending Survey, H2 2014.
Sources: EIB, CESEE Bank Lending Survey, H2 2014.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
ALB
BIH
BGR
HRV
CZE
HUN
MKD
POL
ROU
SRB
SVK
SVN
UKR
Weak
Niche player
Satisfactory
Optimal
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
ALB
BIH
BGR
HRV
CZE
HUN
MKD
POL
ROU
SRB
SVK
SVN
UKR
Low
Medium
High
CEE NEW MEMBER STATES POLICY FORUM
54 INTERNATIONAL MONETARY FUND
from common euro area institutions is needed in order for the NMS-6 to make a more informed
decision on the BU opt-in.
Box 1. Key Elements of the Euro Area Banking Union
For member states participating in the BU, there are three key elements:
Single Supervisory Mechanism
1
Consisting of the ECB and national banking supervisors (the national competent
authorities or NCAs), the SSM unifies banking supervision across participating member states (currently, only euro
area states). The ECB is the overarching supervisory authority, directly supervising 120 significant banksjointly
comprising almost 85 percent of total euro area bank assets 3500
less significant banks in the euro area. The ECB can take over direct supervision of any less significant bank at any
time in order to maintain cross-country consistent and high supervisory standards, or if it deems the bank to have
become significant.
The ECB, in close cooperation with the NCAs, has carried out the assessment of significance based on the criteria
set out in the SSM Regulation and the SSM Framework Regulation, namely:
billion and 20% of GDP of a Member State);
c) significance of cross-border activities (in particular, if the ratio of its cross-border assets or liabilities to its total
assets or liabilities, respectively, is above 20%);
d) a request for, or the receipt of, direct public financial assistance from the European Stability Mechanism (ESM);
e) one of the three most significant credit institutions in a participating Member State.
Single Resolution Mechanism
2
The SRM refers to the system of national resolution authorities and the central
Single Resolution Board (SRB), which is a stand-alone institution. It unifies the bank resolution framework across
participating member states. The SRB oversees the resolution of banks by national resolution authorities (which will
follow the strictures of BRRD), and directly handles the resolution of large and cross-border banks. From January
2016, it can also draw upon a common, industry-funded backstop called the Single Resolution Fund (SRF), in order
to resolve banks under BRRD.
3
of covered deposits in the euro area).
ESM Direct Bank Recapitalization In December 2014, euro area member states gave the European Stability
Mechanism (ESM) the power to directly recapitalize banks (with up to available), mitigating some of the
potential fiscal problems associated with ESM indirect bank recapitalization, when a sovereign borrows from the
ESM and then funnels those funds into its banking system. ESM bank recapitalization will not be available for any
future non-euro area banking union participants, since the ESM Treaty is only open to currency union members.
However, even if it were available, there are doubts about its effectiveness as a common fiscal backstop as currently
formulated. The hurdles for its use are very high and in the event of systemic crisis, the ceiling on the funding
available for recapitalization could be rapidly reached.
___________________
1
The enabling legislation was adopted in October 2013 by the European Union. To prepare the ground for the SSM, the ECB has
undertaken a Comprehensive Assessment of bank balance sheets, with an asset quality review and stress tests.
2
At the EU level, the enabling legislation was adopted April 2014. The SRM began operating in January 2015, shortly after the
SSM.
3
The SRF was adopted by intergovernmental agreement (rather than EU legislation) in May 2014. It will start out with national
compartments which build up over time and are gradually mutualized (60 percent mutualized after 2 years, building to 100
percent after 8 years, in 2024).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 55
Box 2. The SSM Modalities
Oversight will be managed by a Supervisory Board (SB), based within the ECB, which consists of a chair and
vice-chair (the latter also serving on the ECB Executive Board), a single representative from each participating
Member States plus four ECB representatives and who are expected to act in their personal capacities for the
good of the U
national supervisor is not the national central bank, they may request that a representative of the national
central bank also attend. For the purposes of voting however, the representatives of any one member state
are considered as one member.
Governing Council (consisting of
ECB Board members and euro area national central bank heads). Regular draft decisions are passed by simple
majority, while regulatory decisions with SSM-wide import are passed by qualified majority (Article 26 of the
SSM Regulation).
1
The ECB Governing Council then either adopts the decision on a lapse-of-time basis or
objects to it. In case a decision is objected to, then it is referred back to the SB for redrafting, or, as an
intermediary step, goes to a mediation panel which works to resolve the differences in views across national
competent authorities.
Source: https://www.bankingsupervision.europa.eu/organisation/governance/html/index.en.html.
1
A qualified majority is defined in Article 16(4) of the Treaty on European Union (TEU) and Article 3 of Protocol Number 36 on
transitional provisions associated with TEU (reweighted according to the membership of the SSM).
CEE NEW MEMBER STATES POLICY FORUM
56 INTERNATIONAL MONETARY FUND
Box 3. The SRM Modalities
Decision-making in the SRM:
The governing body of the SRM is the Single Resolution Board (SRB), which consists of a chair, vice-chair, three
other full-time members, and one representative from the national resolution authorities of each participating
member state. The chair, vice-chair and other full-time members, constituting the executive of the SRB, are all
appointed by the European Parliament from a short-list of candidates drawn up by the Commission.
Resolution decisions are drafted by the executive of the SRB and are assumed adopted by the SRB unless
there is an objection by one of the representatives of the participating member states (similar to the non-
objection procedure used by the SSM). In the case of an objection, the SRB meets in plenary (all members)
and takes the resolution decision, based on a simple majority rule. In general, the plenary SRB meets at least
twice a year, to review the budget and assess resolution activity, but it may also meet at the behest of the
The resolution procedure also involves close coordination with the European Commission and the EU Council
(see below)
Source: europa.eu/rapid/press-release_MEMO-14-294_en.htm
Contributing to the SRF
Under the SRM Regulation and SRF intergovernmental agreement, all participating member states contribute
ex ante contributions to
the SRF are calculated pro rata with its share of total liabilities minus covered deposits of all banks in
participating member states (plus a risk-adjusted contribution drawing upon BRRD criteria; see the SRM
Regulation, Article 70).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 57
Box 4. Macroprudential Policy Space for the BU members
For BU members, the SSM entails some additional constraints on macroprudential policies. Under the
SSM Regulation (Article 5), national competent authorities (NCAs) can still deploy macroprudential measures
as they deem appropriate, following the usual practice of submitting them to the ESRB for a non-binding
opinion. However, in the case of CRR/CRD IV measures (see below), BU-participating states must also notify
the ECB of their intention 10 working days prior to issuance of their decision. If the ECB objects, then it
supplies a written explanation within 5 working days, which the national authority must take into
consideration. Furthermore, if the ECB wishes, it may apply stricter macroprudential requirements on banks,
irrespective of whether they are under direct SSM supervision or not, than the national authorities (subject
to similar notification and consideration timelines)
The CRR/CRD IV legislative package defines a range of tools over which national macro-prudential
authorities may set stricter requirements (above the industry-wide, micro-prudential minima) based on
systemic risk considerations, macro-prudential concerns, or to address risks at individual firm level. These are
subject to a notification requirement to the ESRB and include:
Pillar I measurescountercyclical capital buffer and additional capital buffers for systemic risk, systemic
important institutions, and capital conservation, as well as the leverage ratio and the level of own funds.
In addition, national authorities can set higher risk weights on real estate exposures and large exposures;
Pillar II measuresa wide range of measures at the level of individual institutions or group of
institutions with similar risk profile, imposed following a supervisory review and evaluation process
aimed at identifying risks they face or pose to the financial system;
Liquidity provisionsliquidity coverage ratio and net stable funding ratio;
Limits on large exposures and intra financial sector exposures.
National macroprudential authorities retain control over macroprudential measures, not specified in Union
law, such as the loan-to-value and debt-to-income ratios, among others (Chart below). This is subject to a
notification requirement to the ESRB and possible intervention by the EU Council. In addition, until the
harmonization of the liquidity requirements in 2015 and the leverage ratio in 2018, member states can set
unilaterally these measures.
Mapping Macroprudential Tools to Objectives
Source: Authors. Mapping to objectives is based on IMF 2013b.
CEE NEW MEMBER STATES POLICY FORUM
58 INTERNATIONAL MONETARY FUND
Box 5. Theoretical Considerations in Designing an Optimal Banking Union
The design of national supervision and safety nets in a multi-country integrated market has to take
into account potential cross-border spillovers. Tighter supervision which makes the domestic banking
system safer may be good for other countries with which this country has close links, by reducing financial
stability risks. But, tighter supervision may also make domestic banks less competitive vis-à-vis foreign
banks. This suggests that while there may be incentives for national supervisors in a financially integrated
region to cooperate, independent regulators may also have an incentive to promote the competitiveness of
When will a centralized solution (“banking union”) be preferred by national supervisors as a way to
achieve their national policy objectives? The theoretical literature suggests that countries that are highly
interlinked and similar in their regulatory preferences will tend to see higher net benefits to coordination,
compared to those that are not. But in order for such national supervisors to prefer a banking union, the
2006).
1
If however, the initial cross-country differences in supervisory preferences are significant, the
centralized solution may not be an optimal choice for all. In more extreme cases, regulatory preferences
may be distorted by vested interests of bank shareholders, debtors, and creditors (Scherf, 2014), in which
Parallels between the decisions to join a banking union and a currency union bring out additional
factors pertinent to the decision.
2
In reality, countries that are contemplating joining a banking union
may be very different, not only in terms of supervisory preferences, but also along other characteristics,
such as their degree of real and financial integration with banking union members, the degree of flexibility
of their economies, the structure of their banking systems, as well as the quality of prudential supervision
members reduces the probability of an idiosyncratic shock driving a wedge between national interests and
that of the banking union. But lower supervisory quality and lower backstops at the national level likely
increase the benefits of having common (tighter) regulatory/supervisory standards and common (larger)
backstops.
3
1
-level regulator is to relinquish its a
ary
condition for a centralized regulator to emerge endogenously as an agreement between the independent regulators is that the
p. 413).
2
See Box 1 in Chapter 1 on Macroeconomic Benefits and Challenges.
3
Recent research in the OCA area highlights the benefits of financial markets integration and of importing prudent economic
management by pegging the domestic currency to that of a dominant economic power (see Iossifov and others, 2009 for an
overview). In the same vein, a common fiscal backstop in a banking union serves the role of an insurance policy, upon which
individual members can draw in the event of an asymmetric shock.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 59
Box 6. Cross-Country Differences in Policymakers’ Relative Preference for Promoting
Domestic Banks
Countries differ in the quality of bank supervision and the balance between financial stability and support of
domestic banks. These two aspects of the policy regime can be, but are not necessarily, linked by the degree
policymakers' revealed preference for promotion of domestically-owned banks can be inferred from the
extent to which domestic banks are allowed to gain market share at the expense of above-average funding
costs.
The data suggest that domestic banks in Bulgaria and the Czech Republic, and to a lesser extent Poland and
Hungary, rapidly expanded their deposit base in the post-crisis period (see figure in the box). Domestic banks
in Bulgaria and Hungary have the largest market shares relative to other NMS-6 countries, whereas in the
Czech Republic the market share of domestic banks remains marginal. In addition, domestic banks have
higher funding costs than their foreign-owned peers in all NMS-6 countries, with the widest margins recorded
in Hungary, Bulgaria, and the Czech Republic.
Overall, results suggest that domestic banks in Bulgaria and Hungary might have benefitted from a more
favorable policy stance compared to their foreign-owned peers. Domestic banks in Bulgaria and Hungary have
the largest market shares relative to other NMS-6 countries, with gains in recent years associated with the
payment of a bigger premium in funding costs over foreign-owned banks. In the case of Hungary, the size of
the premium is likely driven by foreign-
government-mandated cost of restructuring of FX-denominated debt.
2007 2013
Change,
2007-2013
Domestic
banks
Foreign
banks
Difference in
growth rates
(dom. vs frn. banks)
Domestic
banks
Foreign
banks
Difference in rates
(dom. vs frn. banks)
Domestic
banks
Foreign
banks
Difference in ROE
(dom. vs frn. banks)
Bulgaria 21.4 32.1 10.7 133.4 27.4 106.0 4.7 2.9 1.8 5.9 5.7 0.2
Croatia … … … … … … … … … … … …
Czech Republic 1.7 6.4 4.6 167.1 25.1 142.0 2.4 1.4 0.9 9.3 14.1 -4.8
Hungary 41.0 47.0 6.0 14.3 -10.6 24.9 5.5 3.3 2.1 8.5 -6.6 15.0
Poland 32.1 40.2 8.1 79.9 35.8 44.1 3.0 2.6 0.5 11.4 9.2 2.2
Romania 13.5 11.1 -2.4 -7.9 29.6 -37.5 4.4 3.9 0.5 2.4 0.1 2.3
Source: ECB Consolidated Banking Data and Fund staff calculations.
Notes: The cost of funding is estimated as the ratio of interest expenses to the average interest-bearing liabilities (the latter proxied by the difference between total liabilities and equity).
New Member States: Relative Performance of Domestic- versus Foreign-Owned Banks
Cummulative Growth Rate of Deposits
2009-2013
Share of domestic banks in total
bank deposits
Average cost of funding
2009-2013
Return on equity
avg 2008-2013
CEE NEW MEMBER STATES POLICY FORUM
60 INTERNATIONAL MONETARY FUND
Appendix I. Largest Banks in NMS-6 and their Ultimate Owners
Table 1. Largest Banks in NMS-6 and their Ultimate Owners
Sources: BankScope; national sources; Haver Analytics; International Financial Statistics; World Economic Outlook database; and IMF staff calculations.
