Transcript Economic policy under exogenous shocks: EMU and future

Economic policy under exogenous
shocks:
EMU and future prospects for CEE
accession countries
by Katerina Kalcheva
Munich Graduate School of Economics
6th European Workshop
“EMU: Current state and Future Prospects”
August, 24 – 31 2003
Overview

Motivation

Theoretical review

Exchange rate credibility and the
transmission of asymmetric shocks

Structural VAR and Bayesian approach
(extension) - Blanchard and Quah (1989), Doan,
Litterman, Sims (1984)
Research Agenda
My thesis focuses on the costs of various shocks and different shock
transmission mechanisms. The aim is to analyze the main channels
of monetary transmission in CEE countries and to explain their
importance in the transmission in case of possible shocks.
Three essays:
1. The impact of exogenous shocks on the different
exchange rate regimes in CEEC
2. Exogenous shocks, volatility and financial contagion
in transition
3. Brinkmanship and speculative attacks
Motivation:
12 accession countries:

-
Group 1 will join EU in May 2004
Hungary, Poland, Slovenia,Czech Republic, Slovak
Republic, Estonia, Latvia, Lithuania, Cyprus and Malta
-
Group 2 is scheduled to join in 2007
Romania and Bulgaria

1.
2.
3.
Three stages before joining EMU:
Each country is required to meet Maastricht criteria
Participation in ERMII in the course of two years
Fix to euro and denomination in certain period
What still matters is the real convergence, i.e.a similarity of
economic cycles in the countries whose intention is to peg
their exchange rates to each other.
As a result…….

Different regimes in the pre EMU phase – the number of EXR
regimes should be reduced

Maastricht criteria in CEEE
Criterion
Inflation
Interest
FX rate
Deficit
Debt
2001
3.3
10Y
7.4
Deviation
±15%
2001
- 3.0%
2001
60.0%
Bulgaria
7.9
5.2
-1.3
-0.9
72.5
Czech Repub
4.7
5.6
-5.5
-3.2
29.0
Estonia
5.8
4.7
-1.2
1.1
6.2
Hungary
8.5
7.0
-4.4
-3.2
64.4
Latvia
2.5
10.7
2.6
-1.9
12.2
Lithuinia
1.3
7.9
8.6
-1.4
29.0
Poland
5.6
8.3
-8.2
-4.0
38.0
34.5
34.9
-33.3
-3.7
31.2
Slovak Rep
7.3
7.8
-1.8
-7.2
42.7
Slovenia
8.5
-
-7.4
-1.3
25.4
Reference
value
Romania
Source: Deutsche Bank Research (2002, p.27)
Volatility
Gov’t
Real
interest revenue/G
DP
rate**
GDP Terms of
trade**
Real
effective
exchange
rate**
Latin America
4,10
3,74
4,40
8,70
12,66
18,00
6,34
13,18
2,31
2,19
Emerging Asia
4,11
5,92
8,65
2,52
1,82
Advanced
countries
2,09
3,73
5,90
2,07
1,02
CEECs*
*1993-2001
**Only Czech republic, Hungary, Poland and Romania
Source: Coricelli (2002)
Acceptable exchange rate regimes under
Maastricht criteria for EMU membership:
No optimal EXR regime - heterogeneity
No euroization: risk of misalignment due to wrong conversion
rate or asymmetric shock
No further adjustment of exchange rate possible
Exposed to the fluctuations of the G3 currencies
Case by case
EMU as soon as possible after EU
25 members rotation principle, transparency and accountancy
Four main questions:












The establishment of an appropriate central parity
Choice of appropriate date of exchange rate fixing
Ability to meet Balassa-Samuelson effect :productivity shock
If the magnitude of destabilization risk in various exchange
rate regime is comparable
Fix
Stabilisation phase Czech Rep.
1990-1994
Estonia
Hungary
Latvia
Lithuania
Malta
Poland
Slovakia
Transition phase
1995-2000
Bulgaria
Estonia
Latvia
Lithuania
Malta
Preparatory phase Bulgaria
2001 - ERMII
Estonia
Latvia
Lithuania
Malta
Intermediate
Float
Cyprus
Bulgaria
Slovenia
Romania
Czech Rep.
Cyprus
Hungary
Poland
Slovakia
Slovenia
Romania
Cyprus
Hungary
Czech Rep.
Poland
Slovakia
Slovenia
Romania
De jure classification according to the IMF. Fix: currency board, conventional peg, narrow band;
Intermediate : tightly managed, broad band; Float : managed float, free float
Source: Coricelli (2002)
Aim of the analysis:




