; THE IMPACT OF EMU ON REAL EXCHAN
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

THE IMPACT OF EMU ON REAL EXCHAN

VIEWS: 13 PAGES: 30

  • pg 1
									   EARLY WARNING
SYSTEMS FOR BANKING
       CRISES

Course on Financial Instability at the Estonian Central Bank,
             9-11 December 2009 – Lecture 7

                       E Philip Davis
                NIESR and Brunel University
                        West London
                  e_philip_davis@msn.com
                   www.ephilipdavis.com
          groups.yahoo.com/group/financial_stability
               Introduction
• 3 types of models for early warning, logit,
  signal extraction and binary recursive tree
• We apply the models first to prediction of
  crises in Asia
• And then outline a new logit approach
  which predicts banking crises in OECD
  countries
            Early warning systems
• Multivariate logit model uses macroeconomic,
  institutional and financial variables X as inputs to
  calculate probability of a banking crisis Y as the
  output via logistic function estimator. Suitable for
  answering question “what is the likelihood of a
  banking crisis occurring in the next t years?”
                                     e  'Xit
   Pr obYit  1  F X it  
                                   1  e  'Xit
• Non-parametric signal extraction approach
  tracks individual time series X prior to and
  during crisis episodes to answer question “is
  there a signal S of future crisis or not?” If an
  input variable’s aberrant behaviour can be
  quantitatively defined whenever that variable
  moves from tranquil to abnormal activity, a crisis
  is forewarned.
• { S ij = 1 } = { │ Xij │ > │ X*ij │ } or
• { S ij = 0 } = { │ Xij │ < │ X*ij │ }
• Binary Recursive Tree (BRT) can be used to
  answer question “which non-linear variable
  interactions make an economy more vulnerable to
  crisis than others?” Argued that liquidity, credit
  and market risks are all potentially non-linear.
  Estimator identifies single most important
  discriminator between crisis and non-crisis
  episodes across the entire sample, thereby creating
  two nodes. Nodes are further split into sub-nodes
  based on the behaviour of splitter variables’ non-
  linear interactions with previous splitter variables.
  This generates nodal crisis probabilities and the
  associated splitter threshold values.
  Figure 4: Schematic Diagram                     Entire Sample: 72
  of Binary Recursive Tree                        crises
  (BRT)
                                                  PARENT NODE




                                     X1≤ V1*                                      X1>V1*



                                                Splitter Variable: X1


                     Child Node 1:                                                         Child Node 2:
                     52 crises                                                             20 crises




             Splitter Variable: X2                                                     Splitter Variable: X3

X2≤ V2*
                                                  X2> V2*               X3≤ V3*                                X3≥ V3*

  Terminal Node 3:                   Terminal Node 3:              Terminal Node 4:                            Terminal Node 5:
  48 crises                          4 crises                      17 crises                                   3 crises
      Advantages and disadvantages
• Logistic models are ideally suited to predicting a binary
  outcome (1 = banking crisis, 0 = no banking crisis) using
  multiple explanatory variables selected on the basis of their
  theoretical or observed associations with banking crises.
• Logistic approach is also parametric, generating confidence
  intervals attached to coefficient values and their
  significance, but logit coefficients are not intuitive to
  interpret and they do not reflect the threshold effects that
  may be simultaneously exerted by other variables.
• Signal extraction non parametric and can use high
  frequency data
• Logit approach is the most appropriate for use as a global
  EWS, while signal extraction methods are more appropriate
  for a country-specific EWS (Davis and Karim 2008).
• BRT is able to discover non-linear variable interactions,
  making it especially applicable to large banking crises
  datasets where many cross-sections are necessary to
  generate enough banking crisis observations and
  numerous factors determine the occurrence of systemic
  failure.
• In BRT no specific statistical distribution needs be
  imposed on the explanatory variables. Also not
  necessary to assume all variables follow identical
  distributions or that each variable adopts the same
  distribution across cross-sections.
• Although logistic regression does not require variables
  to follow any specific distribution, Davis and Karim
  (2008) showed that standardising variables displaying
  heterogeneity across countries improved the predictive
  performance of logit models.
• Logistic regressions are also sensitive to outlier effects,
  yet it is precisely the non-linear threshold effects exerted
  by some variables that could generate anomalous values
  in the data.
• In low risk, stable regimes, variables may conform to a
  particular distribution which subsequently jumps to a
  regime of financial instability. Non-parametric BRTs
  should handle such data patterns better than logistic
  regressions.
• BRT is extremely intuitive to interpret. The model
  output is represented as a tree which is successively split
  at the threshold values of variables that are deemed as
  important contributors to banking crises.
• Signal extraction is also easier to interpret than logit, but
  is vulnerable to ignoring multivariate patterns at core of
  instability
Illustrative results – logit (Asia)
       Variable              Coefficient        z-Statistic   Coefficient    z-Statistic
     DCRED(-1)               -0.033902         -2.091298      -0.032416     -2.046609
     GDPPC(-1)               -0.000246         -3.451172      -0.000235     -3.535303
     FISCY(-1)                0.010451          0.153806
      INFL(-1)               -0.037791         -1.212934
       RIR(-1)                0.114829          2.528462      0.113567      2.612414
    DEPREC(-1)                0.053493          2.724725      0.044323      2.712526
    DCREDY(-1)                0.022844          2.971898      0.021231      2.959820
      DTT(-1)                 0.007492          0.322193
     DGDP(-1)                -0.261366         -3.853235      -0.276748     -4.192324
     M2RES(-1)               -0.000549         -2.232728      -0.000536     -2.190088
Expectation-Prediction Evaluation for Binary
Specification
Equation: IND_STAND
Date: 12/03/09 Time: 19:46
Success cutoff: C = 0.25