Country Bank
National
rank by
assets
Country of global
ultimate owner
Name of global ultimate owner
Market share
in banking
sector loans
Total assets
(bil EUR)
Total assets
(% of GDP)
Bulgaria UniCredit Bulbank AD 1 Italy UNICREDIT SPA 18% 7.18 18%
Bulgaria DSK Bank Plc 2 Hungary OTP BANK PLC 13% 4.70 12%
Bulgaria First Investment Bank AD 3 Bulgaria FIRST INVESTMENT BANK AD 11% 4.66 12%
Bulgaria Corporate Commercial Bank AD 4 Bulgaria MR TSVETAN RADOEV VASILEV 8% 3.58 9%
Bulgaria United Bulgarian Bank - UBB 5 Greece HELLENIC FINANCIAL STABILITY FUND 9% 3.56 9%
Bulgaria Raiffeisenbank (Bulgaria) EAD 6 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 8% 3.16 8%
Bulgaria Societe Generale Expressbank 7 France SOCIETE GENERALE SA 6% 2.10 5%
Bulgaria Central Cooperative Bank AD 8 Liechtenstein CHIM INVEST ANSTALT 3% 1.99 5%
Bulgaria CIBANK JSC 9 Belgium KBC GROEP NV/ KBC GROUPE SA-KBC GROUP 2% 1.10 3%
Bulgaria Allianz Bank Bulgaria AD-CB Allianz Bulgaria AD 10 Germany ALLIANZ SE 2% 1.05 3%
Croatia Zagrebacka Banka dd 1 Italy UNICREDIT SPA 31% 13.98 33%
Croatia
Privredna Banka Zagreb d.d-Privredna Banka
Zagreb Group
2 Italy INTESA SANPAOLO 18% 8.61 20%
Croatia Erste & Steiermärkische Bank dd 3 Austria ERSTE GROUP BANK AG 18% 7.89 18%
Croatia Raiffeisenbank Austria d.d., Zagreb 4 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 9% 4.91 11%
Croatia Hypo Alpe-Adria-Bank dd 5 Austria REPUBLIK OSTERREICH - GOVERNMENT OF AUSTRIA 8% 4.16 10%
Croatia Societe Generale - Splitska Banka dd 6 France SOCIÉTÉ GÉNÉRALE 7% 3.71 9%
Croatia Hrvatska Postanska Bank DD 7 Croatia GOVERNMENT OF CROATIA 4% 2.52 6%
Croatia OTP banka Hrvatska dd 8 Hungary OTP BANK PLC 3% 1.86 4%
Croatia Sberbank dd 9 Russia PRAVITELSTVO ROSSIISKOI FEDERATSII 2% 1.26 3%
Croatia Kreditna Banka Zagreb 10 Croatia N/A 1% 0.56 1%
Czech Republic Ceskoslovenska Obchodni Banka A.S.- CSOB 1 Belgium KBC GROEP NV/ KBC GROUPE SA-KBC GROUP 17% 39.16 26%
Czech Republic Ceska Sporitelna a.s. 2 Austria ERSTE GROUP BANK AG 16% 36.66 25%
Czech Republic Komercni Banka 3 France SOCIÉTÉ GÉNÉRALE 16% 32.70 22%
Czech Republic Unicredit Bank Czech Republic and Slovakia AS 4 Italy UNICREDIT SPA 10% 17.58 12%
Czech Republic Hypotecni banka a.s. 5 Belgium KBC GROEP NV/ KBC GROUPE SA-KBC GROUP 6% 8.10 5%
Czech Republic Raiffeisenbank akciova spolecnost 6 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 5% 7.46 5%
Czech Republic GE Money Bank as 7 USA
GE CAPITAL INTERNATIONAL HOLDINGS
4% 5.10 3%
Czech Republic J&T Banka as 8 Slovakia TECHNO PLUS, A. S. 2% 4.17 3%
Czech Republic PPF banka a.s. 9 Czech Republic MR KELLNER PETR 1% 3.98 3%
Czech Republic Stavební Sporitelna Ceské Sporitelny as 10 Austria ERSTE GROUP BANK AG 1% 3.76 3%
Hungary OTP Bank Plc 1 Hungary OTP BANK PLC 49% 36.24 36%
Hungary K&H Bank Zrt 2 Belgium KBC GROEP NV/ KBC GROUPE SA-KBC GROUP 8% 8.95 9%
Hungary Erste Bank Hungary Nyrt 3 Austria ERSTE GROUP BANK AG 11% 7.91 8%
Hungary MKB Bank Zrt 4 Germany
FREISTAAT BAYERN BAYERISCHES
STAATSMINISTERIUM DER FINANZEN
10% 6.85 7%
Hungary Raiffeisen Bank Zrt 5 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 10% 6.46 6%
Hungary CIB Bank Ltd-CIB Bank Zrt 6 Italy INTESA SANPAOLO 10% 6.46 6%
Hungary UniCredit Bank Hungary Zrt 7 Italy UNICREDIT SPA 7% 6.18 6%
Hungary OTP Mortgage Bank-OTP Jelzalogbank Rt 8 Hungary OTP BANK PLC 8% 4.49 5%
Hungary
Budapest Bank Nyrt-Budapest Hitel-és Fejleszési
Bank Nyrt
9 USA GE CAPITAL INTERNATIONAL FINANCING CORP INC 4% 3.16 3%
Hungary FHB Mortgage Bank Plc-FHB Jelzalogbank Nyrt. 10 Hungary
FHB MORTGAGE BANK PLC-FHB JELZALOGBANK
NYRT.
3% 2.57 3%
Poland
Powszechna Kasa Oszczednosci Bank Polski SA -
PKO BP SA
1 Poland
POWSZECHNA KASA OSZCZEDNOSCI BANK POLSKI
SA - PKO BP SA
15% 49.80 13%
Poland Bank Polska Kasa Opieki SA-Bank Pekao SA 2 Italy UNICREDIT SPA 10% 39.63 10%
Poland Bank Zachodni WBK S.A. 3 Spain BANCO SANTANDER SA 7% 26.52 7%
Poland mBank SA 4 Germany COMMERZBANK AG 7% 26.07 7%
Poland ING Bank Slaski S.A. - Capital Group 5 Netherlands STICHTING ING AANDELEN 5% 21.69 6%
Poland Getin Noble Bank SA 6 Poland GETIN NOBLE BANK SA 5% 15.90 4%
Poland Bank Millennium 7 Portugal
BANCO COMERCIAL PORTUGUÊS, SA-MILLENNIUM
4% 14.25 4%
Poland Raiffeisen Bank Polska SA 8 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 4% 13.35 3%
Poland Bank Handlowy w Warszawie S.A. 9 USA CITIGROUP INC 2% 11.35 3%
Poland Bank Gospodarki Zywnosciowej SA-Bank BGZ 10 Netherlands
COOPERATIEVE CENTRALE RAIFFEISEN-
BOERENLEENBANK B.A-RABOBANK NEDERLAND
3% 8.94 2%
Romania
Banca Comerciala Romana SA-Romanian
Commercial Bank SA
1 Austria ERSTE GROUP BANK AG 24% 15.43 11%
Romania BRD-Groupe Societe Generale SA 2 France SOCIÉTÉ GÉNÉRALE 18% 11.10 8%
Romania Transilvania Bank-Banca Transilvania SA 3 Romania TRANSILVANIA BANK-BANCA TRANSILVANIA SA 10% 7.46 5%
Romania UniCredit Tiriac Bank SA 4 Italy UNICREDIT SPA 9% 6.35 4%
Romania Raiffeisen Bank SA 5 Austria RAIFFEISEN LANDESBANKEN HOLDING GMBH 9% 6.23 4%
Romania CEC Bank SA 6 Romania STATE OF ROMANIA 7% 6.22 4%
Romania Alpha Bank Romania 7 Greece HELLENIC FINANCIAL STABILITY FUND 6% 3.76 3%
Romania Volksbank Romania 8 Austria VOLKSBANKEN HOLDING REGGENMBH 6% 3.20 2%
Romania Bancpost SA 9 Greece HELLENIC FINANCIAL STABILITY FUND 4% 2.75 2%
Romania Banca Romaneasca S.A. 10 Greece HELLENIC FINANCIAL STABILITY FUND 3% 1.72 1%
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 61
References
Bank of International Settlements, 2010, Report and Recommendations of the Cross-border Bank
Resolution Group, available on the internet at www.bis.org/publ/bcbs162.pdf, (Basel).
Directive 2014/59/EU of the European Parliament and of the Council of 15 May
2014 Establishing a Framework for the Recovery and Resolution of Credit Institutions and
Investment Firms and Amending Council Directive 82/891/EEC, and Directives 2001/24/EC,
2002/47/EC, 2004/25/EC, 2005/56/EC, 2007/36/EC, 2011/35/EU, 2012/30/EU and
2013/36/EU, and Regulations (EU) No 1093/2010 and (EU) No 648/2012, of the European
Parliament and of the Council Text with EEA Relevance, available on the internet at
http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:32014L0059 .
Without Borders?
Implications of Ring-Fencing for European Cross-
(International Monetary Fund).
CRR Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26
June 2013 on Prudential Requirements for Credit Institutions and Investment Firms and
Amending http://eur-
lex.europa.eu/legal-content/EN/TXT/?uri=uriserv:OJ.L_.2013.176.01.0001.01.ENG .
CRD IV Directive 2013/36/EU of the European Parliament and of the Council of 26 June
2013 on Access to the Activity of Credit Institutions and the Prudential Supervision of Credit
Institutions and Investment Firms, Amending Directive 2002/87/EC and Repealing Directives
http://eur-lex.europa.eu/legal-
content/EN/TXT/?uri=uriserv:OJ.L_.2013.176.01.0338.01.ENG .
Journal of Financial Economics, 79(2), pp. 401430.
DGSD Directive 2014/49/EU of the European Parliament and of the Council of 16 April
2014 on Deposit Guarantee Schemes Text with EEA Relevance
http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32014L0049 .
ESRB Regulation: Regulation (EU) No 1092/2010 of the European Parliament and of the Council of
24 November 2010 on European Union Macro-Prudential Oversight of the Financial System
and Establishing a European Systemic Risk Boardhttp://eur-
lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32010R1092 .
European Commission,
CEE NEW MEMBER STATES POLICY FORUM
62 INTERNATIONAL MONETARY FUND
the internet at http://eur-lex.europa.eu/legal-
content/EN/TXT/PDF/?uri=CELEX:52012DC0510&from=EN, (Brussels).
European Commission
Fina
the internet at http://europa.eu/rapid/press-release_MEMO-13-679_en.pdf, (Brussels).
European Central Bank
Markets,
internet at www.ecb.europa.eu/press/key/date/2012/html/sp121008_1.en.html, October,
(Frankfurt).
-border Coordination of Prudential Supervision and
Journal of Financial Stability, 7(3), pp. 155164.
Rapid Credit Growth and the Role of Supervisory and Prudential Policie
Paper No. 05/151, available via internet at:
www.imf.org/external/pubs/ft/wp/2005/wp05151.pdf (Washington: International Monetary
Fund).
International Monetary Fund, 201IMF Staff Discussion Note,
13/01, available via internet at: www.imf.org/external/pubs/ft/sdn/2013/sdn1301.pdf,
(Washington).
International Monetary Fund
available via internet at: www.imf.org/external/np/pp/eng/2013/061013b.pdf, (Washington).
International Monetary Fund, 2013cGlobal Financial Stability
Report, October , available via internet at:
www.imf.org/External/Pubs/FT/GFSR/2013/02/pdf/text.pdf, (Washington).
No, 09/260, available on the
internet at www.imf.org/external/pubs/ft/wp/2009/wp09260.pdf, (Washington: International
Monetary Fund).
Kisgergely,
Assessment of a Possible Close Cooperation, MNB Occasional Papers 115, available on the
internet at
http://english.mnb.hu/Root/Dokumentumtar/ENMNB/Kiadvanyok/mnben_muhelytanulmany
ok/Banking_Union_MNB_OP_115.pdf (Budapest: Magyar Nemzeti Bank).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 63
Journal of Common Market Studies, Vol. 42(4), pp. 689-715.
Ozkan, Gulcin athe use of Monetary and Macroprudential Policies for Small
Open Economies,” IMF Working Paper No. 14/112, available on the internet at
www.imf.org/external/pubs/ft/wp/2014/wp14112.pdf, (Washington: International Monetary
Fund).
Scherf, Gundbert, 2014, Financial Stability Policy in the Euro Zone—The Political Economy of National
Banking Regulation in an Integrating Monetary Union, (Wiesbaden: Springer Gabler).
The World Economy, vol.
16(6), pp. 663-685.
CEE NEW MEMBER STATES POLICY FORUM
64 INTERNATIONAL MONETARY FUND
THE EU FISCAL FRAMEWORK AND PENSION REFORM
1
1
Prepared by Greetje Everaert, Nan Geng, John Ralyea, Yan Sun, and Johannes Wiegand. Min Song and Jessie Yang
provided excellent research assistance. The authors are grateful to Ludwig Kotecki (Ministry of Finance of Poland),
Dusan Hadril (Ministry of Finance of the Czech Republic), and Per Eckefeldt (European Commission) for discussing the
paper at the New Member States Policy Forum in Warsaw on December 12, 2014, and to the NMS-6 country
representatives that reviewed the paper during bilateral visits in November 2014. Anita Schwarz (World Bank), Csaba
Feher, and Mauricio Soto (both IMF, Fiscal Affairs Department) provided support and guidance throughout this
project.
Summary
The NMS were among the earliest pension reformers in Europe. Pressures on their public pensions
emerged in the 1990s, reflecting inter alia falling labor force participation and high unemployment
during transition that rendered the inherited pension systems insolvent.
As a result, public pension systems in many NMS are technically sustainable, but often at the cost
of low replacement rates. In spite of severe demographic pressures, average public pension spending is
projected to remain at a level of about 10 percent of GDP in 2060, broadly unchanged from today. The
replacement rate, however, is often projected to fall sharply, giving rise to doubts about whether the
public, defined-benefit, pay-as-you-go (PAYG) pension systems are socially sustainable.
Along with parametric reforms to PAYG systems, most NMS introduced private, mandatory, pre-
funded pension (“Pillar II”) schemes with individual accounts, in order to supplement retirement
incomes. In Hungary and Poland, as well as Estonia and the Slovak Republic, second pillar contributions
exceeded 1 percent of GDP per year in the mid-2000s.
The performance of Pillar II schemes has been mixed. In particular, there is no systematic evidence
that the introduction of Pillar II schemes increased in national savingswhich, ultimately, is required to
generate higher pension incomeas contributions were diverted from PAYG systems and the resulting
fiscal impact was accommodated with higher deficits. Returns were generally modest, while exceeding
basic benchmarks. Management fees remain elevated, even though they have come down somewhat.