On the road to EMU transition countries will choose
system that combines capital mobility with fixed
but adjustable EXR regimes
However, from theoretical point of view (Friedman
1953, Poole 1970; DeGrauwe 1996, Chang &
Velasco 1998) exists higher vulnerability of fixed
EXR regime to external shocks
The analysis attempts to determine to what extent an
external shock can increase volatility in different
regimes and impair the full EMU membership for
accession countries
The choice of the exchange rate regime is a
significant factor in the way different shocks are
transmitted through the monetary sector
Why are shocks important? Theoretical review


Exchange rate regime : absorber of shocks or a source of
shocks Friedman (1953) and Mundell (1961)
Poole W.(1970), DeGrauwe (1996)

Financial shock will be better offset at fixed exchange rate
Buiter (2001) introduces “new” OCA. Finds that exchange rate
is not just a shock absorber, or part of the transmission
mechanism for fundamental shocks originating outside the
foreign exchange markets, but a source of excess volatility,
unnecessary shocks, instability and misalignment.
Artis and Ehrmann (2002) if shocks are symmetric or
asymmetric. Using SVAR technique - how strongly the
exchange rate responds to asymmetric supply and demand
shocks, if it helps to stabilize the economy, also exchange rate is
driven by shocks in the exchange rate market and whether
these shocks have the potential to distort output or prices
Fidurmuc (2002) optimality of currency union using two types
of criteria: business cycle synchronization and existence of
effective adjustment
Moreno and Trehan (2000) : common external shock explains
between sixty to eighty percent of the variation in the total
number of currency crises over the post Bretton woods period





Real shock will lead to more variability in the output at fixed
exchange rate
Theoretical review (cont’d)



Eichengreen, Rose and Wypolsz (1995) and Kaminsky and
Reinhart (1999) model the influence of external conditions by
focusing on the country specific variables.
The impact of supply and demand shocks also related to the
monetary policy (e.g. the slopes of IS, LM and BP curves) has been
extensively studied by Fry and Lilien (1986), Bayoumi and
Eichengreen (1993), Gross (2001), Fidurmuc and Korhonen
(2003).
Dehejia and Rowe (2001) model fixed exchange rates vs. inflation
targeting vs. price level targeting. The difference is explained in
terms of unforeseen observed price shock. In contrast to the
traditional literature, the authors do not emphasize on the source of
shocks but whether a given shock is observed or unobserved. Price
level targeting best stabilizes output and the expected real exchange
rate and enables the central bank to respond to observed shocks
Why shocks are important for CEEC?





Asymmetric shocks arguably have been the reason
behind the collapse of most fixed exchange rate
systems
Evolution of shocks :price liberalization and structural
change, trade opening, capital liberalization
Begg et al. (2002) warn of the danger of enlarged
capital flows which can increase the probability of
crisis if reversed or of overheating and disinflation if
do not reversed.
Asymmetry between current EMU members and
transition countries (see Fidrmuc and Krohonen
2001; Horvath, 2002) – supply shocks are largely
uncorrelated
Habib (2002): high sensitivity to external shocks
(change in risk premium). Poland and Czech
Republic: Exchange rate follows EBC+ shocks.
Hungary and Slovenia: interest rate reacts.
Why shocks are important for CEEC? (cont’d)




External changes may affect the foreign trade
transactions in the EMU-11 and CEEC comparably. But
different in scale and a presumably passive reaction of
the ECB imply de facto an asymmetric character of these
shocks.
Correlation of shocks may change over time – time
varying coefficients of demand and supply shocks.
Babetski, Boon and Maurel (2002)
The importance of adjustment mechanisms:
Current account surplus, mobility of labor and
capital, price flexibility or a system of fiscal risk
sharing by means of intra-union transfers
Otherwise in the extreme case - withdraw from the
union
Current accounts (as % of GDP)
Source: The Economist intelligence country data 2002
Exchange rate credibility and the transmission of
asymmetric shocks

Traditional argument the “endogeneity” of OCA carries important
implications for the credibility of CEEC exchange rate regimes.