                      Estimated Equation
                    Dep=0     Dep=1      Total

P(Dep=1)<=C             80          7         87
 P(Dep=1)>C             34         41         75
    Total              114         48        162
   Correct              80         41        121
  % Correct          70.18      85.42      74.69
 % Incorrect         29.82      14.58      25.31
 Total Gain*         70.18     -14.58      45.06
Percent Gain**       70.18        NA       64.04
                   Signal extraction - Asia
        2
       1.8
       1.6
       1.4
       1.2
NTSR




        1
       0.8
       0.6
       0.4
       0.2
        0
             0.5   1          2          3.5        4        4.5         6         8         10    20
                                               Percentile Threshold


                    GDP grow th                  Change Terms of Trade       Depreciation
                    Real Interest Rate           Inflation                   Fiscal Surplus/ GDP
                    M2/Reserves                  GDP per Capita

                    Real GDP growth Fiscal surplus/GDP Depreciation
 % crises correct         10                 8              6
% no crises correct       99                 98            98
 % total correct          65                 64            62
                               BRT - Asia
                                            Node 1
                                       Class Cases %
                                         0    120 71.4
                                         1     48 28.6



                                                                FISCY > -1.14
                  FISCY <= -1.14
                                                                    Terminal
                     Node 2
                                                                    Node 4
                Class Cases %
                                                               Class Cases %
                  0     41 51.3
                                                                 0     79 89.8
                  1     39 48.8
                                                                 1      9 10.2