In the wake of the 2008/09 financial crisis, many countries unwound their second pillars and
redirected contributions to the budget, as governments struggled with severe fiscal pressures. The
under the stability and growth pact as arguably been a factor in the
reversal: deficit ceilings were defined mostly in headline terms, granting only partial allowances for Pillar
II transition costs. This provided an incentiveespecially for countries with large second pillarsto
reverse the reform. By contrast, neither returns on Pillar II assets nor management fees appear to have
been a systematic factor in the reversal.
While recent reforms render the EC’s fiscal framework more flexible toward Pillar II systems, it
continues to fall short of neutr
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 65
A. Public Pension Systems in New Member States: The Broad Picture
1. The EU New Member States were among the earliest pension system reformers in
Europe. Pressures on the their pay-as-you-go (PAYG) pension systems arose early during the
economic transition of the 1990s, as labor participation declined sharply, unemployment rose, the
informal economy grew, and high evasion complicated the collection of social security contributions.
Reforms to reduce these pressures included changing pension formulas to control the increase in
benefitsfor example by linking pension increases to inflation rather than wage growthincreasing
retirement ages, and reducing incentives for early retirement (see Appendix I).
2. However, demographic change continues to put pressure on PAYG systems. Fertility
rates in the NMS-6 have fallen by an average of 30 percent during 1990-2010. Life expectancy has
also increased since 1990, and is projected to continue trending upward (Figure 1). Moreover, some
countries, including Bulgaria, Romania, and Poland, experienced substantial outmigration in the past
decade, mostly of the younger generation. Overall by 2060 the working age population is projected
to fall to 6080 percent of its 2010 level. Combined with longer life expectancy, this will result in a
sharp increase in the dependency ratio: in 2060, an old person (65+) is projected to be supported by
1½ workers, compared to four workers in 2010.
Figure 1. Selected European Countries: Key Demographic Data
1
1.2
1.4
1.6
1.8
2
HRV
EST
BGR
LTU
SVN
CZE
SVK
POL
ROU
HUN
LVA
CESEE avg.
2010
2060
Fertility rate
(Percent)
Sources: Eurostat; United Nation; and IMF staff calculations.
0
5
10
15
20
25
HRV
SVN
CZE
POL
HUN
EST
ROU
SVK
LVA
BGR
LTU
CESEE avg.
2010
2060
Life expectancy at 65, Men
(Years)
0
5
10
15
20
25
30
HRV
SVN
EST
POL
HUN
CZE
LVA
SVK
LTU
ROU
BGR
CESEE avg.
2010
2060
Life expectancy at 65, Women
(Years)
0
10
20
30
40
50
60
70
80
LVA
POL
ROU
SVK
BGR
SVN
HUN
LTU
EST
CZE
HRV
CESEE avg.
2010
2060
Old-Age Dependency Ratio
(Percent)
CEE NEW MEMBER STATES POLICY FORUM
66 INTERNATIONAL MONETARY FUND
3. Even with the sharp increase in the dependency ratio, most NMS-6 pension systems
appear technically sustainable if already reforms are implemented as foreseen.
2
The reforms
are projected to keep average pension spending at about 11 percent of GDP, only a percentage
point more than todaywhich compares to public pension spending of almost 23 percent of GDP if
reforms are not implemented (Table 1). Most notably, legislated reforms in Poland are projected to
yield an annual reduction of 25 percent of GDP in pension spending (European Commission, 2012).
Still, in some other countries, pension expenditures are projected to increase, yielding rising pension
deficits in their PAYG systems (Table 2).
Table 1. Pension Spending Projections
1/
4. At the same time, most reforms imply a steep cut in pension benefits, calling into
question the social—and therefore political—sustainability of PAYG systems. For some NMS-6
PAYG systems, the replacement rate of income in retirement is projected to fall below 20 percent,
compared to around 40 percent today.
3
In fact, some countries have already reversed recent
reforms, in response to political pressures that increased when benefit reductions started taking
effect (Schwarz and Arias, 2014).
2
Most data for this section are drawn . Croatian data are
from the World Bank (Croatia was not yet an EU member in 2012 and is thus not covered by the 2012 Ageing
Report). The base year is 2010. The next Ageing Report is expected to be published in May 2015.
3
Different tax treatment on pension and non-pension income will affect income replacement rate net of taxes. For
some countries (e.g. Hungary), the replacement rate can be higher because pensions are not taxed.
2010 2020 2030 2040 2050 2060 2010 2020 2030 2040 2050 2060
Bulgaria 9.9 9.2 9.6 10.1 11.1 11.1 9.9 11.9 14.2 16.6 20.0 21.2
Croatia 2/ 6.0 4.4 3.3 2.6 2.5 3.3 6.0 5.5 5.8 5.6 6.3 7.0
Czech Republic 9.1 8.7 8.9 9.7 11.0 11.8 9.1 11.9 13.7 15.9 19.2 21.0
Estonia 8.9 7.7 8.2 8.1 8.0 7.7 8.9 9.9 11.3 12.4 14.8 16.8
Latvia 9.7 7.3 6.5 6.3 6.4 5.9 9.7 10.4 11.9 13.8 17.4 21.0
Lithuania 8.6 7.6 8.4 9.6 10.8 12.1 8.6 9.0 11.2 12.9 14.4 17.2
Hungary 11.9 10.5 9.3 9.8 11.2 12.4 11.9 13.6 14.3 17.2 21.3 24.4
Poland 11.8 10.9 10.9 10.3 10.0 9.6 11.8 15.4 19.9 22.8 29.7 35.5
Romania 9.8 9.2 10.3 11.6 12.8 13.5 9.8 11.6 13.9 18.8 24.6 28.8
Slovakia 8.0 8.6 9.5 10.6 12.2 13.2 8.0 10.3 13.1 16.1 21.3 25.1
2/ For Croatia, numbers are for old age related pension spending only and do not include other pension payments
from the PAYG scheme.
1. "With reforms" block refer to projected spending under currently legislated reforms, and correspond broadly to the
baseline projection in the EU Ageing Report. "Without reforms" block refer to pension spending that would occur if
pension parameters were kept constant at 2010 levels without considering changes from legislated reforms.
With reforms
Without reforms
Source: Fund staff calculations based on data in the EU Aging Report (2012), 2012 EU Fiscal Sustainability Report,
and information provided by the World Bank on Croatia.
(in percent of GDP)
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 67
5. As a result, further steps are called for to improve both fiscal and social sustainability
of NMS pension systems. Two areas with significant reform potential are aligning the mandatory
retirement age with longevity, and reducing incentives for early retirement. Several NMS-6 countries
continue to have relatively generous early retirement incentives and special pension regimes for
privileged groups. As a result, the share of relatively young beneficiaries remains high. The gains
from embracing these reforms could be used to further improve fiscal sustainability, increase the
social sustainability of replacement rates, or a combination of the two.
Table 2. Net Present Value of Pension Deficits
EE Pension Cost, Replacement Rate, and Young Pensioners
2010-
2030
2031-
2060
2010-
2060
2010-
2030
2031-
2060
2010-
2060
2010-
2030
2031-
2060
2010-
2060
Bulgaria
34
79
114
31
55
86
29
38
67
Croatia 1/
-38
-105
-144
-34
-74
-108
-30
-53
-83
Czech Rep.
7
65
72
6
44
50
6
30
36
Estonia
16
-13
4
15
-9
7
14
-6
8
Latvia
51
66
116
46
47
93
42
34
77
Lithuania
14
61
75
13
41
55
12
29
41
Hungary
39
87
126
36
59
95
33
41
74
Poland
76
48
124
70
35
105
65
25
90
Romania
27
10
37
25
7
33
24
5
29
Slovakia
77
176
253
69
122
191
63
86
149
Average
30
47
78
28
33
61
26
23
49
Source: EC (2012), World Bank, and Fund staff calculation.
Table 2. Net Present Value of Pension Deficits
Discount rate=1
Discount rate=2
Discount rate=0
(in percent of GDP)
1/ For Croatia, only old age related pension spending is included. PAYG balance including
other pension payments will be smaller.
Source: EU 2012 Ageing Report, World Bank.
1/ Beneficiaries include old age, disability, survivor. For some countries only old age and disability beneficieries included. In
some countries orphans are included.
BGR
CZE
EST
HUN
LVA
LTU
POL
ROU
SVK
HRV
y = 1.6695x + 11.916
R² = 0.2793
10
15
20
25
30
35
40
45
50
0 5 10 15
Wage replacemenr rate
PAYG Pension Spending (percent of GDP)
PYAG Pension Cost and Replacement Rate, 2060
(For selected EE countries)
0
10
20
30
40
50
60
ROU
RUS
BLR
SVK
HRV
BIH
POL
ALB
BGR
LVA
LTU
Pension Beneficiaries Below the Age of 65 1/
(Percent, 2013)
CEE NEW MEMBER STATES POLICY FORUM
68 INTERNATIONAL MONETARY FUND
B. Pillar II Pension Schemes: History, Rationale, Performance
6. A second component of NMS pension reforms—accompanying the parametric changes
to the PAYG systems mentioned above—has often been the introduction of private,
mandatory, pre-funded, “pillar II” pensions. Pillar II-type systems require contributions to be
channeled into privately held accounts, with accumulated savings to be paid out upon retirement.
Chile was the first country to introduce such a system in the early 1980s. Its example inspired many
countries in Eastern Europe to follow. Hungary (1998) and Poland (1999) led reform efforts in the
region, followed by Bulgaria, Croatia (both 2002), and Romania (2007) (Appendix II). The exception
among the NMS-6 is the Czech Republic that, instead of making contributions mandatory,
strengthened tax incentives for accumulating savings on a voluntary basis.
Conceptual Issues: Why a Second Pension Pillar?
7. A key objective of second pension pillars is to generate additional income out of which
supplementary pensions can be paid. In the NMS, second pillars were often designed to
compensate for the pension income losses from parametric PAYG reforms. At the aggregate level,
this objective can be achieved only if national savings increase. Further, for higher savings to
materialize, the increase in private savings generated by the establishment of a second pillar must
not be offset by simultaneous public sector dis-saving. To avoid this, the contributions to a Pillar II
plan should not be accompanied by decreases in contributions to the PAYG scheme. Alternatively,
the resulting budgetary shortfall needs to be compensated with fiscal savings elsewhere.
8. By the time of their introduction, several other benefits associated with second pillar
programs were touted.
Risk diversification. With a second pillar, pension benefits are paid not only from the wage
base (PAYG) but also from returns on capital (Pillar II). Further, second pillars can allow for better
diversified portfolios by enabling investments in foreign assets, especially when domestic capital
markets are thin.
Ownership, labor market participation, and capital market development. Second pillar
schemes can link pension income transparently to contributions, thus enhancing ownership and
awareness on the part of contributors of the need to save for retirement income. They may also
provide incentives for higher labor market participationor for shifting from the informal to the
formal labor marketand for capital market development, by developing longer-term financial
instruments.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 69
Pillar II Pension Systems in the NMS: How Have They Performed?
9. This section reviews some performance indicators for Pillar II systems in the NMS. As a
general caveat, such assessments are not straightforward. For once, it can be difficult to identify
clear-cut metrics of success: while the capacity to pay higher pensions will ultimately require higher
national savings, for example, the national savings rate is affected by many factors other than
second pillars. This renders it problematic to link changes in savings (or the lack thereof) directly to
the success of reform efforts. Other performance indicators are not fully comparable across
countries and time periods: lack of asset diversification or elevated administrative fees, for
examplefeatures noted beloware fairly common among young pension funds, such as those of
the NMS. However, these indicators tend to improve as funds mature.
10. With these caveats, there is no systematic evidence that Pillar II systems have increased
savings (Figure 2).
National savings as a share of GDP do not reveal a clear pattern around the dates of Pillar II
introductiononly Croatia, Estonia, Lithuania, Poland and Slovakia show a sizeable increase. This
observation is consistent with governments offsetting the increase in private savings triggered
by the establishment of Pillar II funds with higher fiscal deficits.
4
Social security contributions. As for more direct evidence of public dis-saving, upon Pillar II
introduction PAYG contributions were typically cut to the same degree as Pillar II contributions
were levied, to avoid excessive burdens on wage earnersespecially as at the time of Pillar II
introduction many economies were still weak from transitionand prevent higher tax wedges.
5
Investments in government bonds. Further, large holdings of government bonds by many
Pillar II funds suggest that governments covered transition cost largely with debt financing
rather than fiscal consolidationtapping the very Pillar II funds for financing that they had
created. Reflecting this pattern, Pillar II investments in government securities are more than
twice as high in the NMS than for private pension funds in other OECD countries.
6
11. Prior to the 2008/09 financial crisis, gross investment returns on Pillar II funds were
generally modest, even though, barring a few exceptions, they exceeded real returns from
investing in domestic long-term government bonds. Once the crisis struck, pre-crisis gains were
eliminated. As Pillar II funds have a long-term investment horizon, comparing their returns against
4
Pillar II reforms were sometimes hoped to be self-financing, through increased labor force participation and, as a
consequence, higher tax revenues. In the case of Poland, labor force participation rates were expected to increase by
some 20 percentage points with Pillar II pension introduction (Epstein and Velculescu, 2011). In the event,
participation remained broadly constant in the 2000s.
5
Exceptions are Estonia and Lithuania, where the government matched additional individual contributions with
additional contributions from the state (from the outset in Estonia, only recently in Lithuania).
6
Note that the OECD comparator group includes not only second pillar funds but also voluntary and occupational
retirement funds.
CEE NEW MEMBER STATES POLICY FORUM
70 INTERNATIONAL MONETARY FUND
short-term benchmarks is problematic, however, especially if funds invest into riskier assets such as
equity. As a result, it is too early to assess to what extent the crisis losses can be recouped.