CEEC will satisfy OCA properties ex post

asymmetries will be mitigated by the financial integration
before accession all CEE countries have to choose the most credible
exchange rate regime – the fixed exchange rate regime





contradicts with the prerequisite for fulfillment of Maastricht criteria
However, as I show in the paper the main issue for CEEC is not to
choose the most credible regime but rather which regime will
increase the prospects for real convergence ex ante.
A high incidence of asymmetric shocks, differences in the economic
structure or swings in foreign financing might bring serious
deviations from the criteria if countries rely only on credibility
High pass-through and problem with inflation targeting, esp. in
Slovenia and Hungary e.g. Darvas (2001); Coricelli et al. 2002
Hypotheses to be tested and questions to be answered:








EXR regime matters as shock absorber in the pre EMU
phase
Which EXR regime is mostly affected by external
disturbances ?
Which exogenous shocks can have permanent effect
depending on the exchange rate regime?
What makes different exchange rate regimes
sustainable to external shocks?
Which are the main transmission mechanisms?
How well prepared the accession countries are to
absorb shocks?
What is the speed of after shock adjustment?
Can national or/and supranational regulation smooth
shocks?
What is new, my contribution?




Exogenous shocks and volatility in different
exchange rate regimes - a comparative study
Test empirically using SVAR
In order to have more efficient and reliable
estimates and because of the relatively short
transition period we use Bayesian VAR
Three types of exchange rate regimes – currency
board (Estonia), intermediate (Hungary) and
floating (Czech Republic)
 Impulse responses to innovations in Real
effective exchange rate (REER),
Export+Import/GDP(TOT), M2/International
Reserves(M2R)
 Following Mendis (2002), we create dummy
variable for each year if the regime has been
hit by a shock
Research in progress
So far my research follows Blanchard and Quah (1989),
Doan, Litterman, Sims (1984)
to follow the shock transmission mechanism
 to look at the speed of after shock adjustment
A structural VAR model is constructed (see Hamilton
1994, Green 2003) with three dependent variables – real
effective exchange rate (REER), the ratio of M2 and
foreign reserves (M2R) and export+import / GDP (TOT).
Two lagged model using data from 1994:1 to 2001:4

Source: IFS and DataStream
The joint process can be written as an infinite moving average
representation of disturbances:
REERt = aY + ∑ bRR,jRt-1-j + ∑ bRM, Tt-1-j + ∑ bRT,jMt-1-j +Rt
M2Rt = aE + ∑ bMR,jRt-1-j + ∑ bMT, Tt-1-j+ ∑ bMM,jMt-1-j+Mt
TOTt = aP + ∑ bTR,jREERt-1-j + ∑ bTT, Tt-1-j+ ∑ bTM,jMt-1-j+Tt
Y = (A1 + A2D)Yt-1 + 
Orthogonalized impulse response function:
Yt= a + et +et-1+ et-2+….= (L)et
Y
ts   a
e
s j
t
Choleski Decomposition of the variance
E[Tt]2=[∑ bTR,j]2E[Mt]2 + [∑ bTR,j]2E[Rt]2 + E[Tt]
Summary:



