Contagious Currency Crises

1
Contagious Currency Crises
The Academy of Economic Studies Doctoral School
of Banking and Finance

• - Dissertation Paper-

Student Dumitru Delia Supervisor Prof.
Moisã Altãr
Bucharest, July 2003
2
Objectives

• The Currency Crisis from Russia, august 1998
  testing for the existence of a contagion effect
• Determine whether the macroeconomic similarities
  between countries represented a channel of
  contagion
• Determine the domestic economic fundamentals that
  influenced the pressure on the exchange market.

3
Definitions

• A currency crisis is usually defined as a
  situation in which an attack on the currency
  leads to a sharp depreciation of the exchange
  rate.
• Testing for contagion means searching whether the
  probability of a crisis in a country at a point
  in time increases the probability of crises in
  other countries after controlling for the effect
  of political and economic fundamentals.

4
Litherature Review
Three generations of models referring to currency
crises
Contagious Currency Crises

• Krugmans Model (1979) - crises were caused by
  weak economic fundamentals
• Obstfelds Model (1986) - self-fulfilling crises
• Early Warning System Models
• -Kaminsky, Lizondo and Reinhart, 1998
• -Eichengreen, Rose and Wyplosz, 1996

• Gerlach and Smets (1995)- trade links
• Goldfajn and Valdes (1995) illiquidity
• Eichengreen, Rose and Wyplosz (1996)- trade and
  similarity links
• Sachs, Tornell and Velasco (1996)- contagion due
  to similar economic features.

5
The Data

• Countries Russia, Ukraine, Latvia, Lithuania,
  Estonia, Poland, Hungary, the Czech Republic, the
  Slovak Republic, Romania and Bulgaria
• Quarterly Data Q11993- Q12003
• Date Sources International Financial Statistics
  , IMF-World Bank-OECD-BIS joint table .

6
When did speculative attacks take place?

• Index of exchange market pressure
• where ei,t - the price of a USD in
  countrys i currency at time t
• ?ii,t - the variation of
  short term interest rate
• ?ri,t - the variation of
  international reserves
• a, ß, ? - weights.

7
When did speculative attacks take place?

• Extreme values of EMP
• 1, if EMPi,t1.5sEMPµEMP
• Crisisi,t
• 0, otherwise.
• Results

Quarter RUS UKR SLO POL LIT LAT HUN EST CZH BUL ROM
19983 1 0 0 0 0 0 0 0 0 0 0
19984 0 0 0 0 0 0 0 0 0 0 0
19991 0 0 1 1 0 0 0 1 1 0 0
19992 0 0 0 0 0 0 0 0 0 0 1
8
The Model

• Equation
• Fundamentals
• - domestic credit - current account
• - CPI growth - employment
• - GDP growth - unemployment
• - money - government
  deficit
• - ratio of short term debt to reserves
• - deviation of the real exchange rate from
  the trend.

9
The Model

• Determine the macroeconomic similarities whose
  existence might be a potential channel for
  contagion.
• Being similar means having similar
  macroeconomic conditions
• Similarity weights
• Variables domestic credit, money, CPI, output
  growth and current account.

10
The Czech Republic
EMP index

• Russia EMP- significant positive coefficient
• Current account similarity significance (1)
  domestic credit and money-no sign.
• Domestic influences
• - domestic credit()
• - ratio of short term debt to reserves()
• - percentage of current account in GDP(-)
• - economic growth(-).

variable Coefficient T-statistic Prob.
D(pctcrt(-1)) -1.135928 -5.946943 0.0000
D(domcred(-2)) 0.000823 4.943256 0.0000
DGDP(-2) -0.050498 -1.582290 0.1252
Emp1rus(-2) 0.081906 2.714749 0.0114

• R-squared 0.628581
• Adjusted R-squared 0.559800
• S.E. of regression 0.042134
• Schwarz criterion -3.060881
• Akaike info criterion -3.332973

11
Bulgaria

• The probability that Russia EMP might be
  significant is around 50
• Domestic fundamentals found significant
• - CPI inflation()
• - current account(-)
• - ratio of short term debt to reserves()
• - deviation of real exchange rate from trend().

EMP Index
variable Coefficient T-statistic Prob.
CPIL 0.515084 12.09939 0.0000
ctcrt -0.000188 -2.383844 0.0232
Devreer(-1) 0.582110 13.53560 0.0000
dtsrez 0.103150 1.625161 0.1139

• R-squared 0.836911
• Adjusted R-squared 0.816524
• S.E. of regression 0.192737
• Schwarz criterion -0.112205
• Akaike info criterion -0.329896

12
Estonia

• Russia EMP - significant positive
  coefficient(1)
• GDP similarity best results
• Significant influence
• - domestic credit()
• - percentage of current account in GDP(-)
• - CPI inflation().