Figure 2. Second Pension Pillars and National Savings
83
70
68
58
56
56
43
39
26
12
6
26
18
35
42
37
57
58
5
24
3
7
8
3
20
4
16
17
ROU
SVK
HRV
BGR
POL
HUN
LVA
LTU
EST
Bonds
Equity
Bank Deposit
Other
Second Pillar Pension Fund Asset Allocation (2012 or latest)
BGR
HRV
HUN
POL
ROU
EST
LVA
LTU
SVK
y = -1.9954x + 28.693
R² = 0.4361
0
5
10
15
20
25
30
0 2 4 6 8 10
Contribution rate of PAYG in 2007 (percent of gross wage)
Contribution rate of Pillar II in 2007 (percent of gross
wage)
Contribution Rate of PAYG and Pillar II
0
5
10
15
20
25
30
35
40
1990
1994
1998
2002
2006
2010
2014
BGR
HRV
CZE
HUN
POL
ROU
EST
LVA
LTU
SVK
Gross National Savings
(Percent of GDP)
0
5
10
15
20
25
BGR ('02)
HRV('02)
HUN ('98)
POL ('99)
ROU ('08)
EST ('02)
LVA ('01)
LTU ('04)
SVK ('05)
5y avg before reform
5y avg after reform
National Savings per GDP
(Percent)
Sources: WEO; Haver; national authorities; EC 2012 Ageing Report; WB (2014) and IMF staff calculations.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 71
12. Asset management fees charged by Pillar II funds have come down, but they remain
high. Total fees for Pillar II funds have fallen from an average of more than 2 percent of assets in
2006 to about 1¼ percent in 2011. In part, this may reflect the realization of economies of scale as
Pillar II funds grow larger, although it may also relate to political pressures that funds faced in the
wake of the financial crisis. Still, operating expensesone element featuring in the fee structure
remain more than 50 percent higher than in other OECD countries (Figure 3).
7
13. Only a few countries with second pillars have taken advantage of the opportunity to
diversify away from domestic risk. As shown above, Pillar II portfolios contain a high share of
government securities. The Baltic countriesthat have low levels of public debtare the main
exception. In many countries, Pillar II regulations enforce a high share of risk-free assets and a
minimum domestic investment requirement, creating a bias toward government securities, bank
deposits, and cash.
C. The EU Fiscal Framework and Pillar II Reversals
The EU Fiscal Framework and Pillar II Pensions Prior to the Crisis
14. The treatment of second pillars within the EU fiscal framework assumed practical
relevance only in 2004, when the first NMS joined the EU. Among the older EU members, only
Sweden had a second pillar. Further, prior to 2004, second pension pillars were recorded as part of
the public sector when calculating government deficits and debt. In 2004, however, Eurostat
reclassified Pillar II to within the private sector. This gave rise to higher recorded fiscal deficits, as
Pillar II contributions counted no longer as fiscal revenue a phenomenon
15. Starting with a 2005 reform of the Stability and Growth Pact (SGP), the EU fiscal
framework began to partially accommodate Pillar II transition costs (Table 3).
A formal request to fully exempt transition cost was turned down in 2005. The request had been
brought forward by countries that had already put in place a second pillar at the time, which
included Hungary, Poland, and Sweden.
However, the European Council agreed on limited exemptions in the context of the Excessive
Deficit Procedure (EDP). Specifically, it allowed adjustments to the deficit for a maximum of five
yearson a degressive linear scale
7
As a rule of thumb, one percent of assets spent in fees and other charges reduces life-time pension earnings by
about 20 percent, see Barr (2000).
CEE NEW MEMBER STATES POLICY FORUM
72 INTERNATIONAL MONETARY FUND
Figure 3. Pillar II Pension Funds: Returns and Fees
Sources: WEO, OECD and IMF staff calculations.
1/ 2005-07 for Slovak Rep.
2/Excl. Czech Republic.
3/ CEE OECD include Czech Republic, Estonia, Hungary, Poland and Slovak Republic.
-40
-30
-20
-10
0
10
20
2000
2002
2004
2006
2008
2010
2012
Min
Max
Avg
Median
Annual Gross Real Returns on Pillar II Pension
Investments in NMS+4 2/ (Percent)
-8
-6
-4
-2
0
2
4
6
8
10
POL
HUN
BGR
HRV
LVA
LTU
SVK
EST
Weighted Average Returns, 2004-07 1/
(Percent, real)
0.0
0.2
0.4
0.6
0.8
1.0
Average CEE OECD
Average non-CEE OECD
Operating Expenses, % of total assets
2006
2011
Operating Expenses of Pillar II in CEE OECD 3/
BGR
HRV
POL
ROU
EST
LVA
LTU
SVK
y = -0.0647x + 1.8144
R² = 0.3956
0.0
0.5
1.0
1.5
2.0
2.5
0 5 10 15 20
Fund management fees (pct. of total assets, 2011)
Mandatory pension fund savings (pct. of GDP, 2012)
Pension Fund Management Fees and Mandatory
Pension Fund Savings
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 73
Table 3. Treatment of Net Cost of Systemic Pension Reforms in the EU Fiscal Framework
(the Stability and Growth Pact (SGP))
As for the preventive arm of the SGP, net cost of systemic pension reforms with an impact on
medium-term objective (MTO).
16. These exemptions were insufficient to eliminate disincentives for maintaining a second
pillar, especially as regards the EDP. The ceiling for the deficit allowance remained unspecified,
but a common understanding was that it would be at most ½ of a percent of GDP. This compares to
average Pillar II transition costs of about one percent of GDP pre-crisis, with significantly higher
fiscal burdens for countries with large second pillars, such as Poland and Hungary. Further, the
exemption period was far shorter than the actual transition period during which a second pillar
2005 reform of the Pact 2011 reform of the Pact
Can be partially reflected through the
MTO
ILD
indicator, but subject to peer
review and endorsement by the
EPC.*
Can be partially reflected through the
MTO
ILD
indicator, but subject to peer
review and endorsement by the EPC.*
N/A
Can be taken into account when defining
the path or allowing a temporary deviation
from the path with two conditions: 1) a
safety margin to ensure the respect of the
3% of GDP reference value for the deficit
is guaranteed; and 2) the budgetary
position is expected to return to the MTO
within the period covered by the Stability
or Convergence Program.
Launch of EDP Government debt
No requirement
Does not exceed the Maastricht reference
value
Government deficit
1) Close to the Maastricht reference
value; and 2)excess reflects the net
cost of the reform.
1) Does not significantly exceed what can
be considered close to the Maastrict
reference value; and 2) Excess is explained
by reform costs.
Other criteria
Considered only over five years and in
resgressive scale
Overall fiscal sustainability is maintained
Aboragation of
EDP
Government debt
No requirement
Does not exceed the Maastrict reference
value
Government deficit
1) Has declined substantially and
continuoursly; and 2) Close to the
Maastrict reference value.
1) Has declined substantially and
continuoursly; and 2) Close to the
Maastrict reference value.
Other criteria
Considered only over five years and in
resgressive scale
Overall fiscal sustainability is maintained
N/A N/A
Table 3. Treatment of net cost of systemic pension reforms in the EU fiscal framwork (the Stability and Growth Pact (SGP))
MTO revision
Adjustment path toward the MTO
Assessment of compliance with the debt ceilings
* MTO is defined as the maximum among three components, MTO
MB
(the "minimum benchmark" as agreed by the EFC), MTO
Euro/ERM2
(the Pact obligation for euro
area Member States and Member States participating in ERM II to have an MTO not lower than –1% of GDP), and MTO
ILD
. The component MTO
ILD
, which has an
ageing component in it, relates to explicit liabilities and a fraction of implicit liabilities. (Detailed description can be found in Chapter 3 of Part II of2013 Public
Finance Report available at http://ec.europa.eu/economy_finance/publications/european_economy/2013/pdf/ee-2013-4.pdf
Preventive arm of the pact
Corrective arm of the pact
Criteria/conditions for consideration
CEE NEW MEMBER STATES POLICY FORUM
74 INTERNATIONAL MONETARY FUND
creates net budgetary costthis period can last 4050 yearsand there was no allowance under
the debt criterion.
Pillar II Pensions Reversals: Why Did They Occur?
17. Several second pillar reforms were reversed following the 2008/09 crisis, although not
in all countries the reversal has been permanent (Figure 4). As regards temporary reversals, the
Baltic countries are currently in the process of or have finished restoring their Pillar II systems.
Romania delayed somewhat the built-up of its second pillar. By contrast, Slovakia and Poland
significantly reduced the size of their Pillar II schemes, with no declared intention for restoration,
and Hungary eliminated its second pension pillar altogether. Disappointing financial performance
and high private management fees were often cited as reasons for the reform reversals. The
exception among the NMS-6 are Bulgaria and Croatia, both of which have maintained their second
pillars throughout the crisis period, and without a major change in parameters.
8
18. To gauge why countries unwound second pillar reforms, we correlate the size of the
Pillar II reversal with characteristics of a country’s Pillar II fund. The results have to be
interpreted with some caution, given the small number of observations.
9
19. With this caveat, an important trigger for reversing Pillar II reforms appear to have
been fiscal pressures, including the need to stay within—or return to—the EDP’s deficit
ceilings.
Countries with the largest second pillars—as measured by the Pillar II contribution rate—
had the largest reform reversals. For them, the amount of contributions channeled into the
second pillar was large, providing a strong incentive to divert second pillar contributions to the
budget in order to reduce the headline fiscal deficit.
By contrast, countries with smaller Pillar II schemes were generally able to maintain them,
as these schemes imposed less of a fiscal burden. Pillar II funds with contribution rates of up to
5 percenttriggering a loss in fiscal revenue of ½-1 percent of GDPdid in general survive the
financial crisis and its aftermath.
As for the preventive arm of the SGP, net cost of systemic pension reforms with an impact on
long term fiscal sustainability could be partially taken into
medium-term objective (MTO).
8
Including Bulgaria which recently allowed participants in the second pillar to opt back into the PAYG system.
9
Further, the small number of observations allows only for univariate analysis.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 75
Figure 4. Pillar II Pension Reversals
BGR
HRV
HUN
POL
EST
LVA
LTU
SVK
y = -0.5676x - 0.5992
R² = 0.1589
-8
-6
-4
-2
0
2
4
6
8
10
-10 -8 -6 -4 -2 0 2
Weighted Avg Real Return (2005-07)
Change of Pillar II Contribution Rate over 2007-13
Reform Reversals and Pillar II Asset Returns
BGR
HRV
HUN
POL
EST
LVA
LTU
SVK
y = -0.0583x + 2.0141
R² = 0.0402
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
-10 -8 -6 -4 -2 0 2
Administration Fees (pct. of total assets, 2006)
Change of Pillar II Contribution Rate over 2007-13
Reform Reversals and Pillar II Fees
Sources: WB (2014) and staff calculations.
BGR
HRV
HUN
POL
ROU
EST
LVA
LTU
SVK
y = -0.5389x + 4.2886
R² = 0.7304
0
1
2
3
4
5
6
7
8
9
10
-10 -8 -6 -4 -2 0 2 4
Contribution Rate of Pillar II in 2007
Change of Pillar II Contribution Rate over 2007-13
Reform Reversals and Pillar II Contribution
Rates (in percent of gross wages)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
BGR ('02)
HRV ('02)
CZE ('13)
HUN ('98)
POL ('99)
ROU ('08)
EST ('02)
LVA ('01)
LTU ('04)
SVK ('05)
Average before crisis
Average after crisis
Budgetary Cost of Second Pension Pillars
(Percent of GDP)
CEE NEW MEMBER STATES POLICY FORUM
76 INTERNATIONAL MONETARY FUND
This said, the on the reversal is difficult to disentangle from
market pressures during crisisCuevas et al (2008) find that investors put a higher weight on
explicit government debt than on implicit pension liabilities in their assessment of country
creditworthiness.
20. There is no evidence that reform reversals were related to poor investment
performance. There is even a negative correlation of pre-crisis returns with reform reversals
hence, if anything the Pillar II funds that were dissolved or cut back were relatively more profitable.
Similarly, there is no significant relationship between the level of asset management fees and the
extent of Pillar II reversals after 2008 for the countries under consideration.
The EU’s Fiscal Framework Post-Crisis
21. In 2011, another SGP reform expanded the framework’s flexibility vis-à-vis second
pillar funds.
Deficit exemption. The regressive scale for the deficit allowance under the EDP was
eliminatedin other words, it was made permanentbut the allowance remains subject to tight
approval restrictions. Further, the size of the exemption remains limited.
MTOs. In addition, under the preventive arm of SGP, net cost of systemic pension reforms can
now also be partially taken account in part when definingor allowing a temporary deviation
fromadjustment path toward its MTO (and not just the MTO itself).
Further, in September 2014 the fiscal accounting framework ESA 2010 entered into force. Under ESA
2010, lump-sum transfers of assets from the second pension pillar fund to the general government
sector have no longer a direct impact on the general government budget balance, somewhat
reducing the incentive to abolish Pillar II funds in the context of the EDP.
10
22. While the reforms have increased the flexibility of the EU’s fiscal framework, they
remain short of a neutral treatment of different types of pension regimes. Shortcomings exist
Preventive arm. Improvements in fiscal sustainability from pension reforms are only partially
recognized in MTOs. In practice, there appears to be little relation between the existence of a
Corrective arm. However, the larger deficiencies persist arguably in the corrective arm. In the
context of the EDP, the adjustment to the deficit criterion for second pillars remains small and
10
It does not eliminate the incentive entirely, as (i) the increase in higher social security contributions from re-
integrating the second pension pillar into the budget continues to lower the recorded fiscal deficit, (ii) pillar II asset
transfers continue to reduce recorded government debt and, relatedly, (iii) interest savings as a result of debt
reduction also continue to lower the recorded fiscal deficit.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 77
subject to many conditions. No flexibility is allowed when assessing compliance with the debt
ceilings. As a result, the current framework continues to discriminate against large second
pension pillars. While ESA 2010 reduces incentives to dissolve Pillar II funds, disincentives to
setting up or enlarging Pillar II funds remain.
Frequent ad hoc changes to the fiscal framework, its complexity and the degree of discretion
in its application trigger substantial policy uncertainty.
23. The gap between the EU’s fiscal framework and a fiscally neutral treatment of a
country’s pension regime is illustrated by a comparison between Poland and Germany, using
the concept of the pension-adjusted balance
(PABF) developed by Soto et al. (2011) and
balancethat remains the basis for EDP
has been almost 4 percentage points stronger
pension-adjusted balances were, on average,
almost identical, reflecting cost-saving
and the existence of its second pillar. Put
differently, while Germany accumulated less
explicit debt in this period, it accumulated more
implicit debt than Poland.