EMU participation requires stronger adjustment efforts
than EU membership
Asymmetric shocks have significant importance for the
CEEC
Forex regime and rates – of decisive importance:
 They reflect the performance of all reshaping
markets and affect them as well
 Large benefits of enlargement
 However risk of destabilizing shocks
 ERM II regimes (+- 15% band or tighter are
particularly vulnerable to speculative attacks)
CEEC current economic situation 2000-2001
External
External
debt/GDP debt/Exports
Current
Account/
GDP
FDI
/GDP
Bulgaria
Czech
Estonia
Hungary
Latvia
Lithuania
Poland
Romania
Slovakia
Slovenia
86.4
42.8
61.4
67.3
65.9
42.9
42.9
27.0
56.3
34.3
148.3
56.2
64.6
97.3
144.0
95.1
214.5
81.7
76.5
58.1
8.3
9.1
6.4
2.6
5.6
3.3
5.9
2.7
10.7
0.2
5.9
4.8
6.8
3.9
6.8
6.0
6.3
3.7
3.7
3.3
avg.
CEECs
52.7
103.6
5.5
5.1
Source: Coricelli (2002)
CURRENCY BOARDS
Estonia
Exchange rate versus Euro
(100 = 01/93)
350
Lithuania
800
700
300
April 1994
Currency board
600
250
500
200
400
300
150
200
100
100
50
0
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Bulgarie
1200
1000
800
600
400
July 1997
Currency board
200
0
-200
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Basis 100 = January 1993
An increase in the real and nominal exchange rates stands for an appreciation.
Source : ECB, IMF
Nominal exchange rate
RER calculated with CPI prices
RER calculated with PPI prices
PEGS
Exchange rate versus Euro
(100 = 01/93)
LATVIA
CZECH REPUBLIC
450
180
Peg
400
160
350
300
April 1997
Managed float
140
250
120
200
150
100
100
80
50
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Nominal exchange rate
Nominal exchange
RER calculated
with CPIrate
prices
Basis 100 = January 1993
RER calculated
with CPI
prices
RER calculated
with PPI
prices
RER calculated with PPI prices
An increase in the real and nominal exchange rates stands for an appreciation.
Source : ECB, IMF
CRAWLING PEGS
Exchange rate versus Euro
(100 = 01/93)
HUNGARY
POLAND
140
200
Crawling peg
180
120
Crawling peg
160
100
80
140
120
April 2001
Float
100
60
80
April 2000
Float
60
40
40
20
20
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Nominal exchange rate
RER calculated with CPI prices
Basis 100 = January 1993
RER calculated with PPI prices
An increase in the real and nominal exchange rates stands for an appreciation.
Source : ECB, IMF
MANAGED FLOATS
Exchange rate versus Euro
Slovak Republic
(100 = 01/93)
160
120
150
110
140
100
130
Slovenia
90
120
80
110
70
100
60
90
80
50
70
40
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Roumanie
250
200
150
100
Nominal exchange rate
50
RER calculated with CPI prices
0
RER calculated with PPI prices
-50
01/93 01/94 01/95 01/96 01/97 01/98 01/99 01/00 01/01 01/02
Basis 100 = January 1993
An increase in the real and nominal exchange rates stands for an appreciation.
Source : ECB, IMF
Maastricht Conditions for EMU Membership:





Inflation: (no more than 1.5% above average of 3 lowest inflation
countries).
Nominal interest rate: ( no more than 2.0% above average of the 3
lowest inflation countries).
Nominal exchange rate:
 Respect normal fluctuation margins for ERM without severe tensions
for at least 2 years before the examination. No devaluation ‘on own
initiative’. [At least 2 years of ERMII plus unrestricted financial
capital mobility: risk of speculative attacks and crises]. Italy
and Finland precedents 1998/9; Greece precedent 2000/1.
 Council of Ministers decides conversion rate.
Fiscal criteria:
 Budget deficit should not be higher than 3% of GDP
 public debt should not be higher 60% of GDP
Central Bank independence
The theory of optimum currency areas
The theory of OCAs is a collection of various economic indicators
determining how a currency area will function after bilateral exchange
rates are fixed.
The most famous characteristics for participation in OCA are:
 countries face symmetrical disturbances (as type, direction and
speed of adjustment) (R.Mundell, 1961),
 degree of factor mobility and similarity of production structures
(R.Mundell, 1961),
 openness of the economy (R.I.McKinnon, Optimum Currency Areas, —
The American Economic Review, vol. 53, no. 4/1963)
 price and wage flexibility (B. Eichengreen, European Monetary
Unification, —Journal of Economic Literaturel,, vol. 31, no. 3/1993),
 low inflation rates differentials (G.Haberler, The International Monetary
System: Some Recent Developments and Discussions in: Approaches to
Greater Flexibility of Exchange Rates, ed.G. Halm, Princeton University
Press,1970; J. Fleming , On Exchange Rate Unification, in Economic
Journalli, vol. 81/1971.)
The latter condition has been adopted as the Maastricht convergence
criterion on price stability.

All accession candidates should aim to become full EMU members ASP.

Pre-EU



free float with inflation targeting
most credible fixed exchange rate regime. Unilateral euroisation inconsistent
with future EMU membership. ‘Consensual’ euroisation worth pursuing.
Post-EU but pre-EMU.

First-best: achieve inflation convergence and join EMU as soon as possible after
joining EMU. Could even be at same time as EU, if ERM membership (for at
least 2 years) is not required to satisfy normal ERM fluctuation margins.
Precedents: Italy, Finland, Greece.