 DGDP <= 4.75
                                    DGDP >   4.75
     Terminal
                                        Node 3
      Node 1
                                   Class Cases %
Class Cases %
                                     0    34 73.9
  0      7 20.6                                                                    Asia
                                     1    12 26.1
  1     27 79.4
                                                                % crises correct    46
                                                               % no crises correct 90
                     DCREDY <= 60.49         DCREDY > 60.49     % total correct     84
                          Terminal                Terminal
                           Node 2                 Node 3
                     Class Cases %           Class Cases %
                       0     18 100.0          0     16 57.1
                       1      0    0.0         1     12 42.9
Leading indicator selection
                                         Asia
                                       Signal
                               Logit   Extraction   Tree
Real GDP Growth                                   
Real Interest Rate
Inflation
Fiscal Surplus/ GDP                                 
M2/ Foreign Exchange
Reserves                          
Real Domestic Credit Growth       
Real GDP per capita               
Domestic credit/GDP                                 
Depreciation                            
Terms of Trade                          
Current account/GDP
External short term debt/GDP
    A new model for the OECD
• Existing work on early warning systems (EWS)
  for banking crises generally omits bank capital,
  bank liquidity and property prices, despite their
  relevance to the probability of crisis in the mind of
  bankers, policymakers and the public. One reason
  for this neglect is that most work on EWS to date
  has been for heterogeneous global samples
  dominated by emerging market crises. For such
  countries, time series data on bank capital
  adequacy and property prices are typically absent,
  while other variables affecting crises may also
  differ in OECD countries.
• We argue results are misspecified
• Triggers of crisis depend on the type of economy and
  banking system. In OECD countries with high levels of
  banking intermediation and developed financial markets,
  shocks to terms of trade are less important crisis triggers
  than, say, property price bubbles.
• Also developed economy banking systems are more
  likely to be regulated in terms of capital adequacy and
  liquidity ratios
• Accordingly, we estimate logit models of crisis for
  OECD countries only and find strong effects of capital
  adequacy, liquidity ratios and property prices, such as to
  exclude most traditional variables. Our results imply that
  higher unweighted capital adequacy as well as liquidity
  ratios has a marked effect on the probability of a banking
  crisis, implying long run benefits to offset some of the
  costs that such regulations may impose (e.g. widening of
  bank spreads).
             Methodology and data
• Multivariate logit with dependent variable being crisis
  probability
• Problems of crisis dummies
   – Definition of banking crises
   – Start and end dates ambiguous
   – Focus on switch date in core results
• Data partitioned to 1980-2006 and 2007 to leave subprime
  crisis for out-of-sample
• Variables for bank regulation:
   – Unweighted capital adequacy ratio - ratio of capital and reserves
     for all banks to the end of year total assets
   – Liquidity - ratio of the sum of cash and balances with central
     banks and securities for all banks over the end of year total assets
            Table of crises in sample
       BG   CN   DK   FN   FR   GE   IT   JP   NL   NW   SP   SD   UK   US

1980   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1981   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1982   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1983   0    1    0    0    0    0    0    0    0    0    0    0    0    0
1984   0    0    0    0    0    0    0    0    0    0    0    0    1    0
1985   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1986   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1987   0    0    1    0    0    0    0    0    0    0    0    0    0    0
1988   0    0    0    0    0    0    0    0    0    0    0    0    0    1
1989   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1990   0    0    0    0    0    0    1    0    0    1    0    0    0    0
1991   0    0    0    1    0    0    0    1    0    0    0    1    1    0
1992   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1993   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1994   0    0    0    0    1    0    0    0    0    0    0    0    0    0
1995   0    0    0    0    0    0    0    0    0    0    0    0    1    0
1996   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1997   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1998   0    0    0    0    0    0    0    0    0    0    0    0    0    0
1999   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2000   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2001   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2002   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2003   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2004   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2005   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2006   0    0    0    0    0    0    0    0    0    0    0    0    0    0
2007   0    0    0    0    0    0    0    0    0    0    0    0    1    1
   Box 1: List of Variables (with variable key)

                              1. Real GDP Growth (%) (YG)
      Variables used in
                              2. Real Interest Rate (%) (RIR)
      previous studies:
                              3. Inflation (%) (INFL)
     Demirguc-Kunt and
                              4. Fiscal Surplus/ GDP (%) (BB)
     Detragiache (2005);
   Davis and Karim (2008).    5. M2/ Foreign Exchange Reserves (%) (M2RES)
                              6. Real Domestic Credit Growth (%) (DCG)
                              7. Liquidity ratio (%) (LIQ)
    Variables introduced in
                              8. Unweighted capital adequacy ratio (%) (LEV)
          this study.
                              9. Real Property Price Growth (%) (RHPG)


                                            e  'Xit
Pr obYit  1  F X it                        'Xit
                                         1 e
               n    T
Log e L   Yit log e F  ' X it   1  Yit  log e 1  F  ' X it 
              i 1 t 1
Table 2: The General To Specific Approach