11
D. Conclusions
24. Securing fiscally and socially sustainable pension systems remains a challenge. With
currently legislated parameters, most NMS-6 PAYG systems will pay much lower pensions in the
future, creating a risk of old-age poverty. Absent increases in the retirement age, such an outcome
can be avoided only by accepting higher fiscal burdens or by increasing savings ahead of time in
order to generate additional income out of which pensions can be paid. With countries emerging
from the 2008/09 financial crisis, a renewed focus on long-term challenges is appropriate, including
on strengthening, rather than reversing, the momentum of pension reform.
25. Pillar II reforms are one, but not the only, way of generating higher savings for
retirement.
11
While the EC uses some indicators of implicit liabilities in its fiscal analysis, these complementary indicators do not
carry the same weight as the core assessment of compliance with fiscal rules.
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2007
2008
2009
2010
2011
2012
2013
Germany: healine fiscal balance
Germany: PAFB
Poland: headline fiscal balance
Poland: PAFB
Poland vs. Germany: Headline vs. Pension Adjusted Fiscal
Balance (PAFB) (Percent of GDP)
Sources: EC 2009 and 2012 Ageing Report; and IMF staff calculations.
.
CEE NEW MEMBER STATES POLICY FORUM
78 INTERNATIONAL MONETARY FUND
Countries that choose to maintain a second pillar should seek to strengthen the Pillar II
further reducing administrative costs.
12
Moreover, the transition costs of second pillars should at
least in part be absorbed by the budget, which will require generating more fiscal space.
Countries that choose to abolish second pillars need to cope with the cost of ageing in
other ways. This implies the need to improve fiscal performance and increase public savings,
and strengthen incentives for participation in a third, voluntary pillar. A few countries in the
region (the Czech Republic, Slovakia, Romania and Poland) have introduced such schemes in
recent years. Still, and similar to second pillars, the design of voluntary pensions schemes should
be accompanied by sufficient quality control and include the availability of simple savings
products with low administrative costs (Barr, 2013). As mostly higher earners make use of
voluntary schemes, generous tax breaks should be avoided.
For all countries, aligning the retirement age more closely with longevity and reducing
incentives for early retirement is key to ease the trade-off between fiscal and social sustainability
of pension systems.
26. There remains a case for rendering the EU’s fiscal framework more neutral toward a
country’s choice of pension regime. While the reforms SGP reforms of 2011 have moved the
framework some way in this direction, the discriminatory treatment of second pillars persists.
Admittedly, full neutrality is difficult, as it would arguably require moving away from the headline
deficit as the main assessment toolwhich, in turn, has other drawbacks (for example, the headline
deficit can be computed from observed data, while concepts like pension adjustment balance
require parametric assumptions about discount rates, etc.). Still, such an effort is worthwhile to avoid
discouraging countries from pre-funding ageing costs.
12
Sweden, for example, centralizes the administration and maintenance of individual accounts to reduce costs, and
has established a low-cost default fund to compete with other, more sophisticated investment schemes. Similarly, the
choice (currently five broadly based funds). The accounts are maintained centrally, and fund management is on a
wholesale basisthat is, the fund manager knows only the total volume of resources to be managed, not the details
of which worker owns how much.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 79
References
International Monetary Fund). http://www.imf.org/external/pubs/ft/wp/2000/wp00139.pdf.
, 2013, The pension system in Finland: Adequacy, sustainability and system design, The
Finnish Centre for Pensions.
Cuevas, Alfredo, Maria Gonzalez, Davide Lombardo, and Arnoldo Lopez-Pension
Privatization and Country Risk
Implementation of the Stability and Growth Pact and Guidelines on the Format and Content
of Stability and Convergence Pr
http://ec.europa.eu/economy_finance/economic_governance/sgp/pdf/coc/code_of_conduct_
en.pdf.
Epstein, Natan, and
Report No. 11/167 (Washington: International Monetary Fund).
http://www.imf.org/external/pubs/ft/scr/2011/cr11167.pdf.
European Commission, 2012, The 2012 Ageing Report: Economic and budgetary projections for the
EU27 Member States (2010-2060).
http://epp.eurostat.ec.europa.eu/cache/ITY_PUBLIC/202032004-BP/EN/202032004-BP-
EN.HTML.
Occasional Paper No. 151, May.
Lassila, Jukka, Niku Määttänen, and Tarmo Valkonen, 2014, Linking retirement age to life expectancy
what happens to working lives and income distribution? Finnish Centre for Pensions,
Reports 02/2014.
Sinn, Hans-Werner, 2000, Why a Funded Pension System Is Useful and Why It Is Not Useful, NBER
Working Paper No. 7592 (Cambridge, Massachusetts: National Bureau of Economic
Research). http://ideas.repec.org/p/nbr/nberwo/7592.html.
CEE NEW MEMBER STATES POLICY FORUM
80 INTERNATIONAL MONETARY FUND
Schwarz, Anita M., Omar S. Arias. 2014, The Inverting Pyramid: Pension Systems Facing Demographic
Challenges in Europe and Central Asia. DOI: 10.1596/978-0-8213-9908-8. Washington, DC:
World Bank.
Soto, Mauricio, Benedict Clements, and Frank Eich, 2011, A Fiscal Indicator for Assessing First and
Second pillar Pension Reforms, IMF Staff Discussion Note, SDN/11/09.
World Bank, 1994, Averting the Old Age Crisis: Policies to Protect the Old and Promote Growth
(Washington).
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 81
NMS-6 pay-as-you-go (PAYG) pension schemes and key parameters
Country Type
Earnings
reference
Valorization
variable
Benefit indexation
variables
Statutory
retirement age
BG DB Full career Wages Prices and wages M=63; F=60
CR PS Full career 1/ Prices and wages Prices and wages M=65; F=60.3
CZ DB Full career Wages
Prices and wages 3/
M=63.8; F=60.8
HU DB Full career 2/ Wages Prices M=62; F=62
PL NDC Full career
Wages Prices and wages M=65; F=60
RO PS Full career
Prices (and wages
until 2030)
Prices (and wages
until 2030)
M=64; F=59
Sources: Europen Commission; OECD; World Bank, and IMF staff
DB = Defined benefit; NDC = Notional defined contribution; PS = Point system
2/ Net pay from 2008, moving toward full career.
3/ Suspended in 2009.
1/ For those who are in both PAYG and Pillar II systems, benefits for post-2002 years of service are based
on a basic pension plus the second pillar annuity.
Appendix I. Pension Systems in NMS-6 and Recent Pension
Reforms
New Member States Pension Systems–an Overview
The pension systems of the new member states (Bulgaria, Croatia, Czech Republic, Hungary, Poland,
and Romania, collectively NMS-6) are similar: all NMS-6 provide the bulk of pension entitlements
through statutory pay-as-you-go public pension (PAYG) systems, which generally cover old-age,
irst five components are
provided on an earnings-related basis, while minimum pensions are either means-tested or
delivered as social assistance. Bulgaria, Croatia, the Czech Republic, Hungary, and Romania provide a
defined-benefit (or similar) old-age pension. Poland has a notional defined-contribution system,
account value by a divisor reflecting life expectancy at the date of retirement.
Over the last two decades, the NMS-6 added mandatory second pillars (or a third, voluntary pillar
with state contributions and tax incentives in the case the Czech Republic). Contributions to these
pillars are administrated by private pension managers.
NMS-6 policies concerning drivers of future old-age pension spending are also similar (see table
below). All the member states apply pension benefit formulas in which full career earnings are taken
as a reference to calculate pension entitlements. All contributions paid before retirement are
indexed to wages (valorization)Croatia incorporates prices into the formula as well. Wages, along
with prices, are also used to index pension benefits.
In 2010, the statutory
retirement age was less than
65 for female participants in all
NMS-6 and male participants
other than Polish and Croatian
men. However, in all NMS-6
the statutory male retirement
age will increase gradually to
65 or above by 2022.
Moreover, the Czech
provides for a continuous
increase in the retirement age. Available data on accrual ratesthe annual pension earned through
participation in the PAYG systempaint a mixed picture, with rates rising in Bulgaria and Hungary
over the next 40 years, and falling rates in the Czech Republic.
CEE NEW MEMBER STATES POLICY FORUM
82 INTERNATIONAL MONETARY FUND
Recent Pension Reforms
Bulgaria. Since 2012, the retirement age started to increase by 4 months per year until reaching
65 years of age for men (63 for women), and the required length of insurance started to increase by
four months per year until reaching 40 years for men (37 for women). Pensions were indexed to CPI
only and eligibility requirements for military and police pensions were tightened and contributions
increased. In 2013, significant reform reversals were announced; including (i) a return to th
and (ii) the gradual increase in the retirement age was halted until at least 2014.
Czech Republic. The reform adopted in 2011 increased the statutory retirement age, reduced
disability pensions, curtailed the rate of progressivity in the assessment of contributions, and
extended the insurance period required for accessing a full pension. These and other changes to the
pay-as-you-go (PAYG) system have cut its long-term deficits from 45 percent of GDP to around
2 percent of GDP in 204060, and to less than 1 percent of GDP from 2070. The statutory retirement
age is gradually increased by two months per birth cohort without any upper limit for men (and later
on for women too). The pension eligibility age for women is increased by four months and from
2019 by six months to be unified with that of men. In 2012, a voluntary second pillar (with partial
diversion of premiums from the PAYG plan) was legislated, but the take-up was very low, and the
pillar is expected to be soon abolished.
Croatia. In 2010, female retirement age is raised to 65 for women by 2030. Early retirement age for
both genders was also increased, and a modest late retirement bonus was introduced. In 2013, a
further increase in the retirement age to 67 was legislated, along with the relaxation of the early
retirement rules and more generous indexation. In early 2012, the government abolished privileged
pensions of government officials and members of parliament to improve equity in the pension
system. In August 2012, the list of military occupations subject to early retirement with extended
service period was rationalized. The government reduced pensions that are above HRK 5,000 by
10 percent in December 2013 and conditioned the indexation of privileged part of the pension
benefit with growth and fiscal parameters.
Hungary. Reforms taken in 2008 included eliminating the 13th month public pensions, and
replacing the combined price-wage indexation of pensions with pure price indexation. According to
the 2009 pension reform, since 2014, the statutory retirement age has been gradually increased (by
half a year for every age cohort), with the objective of reaching 65 years in 2021 for those born in
1957 and thereafter. Measures to reduce or eliminate early retirement schemes, terminate special
retirement rules for armed forces, tighten conditions for disability pension eligibility, and overhaul of
allowances were also introduced in 2012. Moreover, to address fiscal pressures and contain public
debt, the government made changes to the second pillar in 2011. Specifically, from November 2010
to December 2011, contributions from mandatory DC plans were diverted to the public scheme, and
the mandatory DC scheme became voluntary in December 2011 with its assets transferred to the
government. Finally, in 2013, the upper ceiling on pension contributions was terminated.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 83
Poland. Retirement ages will gradually increase to 67 from 65 over the period 2013 to 2020 (men)
and 2040 (women). Early retirement (at 62 for women and 65 for men) is possible with pension
reduced by 50% (2012). Several early retirement schemes were abolished at the beginning of 2009.
It is possible to defer both the notional and the funded, defined-contribution pension component
without any age limits. In 2014, the second pillar was scaled-back with the transfer of about half of
pension fund assets (and corresponding liabilities) to the PAYG plan. The changes also entailed, inter
alia, a further redirection of contributions to the PAYG system which initially began in 2011, and the
centralization of the payout phase in the PAYG plan.
Romania. Key reforms taken include the increase of the retirement age 63 years for woman females
(65 for man) by January 1, 2015, and the equalization of both ages at 65 years by January 1, 2030;
and the corresponding contribution period for receiving the full old-age pension. The valorization
and indexation of pension benefits will change gradually to inflation by 2030. Other changes include
reducing the attractiveness of early retirement; tighter eligibility for invalidity pensions; creation of a
Guarantee Fund, funded by private pension operators, to backstop minimum investment return
guarantee (real amount of contributions less commissions) for contributors to Pillar II; and the
consolidation of special-sector pension schemes with the overall public pensions scheme. Lately
there has been pressure to re-establish some of the privileged pensions.
CEE NEW MEMBER STATES POLICY FORUM
84 INTERNATIONAL MONETARY FUND
Appendix II. Evolution of Pillar II Systems and Contribution Rates
Table. Evolution of Pillar II. Systems and Contribution rates
1/
Intro Voluntary/mandatory
SSC of PAYG
participants 1/
Current status of Pillar II
PAYG SSC Pillar II SSC Total
Bulgaria Jan-02
2010
16.0% 11.0% 5.0 % (universal funds) no change 5/
2011-14
17.8% 12.8% 5.0 % (universal funds)
Croatia Jan-02
mandatory for those borne
after 31/12/1961
20.0% 15.0% 5 % 3/ 20% no change
Czech Rep. Jan-13
voluntary, but upon entry no
withdrawal from Pillar II
28.0% 25.0% (3+2)% 7/ (28+2)%
likely to be abolished (merged
into Pillar III)
Hungary Jan-98 1998 31.0% 25.0% 6.0% 31.0% Abolished
1999-2000 30.0% 24.0% 6.0% 30.0%
2001 28.0% 22.0% 6.0% 28.0%
2002-03 26.0% 20.0% 6.0% 26.0%
2004-06
26.5% 18.5% 8.0% 26.5%
2007
29.5% 21.5% 8.0% 29.5%
2008-10
33.5% 25.5% 8.0% 33.5%
2011
34.0% .. 0.0% ..
2012
37.0% .. 0.0% ..