Second-best: (unavoidable problem: purgatory of unrestricted capital mobility,
risk of speculative attacks, collapsing peg, excess volatility, misalignment).

most credible fixed exchange rate regime. Cannot be unilateral
euroisation. Could be currency board. Could be currency board with euro
as parallel currency. Might even be ‘consensual’ euroisation.
Problem with any pre-EMU fixed exchange rate regime : inflation criterion
meets Balassa-Samuelson. Could require unnecessary recession for 1 year.

Target zone with margins < ±15%.
Problem: risk of excessive volatility, speculative attacks, collapse of band,
misalignment.

Second-best: (unavoidable problem: purgatory of
unrestricted capital mobility, risk of speculative attacks,
collapsing peg, excess volatility, misalignment).

most credible fixed exchange rate regime. Cannot
be unilateral euroisation. Could be currency board.
Could be currency board with euro as parallel
currency. Problem with any pre-EMU fixed exchange
rate regime : inflation criterion meets BalassaSamuelson. Could require unnecessary recession for 1
year.

Target zone with margins < ±15%.
Problem: risk of excessive volatility, speculative
attacks, collapse of band, misalignment.
Figure 4: Impulse responses
Estonia - Response to One S.D. Innovations ± 2 S.E.
Response of GDPE to GDPE
Response of GDPE to GLNEE
Response of GDPE to GCPIE
0.08
0.08
0.08
0.06
0.06
0.06
0.04
0.04
0.04
0.02
0.02
0.02
0.00
0.00
0.00
-0.02
-0.02
-0.02
-0.04
-0.04
-0.04
-0.06
-0.06
-0.06
-0.08
1
-0.08
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
-0.08
1
2
3
4
5
6
7
8
7
8
Lithuania - Response to One S.D. Innovations ± 2 S.E.
Response of GDPL to GCPIL
Response of GDPL to GDPL
Response of GDPL to GLNEL
0.08
0.08
0.08
0.06
0.06
0.06
0.04
0.04
0.04
0.02
0.02
0.02
0.00
0.00
0.00
-0.02
-0.02
-0.02
-0.04
-0.04
-0.04
-0.06
-0.06
-0.06
-0.08
1
-0.08
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
-0.08
1
2
3
4
5
6
Bulgaria - Response to One S.D. Innovations ± 2 S.E.
R e s pons e of G D PB t o G D PB
R e s pons e of G D PB t o G C PIB
R e s pons e of G D PB t o G LN EB
0.4
0.4
0.4
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.0
0.0
0.0
-0.1
-0.1
-0.1
-0.2
-0.2
1
2
3
4
5
6
7
8
-0.2
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8


Show the reaction of the output to one standard deviation shock
over a time horizon of two years;
Bulgaria:
 higher increase in the level of output, although the confidence
band of the impulse response is wider over time; the impact of
the supply shock on the output tapers off,
 a price shock and shock in employment provide temporally
increase in the level of output followed by a stabilization within a
year;
Estonia and Lithuania:
 smaller increase in the level of output, as compared to
Bulgaria;
 the confidence band of the impulse response increases;
 the impact of the supply shock is not stable and takes long
time to adjust;
 in Estonia, a shock to employment has negative effect on the
output in the initial period and the positive impact comes after
three quarters
Figure 5: Variance decompositions
Lithuania - Variance Decomposition of GDPL
Estonia - Variance Decomposition of GDPE
100
100
80
80
60
60
40
40
20
20
0
0
1
2
3
GDPE
4
5
GCPIE
6
7
GLNEE
1
8
Bulgaria - Variance Decomposition of GDPB
100
80
60
40
20
0
1
2
3
GDPB
4
5
GCPIB
6
7
GLNEB
8
2
3
GDPL
4
5
GCPIL
6
7
GLNEL
8
Variance decomposition:
gives an information about the the contribution of the different
random shocks to the development of the variables in VAR
Results:
•
•
•
the variation in the output of the three CBA
countries is attributed mainly to the supply shocks –
about 80 percent for Estonia and Lithuania and 90
per cent for Bulgaria;
the price contribution in the output variance is
around 20 per cent in Estonia and Lithuania but it is
much less for Bulgaria;
the employment shocks do not contribute
significantly to the variance of the output.