EMP Index
variable Coefficient T-statistic Prob.
CPIL 0.008067 3.431687 0.0019
pctcrt -0.325318 -4.079743 0.0004
Emp1rus(-2) 0.120280 3.997495 0.0004
D(domcred(-1),2) 5.51E-06 1.562903 0.1297

• R-squared 0.443104
• Adjusted R-squared 0.339975
• S.E. of regression 0.040774
• Schwarz criterion -3.126489
• Akaike info criterion -3.398581

Breusch-Godfrey Serial Correlation LM
Test F-statistic 0.64710 Prob
0.532106 ObsR-squared 0.377274 Prob
0.828087
13
Latvia
Similarity weights

• No evidence of contagion(35)
• Significant influences
• Election()
• Current account()
• - CPI inflation().

variable Coefficient T-statistic Prob.
CPIL 0.004705 2.804809 0.0092
ctcrt 0.000127 6.204786 0.0000
elections 0.019899 2.306785 0.0290

• R-squared 0.576670
• Adjusted R-squared 0.513954
• S.E. of regression 0.017422
• Schwarz criterion -4.890526
• Akaike info criterion -5.119547
• Durbin Watson stat 2.082432

14
Lithuania
EMP Index

• No evidence of contagion
• High current account similarity
• Significant influence
• - domestic credit()
• - money()
• - deviation of real exchange rate from
  trend().

variable Coefficient T-statistic Prob.
D(domcred) 1.29E-05 2.711043 0.0112
D(money) 2.65E-05 1.973154 0.0581
devreer 0.004673 1.765932 0.0879

• R-squared 0.618261
• Adjusted R-squared 0.578770
• S.E. of regression 0.020152
• Schwarz criterion -4.676371
• Akaike info criterion -4.857766
• Durbin Watson stat 2.071606

15
Poland
EMP Index

• EMP Russia significant
• GDP similarity - best results
• Significant influences
• - government deficit(-)
• - domestic credit()
• - deviation of real exchange rate from
  trend().

variable Coefficient T-statistic Prob.
Defbug(-1) -2.18E-06 -2.728212 0.0130
D(domcred) 2.08E-0.6 2.478591 0.0222
devreer 0.557848 4.599356 0.0002
Emp1rus(-2) 0.044744 3.077666 0.0059

• R-squared 0.742046
• Adjusted R-squared 0.677558
• S.E. of regression 0.028311
• Schwarz criterion -3.801626
• F-statistic 11.50665
• Prob(F-statistic) 0.000025
• Akaike info criterion-4.091956

16
The Slovak Republic

• EMP Russia positive coefficient
• High current account similarity
• Influences
• - GDP growth(-)
• - money()
• - deviation of real exchange rate from
  trend()
• - domestic credit()
• - ratio of short term debt to reserves()

EMP Index
Variable Coefficient T-statistic Prob.
DGDP(-1) -0.078978 -4.732886 0.0001
D(money(-1)) 9.62E-06 5.888247 0.0000
dtsrez 0.096540 4.052545 0.0004
D(domcred(-1),2) 3.21E-07 4.636595 0.0001
devreer 0.041408 4.985542 0.0000
Emp1rus(-2) 0.027616 1.542939 0.1345

• R-squared 0.777728
• Adjusted R-squared 0.728334
• S.E. of regression 0.023180
• Schwarz criterion -4.195540
• Akaike info criterion-4.091956

17
Ukraine

• EMP Russia significant
• All similarity coefficients are high
• Significant influences
• - money()
• - current account(-).

EMP Indexes
Variable Coefficient T-statistic Prob.
D(ctcrt) -9.12E-05 -2. 880143 0.0114
Emp2rus 0.488519 8.394680 0.0000

• R-squared 0. 854556
• Adjusted R-squared 0.806074
• S.E. of regression 0.077818
• Schwarz criterion -1.735484
• Akaike info criterion-2.033919
• Durbin-Watson 1.783723

18
Hungary

• No evidence of contagion
• Significant influence
• - CPI inflation()
• - deviation of real exchange rate from
  trend()
• - domestic credit()
• - employment(-)
• - money()
• - current account(-).

