Sudden Stop anno 2008: Why Emerging Europe was different
description
Transcript of Sudden Stop anno 2008: Why Emerging Europe was different
1
Sudden Stop anno 2008: Why Emerging Europe
was different
Erik BerglofChief Economist
European Bank for Reconstruction and Development
Sudden Stop anno 2008: Emerging Europe was different
-3.4-4.4
-11.1
-7.8
-11.9-15
-10
-5
0
5
10
15
EmergingEurope
CA andCaucasus
Russia &Ukraine
Latin America Emerging Asia
Per cent
Avg 2007Q4/2008Q1 Avg 2008Q4/2009Q1
Percentage changes in external assets of BIS-reporting banks
Why Emerging Europe was different
• Massive output decline, but
• No traditional emerging market “twin crises” - Despite magnitude of shock
Why?
• Nature of European financial integration
• Policy response – massive and comprehensive
Outline
1. Financial integration and the European transition model: introduction
2. Did financial integration have any tangible benefits?
3. What role did financial integration play in the transmission of the crisis?
4. Did financial integration generate macro-financial vulnerabilities?
5. Policy Response and Lessons
The three pillars of the European transition and convergence model
Political, legal-regulatory integration with EU
Trade integration (both opening, and specifically with the EU)
Financial integration
– Growing external assets and liabilities (but primarily liabilities: via FDI and debt inflows)
– Growing role of EU banking groups
Political, trade, and financial integration have gone hand in hand
0
20
40
60
80
100
New EUMembers
Official EUCandidates
EUAspirants*
EasternPartnershipCountries
Other
0
50
100
150
200
Exports to the EU (left axis) Assets of EU banking groups (right axis)
Per cent of total exports
Per cent of GDP
Financial integration has been rapid, with a boom period from 2004 onwards.
External assets and liabilities as a share of GDP
0
50
100
150
200
25019
94
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Per cent
CEB SEE*EEC RussiaTurkey CA
In CEB and SEE, financial integration has been led by foreign banking groups
0
20
40
60
80
100Per cent CEB SEE EEC Other
Foreign bank asset share, 1998-2008 Foreign bank asset share, end-2008
0
20
40
60
80
100
1998 2000 2002 2004 2006 2008
Per centCEB SEEEEC RussiaTurkey CA
Outline
1. Financial integration and the European transition model: introduction
2. Did financial integration have tangible benefits?
3. What role did financial integration play in the transmission and magnitude of the crisis?
4. Did financial integration generate vulnerabilities that aggravated the crisis?
5. Policy Response and Lessons
The ultimate objective of financial integration: economic growth
Loosen domestic savings constraints to allow more investment
Financial development
– Access to credit allows individuals to access entrepreneurial and educational opportunities,
– Reduced macroeconomic volatility encourages investment
Transfer of skills, technology, and institutions (corporate governance) via FDI
Growth in transition has been associated with capital imports—unlike other regions
Current account balance, per cent of GDP (simple average)
-15
-10
-5
0
5
10
1994 1997 2000 2003 2006Latin America Emerging Asia CEB + SEE
Rising current account deficits have reflected mainly higher investment
10
15
20
25
30
35
1994 1997 2000 2003 2006
-15
-10
-5
0
Saving Investment Current account (right scale)
Current account balance, per cent of GDPSaving and investment, per cent of GDP
In non-transition developing countries, CA surpluses correlated with higher growth
y = 0.1681x + 5.0106
R2 = 0.1201
-2
0
2
4
6
8
10
12
14
-20 -10 0 10
CA/GDP, % (av. 1994-2008)
Gro
wth
GD
P p
er
ca
p,
PP
P
(av
.19
94
-20
08
)
-2
0
2
4
6
8
10
12
14
y = -0.3442x + 5.2813
R2 = 0.2697
02468
101214161820
-15 -10 -5 0 5 10
CA/GDP, % (av. 1994-2008)
Gro
wth
GD
P p
er
ca
p,
PP
P
(av
.19
94
-20
08
)
02468101214161820
… but not in the European transition region.
Non-transition sample Transition sample
Did capital inflows and financial integration cause higher growth in transition countries?