                  -0.118    -0.124 -0.137      -0.135    -0.135    -0.144    -0.147
    LIQ(-1)
                  (-3.55)   (-3.55) (-3.64)    (-3.55)   (-3.45)   (-3.39)   (-3.25)
                  -0.333    -0.239 -0.315      -0.247    -0.271    -0.280    -0.273
    LEV(-1)
                  (-2.85)   (-1.90) (-2.24)    (-1.64)   (-1.67)   (-1.72)   (-1.62)
                   0.113     0.113   0.104      0.100     0.104     0.108     0.110
    RHPG(-3)
                   (2.8)    (2.87) (2.67)       (2.59)    (2.67)   (2.76)     (2.67)
                            -0.099 -0.10         -0.10    -0.10     -0.13     -0.13
    DCG(-1)          -
                            (-1.82) (-1.97)    (-1.86)   (-1.99)   (-1.98)   (-1.98)
                                     0.084      0.085     0.165     0.173     0.166
    RIR(-1)          -          -
                                     (1.37)     (1.40)    (1.41)   (1.46)     (1.30)
                                                 -0.00    -0.00     -0.00     -0.00
    M2RES(-1)        -         -        -
                                                (-1.0)    (-1.0)    (-1.1)    (-1.1)
                                                          -0.13     -0.14     -0.13
    INFL(-1)         -         -        -         -
                                                          (-0.8)    (-0.8)    (-0.7)
                                                                    0.116     0.125
    YG(-1)           -         -        -         -         -
                                                                   (0.65)     (0.66)
                                                                             -0.013
    BB(-1)           -         -        -         -         -         -
                                                                              (-0.1)
Note: estimation period 1980-2006; t-statistics in parentheses; LIQ-liquidity ratio, LEV- unweighted capital
adequacy ratio, YG-real GDP growth, RPHG-real house price inflation, BB-budget balance to GDP ratio,
DCG-domestic credit growth, M2RES-M2 to reserves ratio, RIR-real interest rates, DEP-depreciation, INFL-
inflation.
Table 3: Comparing the Effects of Sample Period on Estimation Results
            Estimation period
          1980-2006 1980-2007


            -0.118      -0.13
   LIQ
            (-3.55)     (-4.1)
            -0.333     -0.261
   LEV
            (-2.85)    (-2.51)
            0.113       0.106
   PHG
             (2.8)      (2.79)




     p(crisis) 
log 
      1 - p(crisis) 
                      = - 0.333 LEV(-1) – 0.118 LIQ(-1) + 0.113 RHPG(-3)
                   

                      (-2.85)         (-3.55)          (2.8)
Marginal effect of 1% rise in
variable on crisis probability
        LIQ     LEV    RHPG
BG     -0.17   -0.49    0.17
CN     -0.22   -0.61    0.21
DK     -0.05   -0.14    0.05
FN     -0.23   -0.65    0.22
FR     -0.78   -2.17    0.74
GE     -0.23   -0.65    0.22
IT     -0.17   -0.46    0.16
JP     -0.38   -1.05    0.36
NL     -0.56   -1.57    0.53
NW     -0.33   -0.91    0.31
SD     -0.12   -0.34    0.12
SP     -0.08   -0.24    0.08
UK     -1.19   -3.32    1.13
US     -0.08   -0.22    0.07
              BG
                   -




                          0.00
                                 0.20
                                        0.40
                                               0.60
                                                      0.80
                                                             1.00
              BG 80
                   -
              BG 90
                   -
              C 00
                 N
                   -
              C 83
                 N
                   -
              C 93
                 N
                   -
              D 03
                 K
                   -
              D 86
                 K
                   -
              D 96
                 K
                   -
               FN 06
                   -
               FN 8 9
                   -
               FR 9 9
                   -
               FR 8 2
                   -
               FR 9 2
                   -
              G 02
                 E
                   -
              G 85
                 E
                   -
              G 95
                 E
                   -0
                IT 5
                   -8
                IT 8
                   -




Probability
               J P 98
                   -
               JP 81
                   -
               JP 91
                   -
               N 01
                 L
                   -




Crisis
               N 84
                 L
                   -
               N 94
                 L
              N - 04
                W
              N - 87
                W
                   -
              SD 97
                   -
              SD 80
                   -
                                                                    Crisis probabilities




              SD 90
                   -
               SP 00
                   -
               SP 83
                   -
               SP 93
                   -
              U 03
                 K
                   -
              U 86
                 K
                   -
              U 96
                 K
                   -
              U 06
                 S
                   -
              U 89
                 S
                   -9
                      9
                In sample prediction
                         Aftermath
        Total                        False
                Crises    of the
        Calls                        Calls
                          Crises             Timing of False Calls relative to Crisis Onset
BG       0        0         0         0
CN       6        1         1         4      next year
DK       0        0         0         0
FN       10       1         1         8      next year
FR       14       1         0         13
GE       4        0         0         4
IT       7        0         2         5      2nd and 3rd years
JP       15       1         6         8      Next 7 years, with a break on the 4th year
NL       18       0         0         18
NW       14       1         2         11     next 2 years
SD       6        1         1         4      next year
SP       2        0         0         2
UK       20       2         0         18
US       0        0         0         0
total   116       8         13        95
Out of sample predictions
      2007   2008   definition1 definition2
BG     X      X          X           X
CN      -      -                     -
DK      -      -
FN      -     X
FR     X      X         X           X
GE      -      -        -           -
IT     X       -                    X
JP      -      -
NL     X       -        X           X
NW     X      X
SD      -      -                    -
SP     X      X                     X
UK     X      X         X           X
US      -      -        -           -
             Country elimination tests
                                                 US and   Norway       Finland   Sweden
          Final      UK not   US not Japan not
                                               Japan not    not          not        not
          panel     included included included
                                                included included     included   included