EMP Index
Variable Coefficient T-statistic Prob.
D(dCPI) 0.517071 2.416113 0.0225
D(money,2) 9.37E-05 4.713845 0.0001
employment -2.60E-05 -8.202428 0.0000
D(domcred,2) 7.31E-05 6.201601 0.0000
devreer 0.006794 5.448621 0.0000
D(ctcrt) 1.99E-05 1.965977 0.0593

• R-squared 0.829776
• Adjusted R-squared 0.793300
• S.E. of regression 0.026885
• Schwarz criterion -3.906587
• Akaike info criterion-4.217656

19
Romania
EMP Index

• EMP Russia positive significant coefficient
• Domestic fundamentals
• - CPI inflation()
• - deviation of real exchange rate from trend()
• - ratio of short term debt to reserves()
• - Government deficit()

20
Romania

• Variable Coefficient Std. Error
  t-Statistic Prob.
• D(CPI,2) 0.000835 0.000405 2.062037
  0.0518
• C 1.029751 0.248285
  4.147457 0.0005
• D(DEF) -1.09E-05 3.83E-06
  -2.840305 0.0098
• DGDP -1.097253 0.249253
  -4.402171 0.0002
• D(DTSREZ) 0.778237 0.181645 4.284391
  0.0003
• D(DEVREER,2) 0.000529 0.000105
  5.034097 0.0001
• EMP1RUS(-3) 0.248825 0.052644 4.726602
  0.0001
• R-squared 0.907751 Mean
  dependent var -0.032190
• Adjusted R-squared 0.877002 S.D. dependent
  var 0.158816
• S.E. of regression 0.055698
  Akaike info criterion -2.708777
• Sum squared resid 0.065149 Schwarz
  criterion -2.331592
• Log likelihood 47.27727
  F-statistic 29.52075
• Durbin-Watson stat 1.843044 Prob(F-statistic)
  0.000000

Breusch-Godfrey Serial Correlation LM
Test F-statistic 0.166708
Probability 0.917425 ObsR-squared 0.000000
Probability 1.000000
21
Romania

• Bilateral trade weights twice the percentage of
  exports and once the percentage of imports with
  Russia
• The Wald test in this case

F-statistic 80.62561 Probability
0.000000 Chi-square 80.62561
Probability 0.000000
22
Conclusions

• A speculative attack in Russia seems to have
  increased significantly the odds of an attack in
  6 of the countries included in the sample - it
  does not represent a definitive proof of
  contagion
• The hypothesis that attacks spread to other
  countries where economic policies and conditions
  are similar is not always confirmed
  similarities are difficult to capture in a
  weighting scheme.
• The fundamental causes of speculative attacks
  differ across countries- it is very difficult to
  find a set of fundamentals underlying all crises.

23
References

• Abiad, A (2003), Early Warning Systems a Survey
  and a Regime Switching Approach, IMF Working
  Paper No.32/2003 ((Washington International
  Monetary Fund).
• Berger, W. and H. Wagner (2002), Spreading
  Currecncy Crises The Role of Economic
  Interdependence, IMF Working Paper No.02/144
  (Washington International Monetary Fund).
• Bussiere, M and M.Fratzcher (2002), Towards a
  New Early Warning System of Financial Crises,
  ECB Working Paper No. 145/2002 (European Central
  Bank).
• Bussiere, M. and C. Mulder (1999), External
  Vulnerability in Emerging market economies How
  High Liquidity can offset Weak Fundamentals and
  the Effects of Contagion, IMF Working Paper
  No.99/88 (Washington International Monetary
  Fund).
• Eichengreen, B., A.K.Rose and C.Wyplosz (1996),
  Contagious Currency Crises, NBER Working Paper
  No.5681 (Cambridge National Bureau of Economic
  Research).
• Frankel, J. and A.K.Rose (1996), Currency
  Crashes in Emerging Markets Empirical
  Indicators, NBER Working Paper No.5437/96
  (Cambridge National Bureau of Economic
  Research).
• Fratzcher, M. (2002), On Currency Crises and
  Contagion, ECB Working Paper No. 139/2002
  (European Central Bank).
• Ghosh, S. and A. Ghosh (2002), Structural
  Vulnerabilities and Currency Crises, IMF Working
  Paper No.02/9 (Washington International Monetary
  Fund).
• Kaminsky, G., S. Lizondo and C.Reinhart (1998),
  Leading Indicators of Currency Crises, Staff
  Papers, International Monetary Fund, Vol.45.
• Kaminsky, G. and C.Reinhart (1996), The Twin
  Crises The Causes of Banking and Balance of
  Payments Problems, International Finance
  Discussion Paper, (Washington Board of Governors
  of the Federal System).
• Kaminsky, G (1999), Currency and banking Crises
  The Early Warnings of Distress, IMF Working
  Paper No.99/178 (Washington International
  Monetary Fund).
• Mathieson, D, J. A.Chan-Lau and J.Y.Yoo, 2002,
  Extreme Contagion in Equity Markets, IMF
  Working Paper No.02/98 (Washington International
  Monetary Fund).