Two approaches
Growth regressions
– Used standard set of controls: initial GDP per capita, life expectancy, trade openness, fiscal balance to GDP ratio, measure for institutional quality
Sector approach
– Key idea: if FI has benefits, it should make sectors with high dependence on external finance grow faster
– Controls for full set of industry and country dummies
Examine effect of capital inflows; levels of financial integration; and asset share of foreign banks
Results: robust evidence backing growth effects of FI in transition economies
Find growth effects in both approaches, and across several proxies for financial integration
Size of growth effect is respectable
– 1 percent of GDP in capital inflows raised average annual growth by 0.15-0.4 percentage points per year
– 10 percentage point higher asset share of foreign banks raised average growth by 0.2-0.4 percentage point per year
– Output in manufacturing firms with average financial dependence grew faster by about 1.5 percentage points per year in high capital inflow countries (75 percentile) than in low capital inflow countries (25 percentile)
No such effects found in non-transition sample
Why is the transition region different?
Hypotheses:
Higher level of financial development
Better institutions (or EU commitment) effect
Threshold effects in financial integration
Find some support for the last idea (with respect to foreign bank presence)
Conclusion (1): Financial integration had tangible growth benefits in the EBRD region
Supported by econometric tests using several methodologies
Magnitude is economically significant
1. Financial integration and the European transition model: introduction
2. Did financial integration have tangible benefits?
3. What role did financial integration play in the transmission and magnitude of the crisis?
4. Did financial integration generate vulnerabilities that aggravated the crisis?
5. Policy response and lessons
Outline
Financial integration was one of the conduits of the international crisis…
But outflows were comparatively modest in parts of the region
-3.4-4.4
-11.1
-7.8
-11.9-15
-10
-5
0
5
10
15
EmergingEurope
CA andCaucasus
Russia &Ukraine
Latin America Emerging Asia
Per cent
Avg 2007Q4/2008Q1 Avg 2008Q4/2009Q1
Percentage changes in external assets of BIS-reporting banks
…so was the collapse of trade in Q4 and Q1…
80
100
120
140
160
180
200
220
240
100
102
104
106
108
110Trade credit*Commodity priceWorld trade volumeEU real GDP (right axis)
Index Index
…resulting in a sharp economic contraction in many countries in the region.
Source: EBRD. Note: For Armenia, Georgia, Kazakhstan, FYR Macedonia, Serbia, and Moldova 2009 Q2 numbers are EBRD projections.
Chart 1: Real GDP Growth (Year-on-year, in percent)
-25
-20
-15
-10
-5
0
5
10
Latv
ia
Esto
nia
Ukra
ine
Turk
ey
Hungary
Georg
ia
Arm
enia
Lithuania
Slo
venia
Czech R
ep.
Cro
atia
Kazakhsta
n
Russia
FY
R M
ac.
Slo
vak
Rep.
Pola
nd
Serb
ia
Rom
ania
Bulg
aria
Mongolia
Mold
ova
Bela
rus
Q4 2008 Q1 2009 Q2 2009
Statistical analysis suggests that foreign bank presence attenuated the outflow
Robust effect
– True for both transition sample and broader developing country sample
– True for both initial shock (Q4 2008 outflows) and Q4 and Q1 2009 combined
Higher foreign bank share of 10 percentage points of assets attenuated Q4 lending outflow by 1.4 percentage points*
*average outflow in transition region was about 6 percent in Q4 2008.
Foreign bank presence is associated with better output performance during the crisis
CzeFYR
GeoAlbBiHPol
MneSrbHun
Kgz
MolSlvRusTur
KazBlr
AzeTaj
BgrHrvRom
EstLtu
Svk
Ukr
LvaArm
-25
-20
-15
-10
-5
0
5
0 20 40 60 80 100 120
Foreign bank ownership, 2007
Ou
tpu
t g
row
th o
ve
r Q
4/0
8-
Q1
/09
, q
oq
, s
.a.
-25
-20
-15
-10
-5
0
5
However, external debt levels are a robust predictor of worse output declines in crisis
-25
-20
-15
-10
-5
0
5
0 20 40 60 80 100 120 140
External debt in percent of GDP, 2007
Ou
tpu
t g
row
th
ov
er
Q4
/08
- Q
1/0
9,
qo
q,
s.a
.