Poland 4/ Jan-99
up to 2010
19.5% 12.3% 7.3% 19.6%
partially abolished -
investments in instruments
other than (domestic)
government bonds remain
2011-12
19.5% 17.2% 2.3% 19.5%
2013
19.5% 16.7% 2.8% 19.5%
Romania Jan-07
2007
31.3% 29.3% 2.0% 31.3% restoration in progress
2008
31.3% 29.3% 2.0% 31.3%
2009
31.3% 29.3% 2.0% 31.3%
2010
31.3% 28.8% 2.5% 31.3%
2011
31.3% 28.3% 3.0% 31.3%
2012
31.3% 27.8% 3.5% 31.3%
2013
31.3% 27.3% 4.0% 31.3%
Other CEE countries
Estonia Jan-02
up to Jun-09
20.0% 16.0% (4+2)% 2/ (20+2)% fully restored
Jul-09 to Dec-10 20.0% 20.0% (0+2)% (20+2)%
2011 20.0% 19.0% 3.0% (20+2)%
2012 20.0% 16.0% (4+2)% (20+2)%
Latvia 4/ Jul-01
up to 2006
20.0% 18.0% 2.0% 20.0% partially restored
2007 20.0% 16.0% 4.0% 20.0%
2008 to Apr-09 20.0% 12.0% 8.0% 20.0%
May-09 to Dec-12 20.0% 18.0% 2.0% 20.0%
2013 20.0% 16.0% 4.0% 20.0%
2014 20.0% 14.0% 6.0% 20.0%
Lithuania Jan-04
2004
26.3% 23.8% 2.5% 26.3% partially restored
2005 26.3% 22.8% 3.5% 26.3%
2006 26.3% 21.8% 4.5% 26.3%
2007 26.3% 20.8% 5.5% 26.3%
2008
26.3% 20.8% 5.5% 26.3%
2009H1 26.3% 23.3% 3.0% 26.3%
2009H2-2011 26.3% 24.3% 2.0% 26.3%
2012 26.3% 24.8% 1.5% 26.3%
2013
26.3% 23.8% 2.5% 26.3%
2014 6/ 26.3% 24.3% (2+1+1)% (26.3+1+1)%
Slovakia Jan-05 up to sept 2012 18.0% 9.0% 9.0% 18.0% partially reversed
2013 6/ 18.0% 14.0% 4.0% 18.0%
1/ SSC to finance pensions only.
2/ 4 ppt of mandatory SSC of the state are redirected to Pillar II. Individuals add 2 ppts of supplementary individual contributions.
3/Inital plan was to increase rate to 10 percent by 2009 but initial law never set a schedule for the increase to take place.
4/ Notionally defined contribution systems.
5/ In 2000, a second pillar-type sytem for workers in hazardous occupations was introduced with the aim to provide for early retirement.
The 2002 reform introduced a mandatory second pillar for all employees .
6/ contributors to the second pillar were also allowed to leave and return to PAYG.
7/ 2 percent supplementary contributions by individuals.
SSC of Pillar II participants 1/
Table. Evolution of Pillar II Systems and Contribution rates 1/
mandatory for those borne
after 31/12/1959
mandatory for new entrants,
voluntary for others
Mandatory for employees
aged 51 and younger;
voluntary for those aged 52
voluntary , but upon entry no
withdrawal from Pillar II
allowed
mandatory for persons born
in 1983 or later, voluntary for
others
mandatory for those borne
after 31/12/1971, voluntary
for those borne between 1962
and 1971
mandatory for those borne
after 31/12/1968, voluntary
for those borne between 1949
and 1969
mandatory for those born
after 1971, voluntary for
those born between 1953 to
1971, those born before 1953
not qualified to join pillar II
2014-17: rate may rise to 8 %
to make up for reduced
revenue
plan to increase pillar II from
2017
1 from individuals, 1 from
other state budget resources
plan to further increase
funding going to Pillar II to 6
percent in 2016 and to 7.5
percent in 2020 (partially
financed from non-SSC
revenues and additional
individual contributions).
plan to increase to 6 % in 0.5
steps
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 85
MAKING THE MOST OF THE EU SINGLE MARKET
1
Summary
Since 1995, NMS-6
2
as a group has experienced a spectacular increase in exports to the EU.
However, this masks considerable heterogeneity across these countries: the Czech Republic,
Hungary and Poland have steadily increased exports to the EU as a share of GDP, while the
performance of Bulgaria and Romania has been less stellar.
Structural and institutional factors explain a significant part of the variance in export
performance. In particular, human capital, labor skills, foreign investment environment, and wage
competitiveness are found to be highly significant in explaining export performance of NMS in the
EU market as are reforms that help countries link up with global supply chains. For countries where
there is a significant room for increasing exports to the EU, these reforms can help maximize
benefits of access to the larger market.
While higher exports help growth, quality improvement is important for sustained income
convergence. For NMS-6, overall quality of exports is relatively high when compared to that of
other exporters of similar products in the world market, but less so when assessed in the EU
market. In both markets, Romania and Bulgaria show the largest room for quality improvement
while Hungary and Czech Republic, countries specializing in mid-quality products, show the least
room tend for improvement among the NMS.
Policy priorities to improve export quality need to be mindful of initial conditions. For
countries specializing at lower end products, priorities include a better foreign investment regime
and higher links with supply chains that would allow access to technology. For countries
specializing in higher quality products, innovation via sustained pursuit of higher education and
R&D spending is key to further improvement of quality.
Significant boost in exports can come from further liberalization of services trade within the
EU. The implementation of the EU Services Directive (SD) resulted in a sizable reduction in import
restrictions and the NMS-6 seems to be benefiting more than other EU members. However,
significant barriers remain in products where the NMS-6 hold a comparative advantage, including
in professional and technical services product. Further liberalization by EU members and a new
impetus to liberalize services trade will help NMS-6 enhance their exports to the single market.
1
Prepared by Jesmin Rahman, Ara Stepanyan, Jessie Yang and Li Zeng. The authors are grateful to Marinela Petrova
(Bulgaria Ministry of Finance) and Christian Buelens (European Central Bank) for discussing the paper at the New
Member States (NMS) Policy Forum in Warsaw on December 12, 2014, to Hylke Vandenbussche of the European
Commission for providing data on export quality and to the NMS-6 country representatives that reviewed the paper
during bilateral visits in November 2014.
2
NMS-6 includes Bulgaria, Croatia, Czech Republic, Hungary, Poland and Romania. NMS includes these countries
and NMS-EA consisting of Estonia, Latvia, Lithuania, Slovakia and Slovenia.
CEE NEW MEMBER STATES POLICY FORUM
86 INTERNATIONAL MONETARY FUND
A. Introduction
1. Being part of the European Union (EU) allows new member states (NMS) access to a
larger market for their products and provides an anchor for growth and convergence.
3
The EU
single market provides opportunities for firms to grow, and, at the same time, subjects them to
stronger competition raising
incentives to improve
productivity. The open trade
and investment regime in turn
also acts as a conduit for
technology transfer that over
time improves quality of
exports. Higher exports and
quality create a virtuous cycle
of growth and convergence
(Hausmann et al. (2007), see
text chart).
2. In this paper, we
examine export performance
of NMS in the EU single
market, with a focus on the
six new member states which
are not yet part of the euro
area (NMS-6): Bulgaria,
Croatia, Czech Republic, Hungary, Poland and Romania. We focus on the following questions: How
successful have these countries been in taking advantage of their unrestricted access to the EU
single market?
4
What structural factors matter most? Has export integration with the EU been
associated with improvement in export quality? What role has services sector played in exports? And
going forward can services exports play a bigger role?
3
NMS includes 11 member states that joined the EU during 200413. They are split into two groups: NMS-6 and
NMS-EA. NMS-6 includes Bulgaria, Croatia, Czech Republic, Hungary, Poland and Romania, and NMS-EA includes
countries that have joined the euro area in recent years: Estonia, Latvia, Lithuania, Slovakia and Slovenia
study refers to all other EU members.
4
While the single market provides an opportunity for all EU firms, there may be domestic factors that prevent firms
from taking full advantage of this opportunity - such as preferential treatment of incumbents or excessive entry
regulations. These obstacles that may narrow the scope of unrestricted access to the EU single market for goods are
outside the scope of this paper.
Export Quality and GDP Per Capita, 2010
Source: WDI, Henn, Papageorgiou, and Spatafora (2013), and IMF staff
calculations.
ROU
BGR
HRV
HUN
POL
CZE
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
0 20000 40000 60000 80000
Export Quality Index
GDP per capita in USD, PPP
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 87
Figure 1. NMS: Gross and Value Added Exports of Goods and Services
Note: More up-to-date data for value added exports of goods and services could not be shown as the World
Input Output Table does not go beyond 2011.
Source: Staff calculations using World Input Output Table, Eurostat and Haver Analytics.
B. Evolution of Exports to EU: Relative Success and its Determinants
3. NMS-6 shows a varying degree of success in taking advantage of the EU single market.
As a group, NMS-6 export more goods and services to the EU than other members measured in
both gross and value added terms (scaled by GDP) (Figure 1).
However, within NMS-6, there is a
range. The Czech Republic and Hungary, being among the most open economies in the world,
derive a quarter of their GDP from value added exports to the EU, while this share is less than one
tenth for Romania. Since the crisis, exports have played a stronger role in growth counting for a
higher share of GDP in all NMS-6 countries except Croatia. This has been driven by higher exports to
the EU, as well as outside the EU (Figure 1).
0
5
10
15
20
25
30
CZE
HUN
POL
BGR
ROU
NMS-6
NMS-EA
Other-
EU
Value Added Exports to EU, 2011
(Percent of GDP)
0
10
20
30
40
50
60
70
80
HUN
CZE
BGR
POL
HRV
ROU
NMS-6
NMS-EA
Other-
EU
Gross Exports to EU, 2013
(Percent of GDP)
0
10
20
30
40
50
60
70
80
90
100
HUN
CZE
BGR
POL
HRV
ROU
NMS
-6
NMS
-EA
Other
-EU
Gross Exports of Goods and Services, 2007-13
(Percent of GDP)
Rest of the World
EU
CEE NEW MEMBER STATES POLICY FORUM
88 INTERNATIONAL MONETARY FUND
4. What factors explain the varying export performance in the single market? We
investigate this question empirically in a sample of ten NMS for the period 2003-11.
5
Our variable of
interest is value added exports to the EU. Scaled by GDP, this variable tells us what share of
economic activity in NMS is generated by import demand from the EU single market. We chose
value added as opposed to gross exports since the former measures exports more accurately taking
out re-exports and imported inputs.
6
5. In what follows, we examine the role of structural factors in export performance. The
choice of structural variables draws on trade literature which emphasizes the importance of human
capital (Bougheas and Riezman 2007, and Bombardini et al. 2012) and institutional quality
(e.g. Anderson and Marcouiller 2002, and Levchenko 2007). These factors affect competitiveness and
export performance of a country by influencing the overall environment in which firms operate. In
selecting variables, we started with a large set that include human capital, labor market efficiency
and flexibility, foreign investment, physical and virtual infrastructure, and governance. The final
selection was made based on data availability and statistical significance. Below are the five variables
that were included in our preferred regression specification, all of which show a strong correlation
with value added exports to the EU.
7
Human capital. Better human capital improves through expansion of
productive capacity over time. Human capital is proxied by two variables: higher education
(upper secondary or tertiary education attainment) and the share of employed participating
in continuous vocational training and skills upgrade. The second variable, which takes into
account on-the-job training and skill upgrade, also implicitly captures the degree of skills match
in the economy (for example, Card and others (2009) found that vocational and on-the-job
training programs tend to lead to better labor market outcomes).
Labor market efficiency. A well-functioning labor market is critical for ensuring an efficient
allocation of labor force and providing incentives to work. We proxy labor market efficiency with
the two variables: inactivity trap, which captures incentives to stay out of work force either
because after-tax income is too low or social benefits are too generous (a larger value indicates
weaker incentive to work), and minimum wage relative to gross average wage, which
captures wage competitiveness of low-skilled labor.
Foreign investment environment. The importance of foreign direct investment in promoting
exports and technology transfer is well-known. This is particularly so for NMS where foreign
5
We were not able to include Croatia in the regression analysis due to lack of data for value added exports and
many of the structural variables.
6
For robustness check, we also use gross exports of goods and services to the EU and exports (both value added and
gross) to the world as the dependent variable. See forthcoming working paper by Rahman and others for details.
7
The details on data sources can be found in Annex I.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 89
capital from the EU has been a main driver of growth in the last decade. We use foreign
investment environment as the fifth structural variable. This variable is an index based on a
survey that captures prevalence of foreign ownership in a country as well as sentiment regarding
whether current regulations discourage foreign ownership. A higher index indicates a more
conducive environment for foreign ownership.
Participation in supply chains. In addition, we also include a measure of supply chain
integration (the share of exports processed through upstream and downstream supply chains)
given the strong role of supply chains in global and EU exports in the past decade (IMF, 2013;
Rahman and Zhao, 2013). The degree of supply chain integration, which varies across time and
country, captures the effects of other structural and institutional variables that may have
included in the regression due to lack of data: quality of export processing infrastructure,
unobserved regulations or obstacles hindering business operation, availability/cost of utilities
and other inputs, and tax advantages. By including this variable, we have a more complete
coverage of structural factors that are relevant for exports.
6. What about gravity factors that are typically found to be important determinants of
trade flows in the literature? In our regression, the measure of supply chain integration is already
capturing many of the gravity factors, such as distance from destination, domestic market size and
income level. According to Rahman and Zhao (2013), about 88 percent of explained variance in
integration with supply chains in a sample of 40 countries over 15 years is captured by these factors.
Nevertheless, we also include main gravity variables -- income per capita, weighted distance from
export partners, and population of the exporting country (which also controls for the size of the
domestic market and the bias that smaller countries typically have a higher exports-to-GDP ratio
compared to larger countries) in our regression analysis. In addition, we control for demand
growth in partner countries (proxied by weighted PPP real GDP growth or weighted consumer
sentiment in partner countries), and price competitiveness (unit labor cost based real effective
exchange rate, REER-ULC).
7. Estimation results show that structural factors explain much of the variance in value
added exports of the NMS to the EU. The estimation results from the panel OLS are shown in
Table 1. A system regression could not be estimated due to the small sample size. We used two
-10 average (column 2, Table 1). The
five structural variables are all statistically highly significant and together capture more than 80
percent of the explained variance in both versions of the regression.
CEE NEW MEMBER STATES POLICY FORUM
90 INTERNATIONAL MONETARY FUND
Table 1. Determinants of Value-Added Exports of Goods and Services to EU: NMS-10, 2003–11
8. Differences in educational attainment, vocational training and skills upgrade and
foreign investment environment seem to be most significant in explaining differences in
exports from NMS to the EU. The relative importance of structural variables included in the
regression is illustrated in Figure 2, which shows the increase in value added exports to the EU
brought about by a one-standard-deviation improvement (LHS panel). The strongest impact comes
from human capital, in particular continuous participation in vocational training and skills upgrade.