          -0.118    -0.143    -0.125    -0.111    -0.119    -0.124    -0.121     -0.115
LIQ(-1)
          (-3.55)   (-2.99)   (-3.55)   (-3.28)   (-3.29)   (-3.59)    (-3.5)    (-3.41)
          -0.333      -0.3    -0.339    -0.344    -0.349    -0.282    -0.293     -0.343
LEV(-1)
          (-2.85)   (-1.78)   (-2.79)   (-2.94)   (-2.86)   (-2.38)   (-2.43)    (-2.87)
           0.113     0.152     0.119     0.111     0.118     0.089     0.083      0.107
PHG(-3)
           (2.8)     (3.44)    (2.82)    (2.74)    (2.76)    (2.04)    (1.84)     (2.58)
Alternative crisis dates
                    Japanese
           Final               US crisis
                     crisis at
          version              at 1984
                      1992

          -0.118     -0.119       -0.12
LIQ(-1)
          (-3.55)    (-3.56)    (-3.58)
          -0.333     -0.332     -0.317
LEV(-1)
          (-2.85)    (-2.85)    (-2.73)
           0.113      0.113      0.104
PHG(-3)
           (2.8)      (2.8)      (2.56)
      Aftermath elimination and
           subprime runup

                    Aftermath                Final
                                                      1980-2007
           Final                                      estimation
                      of the                version
          version                                     with break
                      Crisis
                                            -0.118     -0.128
                                 LIQ(-1)
          -0.118     -0.111                 (-3.55)     (-3.4)
LIQ(-1)                                     -0.333     -0.241
          (-3.55)    (-3.48)     LEV(-1)
                                            (-2.85)    (-1.94)
          -0.333     -0.329                  0.113      0.106
LEV(-1)                         RHPG(-3)
          (-2.85)    (-2.91)                 (2.8)      (2.85)
           0.113      0.111                            -0.029
PHG(-3)                          LIQ(-1)b      -
           (2.8)      (2.74)                           (-0.34)
                                                       -0.045
                                LEV(-1)b       -
                                                       (-0.19)
                                                        0.006
                                RHPG(-3)b      -
                                                        (0.05)
   Further lags and systemic crises


LIQ (-2)   -0.104    LIQ (-1)   -0.121
           (-3.27)              (-2.49)
LEV (-2)   -0.385    LEV (-1)   -0.768
           (-3.22)              (-3.59)
PHG (-3)    0.119    PHG (-3)    0.235
            (3.00)               (3.71)
                   Conclusions
• 3 approaches complementary
• Traditional approaches fruitful for EMEs such as Asia but
  not for OECD countries
• Found relevance of bank capital, liquidity and property
  prices absent from traditional EWS, exclude traditional
  variables
• Can predict crises out of sample and specification is robust
• Warrants policy focus on bank regulation – of
  capital, liquidity but also of terms of mortgages loans
• Also supports measures to reduce procyclicality, adjusting
  capital or provisions countercyclically – and use of simple
  leverage ratio as well as risk weighted capital adequacy
                   References
• Davis, E P and D Karim (2008a), "Comparing early
  warning systems for banking crises", Journal of Financial
  Stability, 4, 89-120
• Davis E P and Karim D (2008b), "Could early warnings
  systems have helped to predict the subprime crisis?",
  National Institute Economic Review, 206, 25-37 and
  Brunel University Economics and Finance Working
  Paper No 08-27
• Barrell R, Davis E P, Karim D and Liadze I (2009),
  "Bank Regulation, Property Prices And Early Warning
  Systems For Banking Crises In OECD Countries",
  NIESR Discussion Paper No. 330

								
To top
;