Lat
Ukr
EstSln
ArmLit
Mol
Hun
CroBul
SerKyrKaz
Alb
AzerBel
Geo
Ru TkySlk
Ro TajCze
FYRM
PolBiH
Both effects hold up when considered jointly, and with other controls.
Conclusion (2): Financial integration had a mixed direct role in the crisis
1. Provided a conduit for financial shocks; (obvious: in financial autarky, no contagion)
2. Some aspects of financial integration made the crisis worse: external debt
3. However, foreign bank presence mitigated the output decline
– Interpretation: foreign banks buffered the financing shock because of commitments to subsidiaries.
Outline
1. Financial integration and the European transition model: introduction
2. Did financial integration have tangible benefits?
3. What role did financial integration play in the transmission and magnitude of the crisis?
4. Did financial integration create vulnerabilities that aggravated the crisis?
5. Policy response and Lessons
Financial integration and crisis vulnerabilities: potential channels
1. Led to higher private external debt: a direct expression of financial integration
2. Did financial integration fuel credit booms? Higher output declines (cf. private external debt)
3. Did financial integration bias the currency composition of borrowing toward FX?
No statistical link with output declines; but probably exacerbated decline in some countries, and complicated the management of the crisis
Capital inflows strongly correlated with credit growth during 2005-08
(Per cent)
Hungary
Slovenia
BiH
CroatiaCzech Rep.
Poland
FYR
Bulgaria
Russia
Tajikistan
Serbia
Albania
Lithuania
Moldova
Estonia Romania
KazakhstanAzerbaijan
Ukraine
Latvia
R2 = 0.2988
0
10
20
30
40
50
60
70
80
90
-20 0 20 40 60 80 100 120 140 160 180 200
Median growth of BIS lending between mid-2005 and mid-2007
Av
era
ge
cre
dit
gro
wth
be
twe
en
mid
-20
05
an
d m
id-
20
07
0
10
20
30
40
50
60
70
80
90
Number of credit boom years (= year with credit growth > 2 p.p. of GDP)
0
1
2
3
4
5
6
7
Slo
ve
nia
Cro
ati
a
Es
ton
ia
La
tvia
Po
lan
d
Hu
ng
ary
Slo
va
k R
ep
.
Cze
ch
Re
p.
Lit
hu
an
ia
FY
R M
ac
ed
on
ia
Se
rbia
Bu
lga
ria
Ro
ma
nia
Alb
an
ia
Ka
za
kh
sta
n
Mo
ldo
va
Ru
ss
ia
Tu
rke
y
Uk
rain
e
Mo
ng
oli
a
Arm
en
ia
Ge
org
ia
Ky
rgy
z R
ep
.
1996 - 2001 2002 - 07
CEB SEE EEC and other
Did financial integration contribute to (excessive) credit booms?
Initial levels of financial integrationbelow median 10.9 10.1 29.0 36.2at or above median 13.8 14.5 39.1 31.9
Change in financial integrationbelow median 10.9 14.5 26.1 34.1at or above median 13.8 10.1 42.0 34.1
Initial levels -0.03 0.08 0.04 0.01Change 0.14 0.09 0.52 -0.10
Cross country-correlations with number of credit boom years
Relative frequency of credit boom years (%) 1
1996-2001 2002-07External Assets + Liabilities
Foreign bank share
External Assets + Liabilities
Foreign bank share
Did financial integration encourage FX lending?Background
Standard causes of “liability dollarisation”:
1. Low monetary policy credibility and/or high inflation volatility;
2. Moral hazard associated with pegged regimes (implicit guarantees)
Did foreign financing make liability dollarisation worse?
If foreign financing is in FX (either through parent bank or wholesale market), and banks want to avoid mismatch, they may want to push FX lending.