This is consistent with the empirical literature that suggests significant productivity gains from
vocational training (for a survey of literature see OECD (1998) and Descy and Tessaring (2005)). In
the version that uses the distance of value added exports from the group average, foreign
investment environment shows the largest impact with higher education and skills also contributing
strongly (Figure 2, RHS panel).
In levels
Relative to NMS-10
average
Structural variables
Upper secondary or tertiary educational attainment
0.16** 0.33***
Participation in continuous vocational training and skills upgrade
0.15*** 0.14***
Inactivity trap
-0.09*** -0.03*
Relative minimum wage
-0.08** -0.16***
Foreign investment environment
1/
1.0** 2.2***
Control variables
Share of exports processed by supply chain
0.44*** 0.34***
PPP GDP per capita
0.00 0.00
Weighted real GDP growth of trading partners 0.06 0.13
Real effective exchange rate (ULC-based) -0.03 -0.07**
Population 0.57 -0.68
Constant -29.4*** 0.03
Observations 73 73
R-squared 0.858 0.859
Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1.
1/
Higher values indicate lower degree of restrictions.
Estimation results
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 91
Figure 2. Structural Factors: Relative Importance for Exports to the EU
9. We also find participation in supply chains and price competitiveness to be statistically
significant (Table 1). Links with supply chains increase exports with the impact being the second
highest when compared to the impact of structural variables included in the regression (Figure 3 2,
LHS). This highlights the role of supply chains as an important conduit for increasing exports. Higher
REER-ULC decreases exports by eroding competitiveness, although this variable was not significant
in all specifications. In contrast, per capita income level, weighted distance from partner countries,
weighted GDP growth in partner countries and population come out as statistically insignificant
although with the expected signs; this could be due to the fact that supply chain participation, which
strongly depends on gravity variables, is partly capturing their impact on exports. We also used time
dummies and dummy variable for the euro area crisis, which were found statistically insignificant.
We did not use country-specific dummy variables as it prevents us from identifying structural factors
that are important for export performance, as most structural variables move slowly over time. This
may imply that the impact of structural variables is somewhat overestimated in our regression.
8
10. Our empirical findings identify country-specific policy priorities. We look at the
relative to its NMS peers. Czech Republic and Hungary show an above average export performance
relative to the other NMS, while Poland, Bulgaria and Romania show a below average performance
during 2003-11. A decomposition of contribution of structural factors based on regression results
shown in Table 1, Column 2 yields the following observations (Figure 3):
8
The coefficients of structural variables are robust to alternative specifications (for details see Rahman and others
(forthcoming).
0.0 0.5 1.0 1.5 2.0 2.5
Vocational training and
skills upgrade
Supply chain
participation
Inactivity trap
Foreign investment and
ownership environment
Higher education
Relative minimum wage
Value added exports to EU (percent of GDP)
Note: This chart illustrates the increase in value added exports to EU that would be
brought about by a one standard deviation improvement in each factor.
Impact of Structural Factors on Value Added Exports to EU
27%
17%
2%
7%
4%
28%
15%
Decompositon of Goodness of Fit 1/
Higher education
Vocational training and skills
upgrade
Inactivity trap
Relative minimum wage
Real effective exchange rate
(ULC-based)
Foreign investment and
ownership environment
Supply chain participation
1/ The decomposition is done according to Shapley value.
CEE NEW MEMBER STATES POLICY FORUM
92 INTERNATIONAL MONETARY FUND
Figure 3. NMS-6: Contribution of Structural Factors to Relative Export Performance in the
EU Market
Source: Staff calculations using regression results in Table 1, column 2.
-2
0
2
4
6
8
10
2003-2006
2007-2010
2011
The Czech Republic
-6
-4
-2
0
2
4
6
8
10
2003-2006
2007-2010
2011
Hungary
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
2003-2006
2007-2010
2011
Poland
-8
-6
-4
-2
0
2
4
6
2003-2006
2007-2010
2011
Bulgaria
-10
-8
-6
-4
-2
0
2
4
2003-2006
2007-2010
2011
Romania
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 93
For the Czech Republic, vocational training and skills upgrade, higher education, a favorable
foreign investment environment and links with supply chains have all contributed positively
during 2003-11, while labor market variables, both wage cost and incentives to work, have not
been a source of competitiveness. For Hungary, wage competitiveness, strong links with supply
chains and foreign investment environment have contributed positively, with contribution from
foreign investment environment declining in recent years. Vocational training and higher
performance relative to the NMS.
For Poland, we have seen an improvement in export performance over time with the gap relative
to the other NMS decreasing over time. Higher education and competitive wages have
contributed positively, while a relatively lower degree of participation in supply chains has been
a drag. Although foreign investment environment contributes negatively, in recent years Poland
has seen improvement in foreign investment environment and a pick-up in off-shoring and
outsourcing of business services (McKenzie 2013). These factors have boosted
to the EU since 2010. For Bulgaria and Romania, the below-average performance in exports has
been persistent with the gap relative to the NMS average worsening over time. This highlights
the need for broad-based reforms, particularly in the areas of human capital and foreign
investment environment (Figure 4).
C. Export Quality in NMS-6: Room for Growth
11. In NMS-6, export quality is generally high when compared to the rest of the world. We
assess the quality of merchandise exports to the world using an index developed by Henn,
Papageorgiou, and Spatafora (2013) based on an estimated relationship between export quality,
export unit value, production cost, and the distance from importers.
9
The overall exports quality for NMS-6 is above the 60
th
percentile when compared to all
countries in the world (Figure 4). The Czech Republic leads with an overall quality level close to
the 90
th
percentile and Romania, at 61
st
percentile, lags others. Our analysis of quality at a more
disaggregated product level shows three tiers
10
: the Czech Republic in the highest tier where the
quality of export goods ranges between 61
st
and 97
th
percentiles; Croatia, Hungary, and Poland in
9
The intuition behind this approach is that after controlling for the production costs and taking into account the fact
that exports to more distant destinations tend to be tilted towards higher-priced goods (because of higher shipping
costs), higher quality goods would have higher export unit values. This approach does not assume that international
export markets are competitive where individual exporters can be price takers. This index defines export quality
within a product group rather than across the entire spectrum of export products. This means that a county that
exports low-end high-tech products (for example, auto parts) may show a lower level of overall export quality
compared with a country that exports high-end low-tech products (for example, designer clothing). The overall
export quality is aggregated using quality indices calculated at the SITC 4-digit product level and export weights.
10
For this product level exercise, we only looked at SITC 4-digit products where a country has a market share of at
least ½ percent in the world.
CEE NEW MEMBER STATES POLICY FORUM
94 INTERNATIONAL MONETARY FUND
the next tier, where the quality ranges between 40
th
and 89
th
percentiles; and Bulgaria and
Romania in the third tier, where the quality ranges between 24
th
and 80
th
percentiles. In other
words, underneath an overall high quality, there seems to be a wide range even for countries
like the Czech Republic.
s standing in
the quality ladder relative to others, but also the average quality absorbed by its importers. A
positive gap indicates the quality demanded by importers is larger than that provided by the
exporting NMS. Our analysis reveals positive quality gaps for the NMS at the overall and product
levels (Figure 4). A look at the SITC 4-digit products indicates significant room for quality
improvement in the following areas: textile products (Bulgaria, Croatia, and Romania); wood
products (Romania), paper products (Poland), beverages and tobacco products (Bulgaria and
Poland), and in footwear products (Croatia) (Figure 4).
12. A comparison of export quality in the EU market using a different methodology draws
somewhat different conclusions. Based on Di Comite, Thisse and Vandenbussche (2014), which
uses firm-level cost data in a mark-up model to capture quality of exports in the EU market relative
to other EU exporters, the NMS-6 show a quality distribution that is concentrated at the low
(Bulgaria, Romania and Poland) and middle (Hungary and Czech Republic) part of the quality
spectrum (see chart). The share of products where quality is below the 50
th
percentile relative to
other EU countries ranges between 62 percent in Poland to 52 percent in Hungary. In other words,
more than half of export
products from the NMS
are in the bottom half of
the quality ladder
showing significant room
for improvement with
respect to other exporters
in the EU market.
Relatively speaking, the
Czech Republic and
Hungary still come out at
the top among the NMS
just as they do with
respect to the world
market with export
products concentrated in
the mid-quality range.
NMS-6: Quality of Exports to the EU Market
Source: Di Comite, Thisse and Vandenbussche (2014).
Note: Quality ranks are normalized b
0
20
40
60
80
100
BGR
CZE
HUN
POL
ROU
Distribution of Export Quality, 2011
(Percent of total exports values to EU15)
0.75 - 1
0.5 - 0.75
0.25 - 0.5
0 - 0.25
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 95
Figure 4. NMS-6: Export Quality and Room for Improvement
uality distribution in the world.
Sources: Staff calculations using Henn, Papageorgiou, and Spatafora (2013) database.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
CZE
HRV
HUN
POL
BGR
ROU
Percentiles
Export Quality, 2010 1/
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
CZE
HRV
HUN
POL
BGR
ROU
Relative percentiles
Room for Quality Upgrade, 2010 2/
0 0.04 0.08 0.12 0.16
Food
Beverages and tobacco
Minerals
Textile products
Non metallic mineral manufactures
Iron and steel
Manufactures of metal
Relative percentiles
Bulgaria: Room for Quality Upgrade
0 0.04 0.08 0.12 0.16
Leather products
Textile products
Footwear
Wood and products
Non metallic mineral manufactures
Transport equipment
Manufact goods
Relative percentiles
Romania: Room for Quality Upgrade
0 0.04 0.08 0.12 0.16
Food
Beverages and tobacco
Wood and products
Paper and products
Rubber manufactures
Non metallic mineral manufactures
Iron and steel
Transport equipment
Manufact goods
Relative percentiles
Poland: Room for Quality Upgrade
0 0.05 0.1 0.15
Wood and products
Footwear
Food
Leather products
Textile products
Iron and steel
Chemical products
Relative percentiles
Croatia: Room for Quality Upgrade
CEE NEW MEMBER STATES POLICY FORUM
96 INTERNATIONAL MONETARY FUND
13. For the NMS, export quality is found to be positively correlated with export value in
both EU and world markets (see figure below on Exports Value and Quality). This is not surprising
as most NMS have experienced an improvement in quality of export producing since 2005 as
exports also grew. Vandenbussche (2014) finds that the estimated price elasticity of quality is around
profits and market share. This shows causality from quality to higher exports. Going forward, this
positive relationship between quality and exports is likely to strengthen. Globally speaking, the NMS
are not countries where low labor costs or labor abundance could be a source of comparative
advantage given ageing population, although this may be currently the case relative to advanced
Europe. So improving quality has to be a part of the strategy to enhance exports and our analysis
shows significant room for improvement particularly in the EU market.
14. How can countries improve export quality over time?
A survey of literature shows that some of the structural reforms that explain differences in
exports performance in our regression analysis are also the ones that tend to explain differences
in export quality: human capital, institutional quality, and foreign investment (see for example
Zhu et al, 2009, Henn et al, 2013 and Weldemicael, 2012). In addition, R&D expenditure is also
important for quality improvement.
The EBRD 2014 Transition Report, which looks at innovation and knowledge-based growth in
transition economies, finds that different factors matter in quality improvement at different
levels of economic development and product quality. At a relatively low level of development
and product quality, when countries are trying to access technology, openness and facilitation of
foreign investment are important. The study also finds firms that are part of the global supply
chains to be more innovative than non-linked firms. The capacity to absorb such technology and
replicate depends on the quality of secondary and undergraduate education, and the
effectiveness of on-the-job training. Thus, to innovate over learnt technology depends on
postgraduate education, quality of scientists and engineers, quality of scientific research,
NMS: Exports Value and Quality
Source: For weighted export quality, staff calculations using Henn, Papgeorgiou and Spatafora (2013) for world exports and Di
Comite, Thisse and Vandenbussche (2014) for EU exports; For value added exports, staff calculations using world input output data.
CZE
HUN
POL
BGR
ROU
SVK
EST
SVN
LTU
LVA
0
5
10
15
20
25
30
0.25 0.35 0.45 0.55 0.65
Value added exports of goods and
services (percent of GDP)
Weighted quality of exports
EU Market
BGR
CZE
HUN
POL
ROU
EST
LVA
LTU
SVK
SVN
20
25
30
35
40
45
50
0.80 0.82 0.84 0.86 0.88 0.90
Value added exports of goods and
services (percent of GDP)
Weighted quality of exports
World Market
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 97
flexibility of product and labor market, effective cooperation between science and industry, and
availability of venture capital which become important for countries with mid-quality products
trying to move up.
15. Based on this, the NMS-6 would have different policy priorities in terms of improving
export quality. For Bulgaria and Romania, the focus should be on improving foreign investment
regime, boosting secondary education, and linking with supply chains which would help with
acquiring technology. For the Czech Republic and Hungary, policies need to focus on improving the
environment for innovation. For these two countries, a comparison outside the region also points to
the need for ramping up higher education and R&D spending (Box 1). Diversification of exports
outside the EU and into new products are other ways to enhance exports for these two given that a
significant part of the GDP is derived from value added exports to the EU.
D. Services Exports: Scope for Further Increase
16. Goods dominate over services in exports from NMS-6 to the EU. Although services
sector contribute to two-thirds of the EU GDP and create 9 out of 10 jobs, its share in intra-EU trade
is low. For the NMS-6, the share of services sector in value added exports is only a third and much
less than that in gross exports (Figure 5). Croatia is the only country among the NMS-6 where
services products, mostly related to the tourism sector, dominate exports to the EU. The lower share
of services in exports, among other things, is explained by specific characteristics of service
products: many services are traditionally non-tradable which can only be delivered at production
location and hence not part of the cross-border trade. But we want to explore whether a lack of
comparative advantage relative to other EU members and restrictive market access may be
contributing to relatively low services exports. When it comes to services exports, the single market
does not work quite as well as it does for goods. Countries face numerous restrictions in the form of
authorization, economic needs test, licenses, territorial restrictions and restrictions on
multidisciplinary activities. In this regard, we want to see to what extent the adoption of Services
Directive (SD) in 2006 has helped reduce barriers to exports.