Foreign bank presence is correlated with a higher share of lending in FX
AZE HUN
LAT
LIT
RUS
UKR
ALB
ARM
BULCRO
EST
KAZ
FYRM
MOLPOL
SER
SLV
TUR
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Asset share of foreign-owned banks
Sh
are
of
fore
ign
cu
rren
cy
len
din
g in
to
tal d
om
esti
c
len
din
g
(but so is L/D ratio, and various other measures of foreign financing)
Did financial integration encourage FX lending? Approach:
1. Firm level regressions based on BEEPS data for 2002-05 LHS variable is currency denomination of last loan
Firm level controls; standard macro + institutional controls (inflation volatility, exchange rate volatility …); add FI variables
2. Test robustness using macro data for same period LHS variable is FX share of bank lending
3. Macro regression over longer (2000-2008) period.
Did financial integration encourage FX lending? Results:
Clear evidence that financial integration had an effect over and above standard causes
Approaches disagree on which measure is main (culprit):
– Firm level regressions: foreign banks (even controlling for other FI measures)
– Macro regressions: more mixed results
– Measures of debt inflows (BIS; L/D ratio) matter more than gross financial integration levels
Conclusion (3): Did financial integration generate macro-financial vulnerabilities?
Yes, but …• Drivers of credit booms and FX lending were fast
inflows, not so much higher levels/stocks• To the extent that stocks were a problem,
it was debt, not FDI stocks• Results not conclusive on role of foreign banks
– Contributed to vulnerabilities as conduits of credit and foreign financing, but little evidence of other effects
– Firm-level evidence on contribution to FX lending – but not always robust in macro regressions
Outline
1. Financial integration and the European transition model: introduction
2. Did financial integration have any tangible benefits?
3. What role did financial integration play in the transmission of the crisis?
4. Did financial integration generate macro-financial vulnerabilities?
5. Policy response and Lessons
Crisis response has been impressive…
• Mature domestic (home and host) policies
• Massive & coordinated international support – IMF resources increased from $250 to $750 bn– EU BOP support raised from €25 to €50 bn– G20 held out substantial rise in MDB funding
• Parent bank engagement
• A new coordination platform– IFI “Vienna Initiative” filled institutional vacuum
The magnitude of official support unprecedented
Official support (percent of GDP)
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Q3 1997: Asian currenciesdevalue sharply
Q4 1997: Closure of Ind. and Thaifinancials, Korean won devalues
Q1 1998: Social unrest in Ind.,Korean debt restructuring,
Q2 1998: Social unrest in Ind.,Russian stock market crash
Q3 1998: Russia default, Braz.stock market crash, LTCM
Q4 2008: Gov. support for largeUS, UK and Swiss banks
Q1 2009: Drop in foreign inflowsand trade
Q2 2009
Korea Thailand
Philippines Indonesia
Hungary Latvia
Romania Ukraine
…yet no time to be complacent
• Second and third round effects of the crisis– Quality of banking portfolios uncertain; rising NPLs– Risks of credit crunch– Rising unemployment
• Regulatory framework still uncertain
=> Use crisis response institutions to mitigate risks to recovery
The Vienna Initiative: Basic Approach
• Incentivise banks (do internalise spillovers)– Regulatory incentives (IMF/EU programs)– Capital infusions (Joint IFI Action Plan)– “Naming/shaming” (memoranda of understanding)
• Intensified collaboration among IFIs– Information-sharing– Within IMF/EU programs (Serbia, Romania…)
Vienna Initiative – next steps
• Weather “second round” effects
• Group “stress testing” – reduce uncertainty
• Manage “controlled deleveraging” of banks
• Restructure real sector and FX exposures
• Build local currency markets (Vienna Plus)
Remaining challenges
• Find appropriate regulatory framework
• Counter unavoidable rise in unemployment
• Shift from private to public sector crisis– Fill large fiscal gaps emerging (Ukraine, Latvia…)– Ensure fiscal burden of bailouts end up in the West
• Build local capital markets…
EC and ECB much needed EC to lead EU response, particularly on
fiscal issues; competition policy evolving ECB targeted liquidity support outside Euro
zone (see Denmark, Sweden or US Fed to Mexico, Brazil)
Reaffirm Euro entry objectives with clear timetable (no rule change) and
De-dollarise and develop domestic capital markets
Lessons
• Financial integration worked but must mitigate risks• Rebalance growth model: more domestic sources• Revamp cross-border collaboration: crisis model• IFI collaboration part of new financial architecture• European integration come out stronger from crisis
… but this is far from over - focus on the next steps…
ANNEX
Unfolding of the crisis in the region
Unfolding of the crisis in the region
Unfolding of the crisis in the region