17. Most services exports from the NMS-6 fall under sectors covered by the SD.
11
The
weighted average share of SD sectors in total services exports from NMS-6 to the EU (70 percent for
NMS-6, almost 90 percent for Croatia) is significantly higher than the NMS-EA (59 percent for NMS-
EA) and other EU (62 percent) (Figure 5). According to the assessment by European Commission, the
implementation of the SD has reduced average restrictions on services imports across products and
countries by about 30 percent since 2006, although with considerable variations (Montiagudi et al,
2012). Exports from the NMS-6 benefited from the SD as services exports from these countries in
11
These include travel, construction services, computer and information services, operational leasing, miscellaneous
business services, royalties, education and other personal, and cultural and recreational services. The rest of the
services exports fall under following six sectors: transportation,
communication services, financial services, insurance services, health, and government services.
CEE NEW MEMBER STATES POLICY FORUM
98 INTERNATIONAL MONETARY FUND
sectors covered by the SD grew more (on average 8 percent annually) than from other EU members
(Figure 5). The increase was even more pronounced when we look at sectors that were most
liberalized after the SD.
Figure 5. NMS-6: Services Exports and Services Directive
Note: Most liberalized sectors include travel agency, real estate agents, tourist guide, and hotels, which are the sectors with
the top percentile percent changes on barriers after the implementation of the Services Directive.
Source: Staff calculations using World Input Output Table, Eurostat and Haver Analytics.
18. However, significant barriers remain regarding services exports, particularly with
regards to professional services. The SD was adopted in 2006 in order to promote competition
and trade in services products. The intended implementation period was 2006-9 during which
member countries were to review their respective regulatory framework for services in order to
identify restrictions that can be removed. However, countries were given considerable leeway in the
sense that pre-existing restrictions could be maintained if they were deemed necessary to protect
public interest and as long as they were non-discriminatory, necessary and proportionate. Countries
worked in clusters of 5 members each for mutual evaluation of abolition/amendment of restrictions.
0
10
20
30
40
50
60
70
80
HUN
CZE
BGR
POL
HRV
ROU
NMS-6
NMS-
EA
Other-
EU
Gross Exports of Goods and Services to EU, 2013
(Percent of GDP)
Services
Goods
0
10
20
30
CZE
HUN
POL
BGR
ROU
NMS-6
NMS-
EA
Other-
EU
Services
Goods
Value Added Exports to EU, 2011
(Percent of GDP)
0
20
40
60
80
100
ROU
POL
HUN
BGR
CZE
HRV
NMS-6
NMS-
EA
Other-
EU
Share of SD Sectors Exports to EU, 2011
(Percent of total services exports to EU)
0
2
4
6
8
10
12
NMS-6
NMS-EA
Other-EU
SD sectors
SD sectors, most liberalized
Growth in Services Exports to EU, 2006-2011
(Percent, annual average)
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 99
Given the broad coverage and the deference to member states for action, liberalization of services
trade under the SD has fallen short of expectations (Corugedo and Ruiz, 2014). Specifically, among
many advanced economies in the EU which are major absorbers of exports from the NMS-6, many
barriers remain on professional and technical services after the implementation of the SD.
19. Our analysis of comparative advantage shows that the NMS-6 would greatly benefit
from further liberalization of services imports by EU member states. Hungary, Bulgaria and
Romania show comparative advantage in higher number of services products than goods
products (Annex II). Croatia, where a RCA analysis based on value added exports was not possible
due to data unavailability, would also likely fall into this group given the high share of services
exports. The Czech Republic, Hungary, Poland and Romania hold comparative advantage in
professional and technical services relative to other EU members (Table 2). The weighted average
share of professional services in total SD exports range between 3040 percent in these four
countries. Further liberalization of services trade, particularly those in professional services, would
greatly help NMS-6 increase exports to the EU.
E. Policy Implications
20. Structural reforms play a key role in maximizing benefits of unrestricted access to the
EU single market. Our analysis shows that improving exports to the EU depends on a competitive
economy underpinned by structural reforms, particularly in the areas of higher education, skills
upgradeforeign investment environment.
Other institutional reforms that promote successful integration with supply chains are also helpful in
enhancing export performance, not just to the EU but to destinations outside the EU.
21. Our analysis identifies some country-specific structural reform priorities that can help
boost export performance. For Bulgaria and Romania, where export performance has been
persistently weak relative to other the NMS, closing the distance with peers will require broad-based
reforms, particularly improvement in skills, education attainment, and foreign investment regime.
For Poland, where export performance has consistently improved over time to almost closing the
Table 2. RCA: Exports on Professional and Technical Services
Sources: Eurostat; and IMF staff calculations. Note: RCA is relative to total services gross exports.
RCA
% of exports in
sectors under SD
Bulgaria 0.6 16
Croatia 0.4 8
Czech Republic 1.1 30
Hungary 1.1 31
Poland 1.1 36
Romania 1.1 40
Estonia 0.8 27
Latvia 0.7 31
Lithuania 0.4 23
Slovak Republic 0.8 20
Slovenia 0.5 16
CEE NEW MEMBER STATES POLICY FORUM
100 INTERNATIONAL MONETARY FUND
gap with the NMS average, a more conducive foreign ownership regime and greater links in
services-based supply chains will help. For Hungary, where the contribution of the operating
environment for foreign investors has declined in recent years, it would be important to strengthen
such an environment to further enhance its integration with the EU.
22. There is room for export quality improvement in all NMS-6. The analysis presented in
this paper, which takes into account both export quality relative to other exporters and quality
demanded by importers, shows that there is room for improvement for exports from the NMS-6,
particularly with respect to other exporters to the EU market. The room for quality improvement is
particularly large for Bulgaria and Romania. We also find a strong positive relationship between
exports value and quality, suggesting that pursuing structural reforms, such as improving human
capital, labor market and business environment, would help increase both exports and quality.
23. For quality improvement, structural reforms need to be mindful of a country’s existing
quality level. For countries producing products at the lower end of quality spectrum such as
Bulgaria and Romania, accessing technology through improving foreign investment environment
and greater links with supply chains are key. Countries that are at the medium-level of quality
spectrum such as the Czech Republic and Hungary, improving skills and higher education, and
innovation through higher R&D spending are priorities.
24. For countries that are already highly integrated with the EU single market and produce
mid-quality products, diversification in products and markets will prove useful. This applies to
the Czech Republic and Hungary, countries that derive a high share of domestic output from
demand in the EU market. Looking outside the EU may be useful for these countries. Given that
export products from these countries are at a relatively high level when compared to other
exporters from the world, increasing quality of existing products would require a significant boost in
R&D expenditure and tertiary education. Diversification of exports into new products would be
another option for these countries to stay on the export-led growth path.
25. Improved market access can significantly increase services exports to the EU. Our
analysis shows that a large part of services exports from the NMS fall under sectors covered by the
SD. A number of the NMS have comparative advantage in these products including in professional
services which remain most restricted. Further dismantling of restrictions by EU members, both
advanced and emerging economies, will help maximize benefits from the single market. In this
regard, a renewed impetus to the SD through third-party review of principles of non-discrimination,
necessity and proportionality to assess public interest may also help.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 101
Box 1. Czech Republic and Hungary: What can be learnt from Korea?
For countries such as the Czech Republic and Hungary, a comparison with countries outside transition
economies may also be instructive despite differences in policy environment. We looked at the experience
-driven countries, such as Japan and Germany, and Koreaa country that
has successfully pursued a sustained period of export-led growth kick-started by the Japanese supply chain
(figure). We focused on the evolution of structural variables indentified in empirical literature as significant for
quality upgrade over time. Since 1970, Japan and Germany have demonstrated a negative room for quality
improvement with respect to what their importers have demanded. This means that they have provided a
quality above and beyond the level demanded by importers helping them maintain market shares and stay as
leaders. Korea joined this group in the early 2000s after a steady improvement in quality. Available time series
data for R&D spending and tertiary education, variables that are identified in literature as important
contributors to quality improvement, shows a ramping up by Korea in both aspects and surpassing the levels of
Germany by early to mid-2000s. For Czech Republic and Hungary, to pursue a similar path of quality upgrade
and movement up the value chain, there seems to be a need for significant improvement on these fronts
(education, R&D expenditure and business environment).
Quality Improvement: Lessons from Korea
Japan and Germany provided a quality above and beyond
that demanded by importers: Korea joined this rank in
2000…
…benefitting from a boost in R&D spending…
… tertiary education
..and a friendly business environment.
1/ Doing business ranking as of 2014, a higher number indicates lower quality.
2/ A higher number for use of talents suggests more efficient use. This index is a composite index of the following indices: pay and productivity;
reliance on professional management; country capacity to retain and attract talent; and relative female participation in the labor force. The value
indicates as of 2014.
Source: The World Bank and World Economic Forum.
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
1970
1974
1978
1982
1986
1990
1994
1998
2002
2006
2010
Percentiles
Room for Quality Upgrade
Korea
Germany
Czech Republic
Japan
0
0.5
1
1.5
2
2.5
3
3.5
4
1996
1998
2000
2002
2004
2006
2008
2010
Percent of GDP
Spending on R&D
Korea
Germany
Czech Republic
Japan
0
5
10
15
20
25
30
35
40
45
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Percent
Share of Labor Force with Tertiary Education
Korea
Germany
Czech Republic
Japan
0
10
20
30
40
50
Czech
Germany
Japan
Korea
Doing Business 1/
3.8
4
4.2
4.4
4.6
4.8
5
5.2
Czech
Germany
Japan
Korea
Use of Talent 2/
CEE NEW MEMBER STATES POLICY FORUM
102 INTERNATIONAL MONETARY FUND
Annex. Data Appendix and Robustness Check for Regression
Analysis for Export Integration
The following table provides definition of variables, sources and statistical properties.
Annex. Table 1. Summary Statistics and Data Sources
Variable
No. of
obs.
Mean
Std.
dev.
Min. Max. Source
Domestic value-added exports
to EU (% of GDP)
90 17.9 5.2 10.0 29.5
World Input-Output Database; World
Economic Outlook database, IMF; and
IMF staff calculations
Upper secondary or tertiary
educational attainment (% of
population aged 20-24 years)
90 85.6 5.5 75.0 94.1
Eurostat
Participation in continuous
vocational training and skills
upgrade (% of total
employed)
77 27.4 14.7 13.3 61.0
LAF database, European
Commission
Inactivity trap 1/ 86 71.6 13.6 42.0 90.0
LAF database, European
Commission
Relative minimum wage
(% of gross average wages)
80 40.3 10.0 23.5 62.4
LAF database, European
Commission
Foreign investment and
ownership environment
80 6.7 1.2 4.3 8.9
Economic Freedom of the World
2013 Annual Report
Share of exports processed
by supply chain (% of gross
exports)
90 76.9 3.8 67.7 82.7
World Input-Output Database and
IMF staff calculations
Weighted real GDP growth of
trading partners
90 0.2 4.1 -11.6 12.2
World Input-Output Database; World
Economic Outlook database, IMF; and
IMF staff calculations
Population 90 10.2 11.0 1.3 38.5
World Economic Outlook database,
IMF
PPP GDP per capita 90 17312.7 5183.8 7828.2 29402.8
World Development Indicators
database, World Bank
GDP weighted distance 90 2197.4 2288.7 423.7 9176.1
GeoDist database, CEPII; World
Economic Outlook database, IMF;
and IMF staff calculations
1/ Inactivity trap is one minus the ratio of difference in net and gross wage, where the difference
is between in-work and out-of work wage for an average worker.
CEE NEW MEMBER STATES POLICY FORUM
INTERNATIONAL MONETARY FUND 103
References
Anderson, James E. and Douglas Marcouiller, 2002, "Insecurity And The Pattern Of Trade: An
Empirical Investigation," The Review of Economics and Statistics, MIT Press, vol. 84(2),
pp. 342352, May.
Bombardini, M., Giovanni Gallipoli, and German Pupato, 2012,
American Economic Review, pp. 102, 232748.
Bougheas, S. and Raymond Riezman, 2007, Journal of
International Economics, pp. 73, 42133.
s: A Meta- The
Economic Journal, 2010, 120, F452-F477.
Cedefop.
Di Comite, F., J. Thisse and H. Vandenbussche, 2014, Veri-zontal Differentiation in Export Markets,
Journal of International Economics, pp. 5066.
Report: Innovation in
Fernandez-Corugedo, Emilio and Esther Perez Ruiz, 2014, "The EU Services Directive: Gains from
Further Liberalization," IMF Working Paper No. 14/113 (Washington: International Monetary
Fund).
Hausmann, Ricardo, Jason Hwang, and Dani RodrikWhat you export matters
Economic Growth.
Henn, Christian, Chris Papageorgiou, and Nikola SpataforaExport Quality in Developing
Countries”, IMF Working Paper No.13/108 (Washington: International Monetary Fund).
International Monetary Fund, 2013, Cluster Report: German Supply Chains (Washington).
Koopman, Robert, Zhi Wang, and Shang-Jin Wei-Added and Double Counting
NBER Working Paper 18579.
CEE NEW MEMBER STATES POLICY FORUM
104 INTERNATIONAL MONETARY FUND
Monteagudi, Jesofa, Aleksander Rutkowski and Dimitri Lorenza
European Commission
Economic Papers 456, June.
Levchencko, Andrei, 2007. "Institutional Quality and International Trade," Review of Economic
Studies, Oxford University Press, vol. 74(3), pp. 791819.
December.
Organisation for Economic Co-operation and Development, 1998, Human Capital Investment. An
International Comparison (Paris: OECD Publishing).
Rahman, Jesmin and Tianli Zhao, 2013,"Export Performance in Europe: What Do We Know from
Supply Chains? IMF Working Papers No. 13/62, (Washington: International Monetary Fund).
Rose, A.K., 2004, "Do We Really Know That the WTO Increases Trade?," American Economic Review,
American Economic Association, vol. 94(1), pp. 98114, March.
Vandenbussche, Hylke, 2014, Economic papers 528, European Commission,
September.
Weldemicael, Ermias, 2012, “Determinants of Export Sophistication”. The University of Melbourne.
Zhu,Shujin, Xiaolan Fu, Mingyong Lai, and Ji Xuan,
SLPTMD Working Paper Series No. 033, University of Oxford.
