The Academy of Economic Studies
Doctoral School of Finance and Banking
Integration of Center and
Eastern European Stock
Markets
MSc student IOSIF ANAIDA
Coordinator Professor Moisă Altăr
Bucharest, July 2007
Dissertation paper outline
 The integration of the emerging stock markets
 The aims of the paper
 Empirical studies concerning stock markets integration
 The Data
 Testing the cointegration
 Testing the correlation
 Conclusions
 References
Different approaches
 Bekaert and Harvey (1997) – market liberalization increase the
correlation between local market returns and the world market but do
not drive up local volatility .
 Forbs and Rigobon (1999) – there was no contagion during the euroAsia crises in 1997, the Mexican peso collapse in 1994 and in the US
stock market crash in 1987. High market co-movements during these
periods where a continuation of strong cross-market linkages
 Egert and Kocenda (2005) – there are no robust cointegration
relationship between emerging and developed markets. But there are
short-term spillover effects in terms of stock returns and stock price
volatility.
 Cappielo, G’erard, Kadareja and Manganelli (2006) – larger new EU
member state exhibit a strong comovements between themselves and
with the euro area. Form the smaller countries only Estonia and Cyprus
show integration bough with the euro zone and the block of large
economies.
The Data
10000
 Initial data series:
Bucharest Exchange
Trading index (BET),
1800
1600
8000
1400
1200
6000
1000
4000
800
600
2000
Prague Exchange index
(PX),
Warsaw Exchange index
(WIG20),
400
0
200
2001
2002
2003
2004
2005
2006
2001
2002
2003
2004
2005
2006
2005
2006
2005
2006
PX
BET
3600
1600
3200
1400
1200
2800
1000
Bulgarian Exchange index
(SOFIX),
Budapest Stock Index
(BUX)
Austrian Traded Index
(ATX).
2400
800
2000
600
1600
400
1200
200
0
800
2001
2002
2003
2004
2005
2001
2006
2002
2003
2004
SOFIX
W IG
28000
5000
24000
4000
20000
3000
 Time length: 20.10.2000
– 04.03.2007
16000
2000
12000
1000
8000
4000
0
2001
2002
2003
2004
BUX
2005
2006
2001
2002
2003
2004
ATX
The Cointegration analyses
 Verify stationarity of series using ADT, Phillips-Perron and KPSS tests:
the series are not stationary in level but is stationary in first difference
 Check the cointegration relationship between the variables using the
Engle Granger residual based cointegration method:
X t  1  i 1iYi ,t   t
n
 Estimate residuals series for each
regression.
 Verify the stationarity of the residual
series using ADF and PP tests.
 Comparing the test statistic with the
critical values estimated by Engle and
Yoo.
Engle-Granger cointegration test
Variables
ADF
PP
ATX
-3.420593*
-3.509189*
BET
-3.648032*
-3.380750*
BUX
-4.589758*
-4.270524*
PX
-5.088.152
-4.560678*
WIG
-3.674414*
-3.486132*
SOFIX
-2.972781*
-3.215374*
Cointegration analysis
 Johansen method in a VAR framework
 Select numbers of lags to include in the VAR using the Akaike informational
criterion
 Check VAR stability
yt  yt k  1yt 1  2 yt2  ...  k 1yt ( k 1)   t
Cointegration - conclusions
 The residuals are not stationary,
the value of t statistic is higher then
the critical value
 The Johansen method – the test
statistic is smaller then the critical
values
 There is no cointegration
relationship between the series.
Johansen cointegration test-eigenvalue trace
No.of
Eigenvalue Test statistic 5% critical 1% critical
cointegration
value
value
None
0.020503
79.57171
94.15
103.18
At most 1
0.011279
49.57410
68.52
76.07
At most 2
0.010257
33.14909
47.21
54.46
At most 3
0.007897
18.22059
29.68
35.65
At most 4
0.004609
6.740067
15.41
20.04
At most 5
3.56E-05
0.051499
3.76
6.65
Johansen cointegration test-eigenvalue max
No. of
Eigenvalue Test statistic 5% critical 1% critical
cointegration
value
value
None
0.020503
29.99762
39.37
45.10
At most 1
0.011279
16.42501
33.46
38.77
At most 2
0.010257
14.92850
27.07
32.24
At most 3
0.007897
11.48052
20.97
25.52
At most 4
0.004609
6.688568
14.07
18.63
At most 5
3.56E-05
0.051499
3.76
6.65
The correlation analysis for the returns
 Calculating the returns: dl_index
Correlation matrix for returns
ATX
BET
BUX
PX
SOFIX
WIG
ATX
1.000000
0.037999
0.445384
0.487496
-0.014321
0.375169
BET
0.037999
1.000000
0.016596
-0.021922
0.016283
0.017652
BUX
0.445384
0.016596
1.000000
0.545900
0.027081
0.536316
PX
0.487496
-0.021922
0.545900
1.000000
0.044288
0.497050
SOFIX
-0.014321
0.016283
0.027081
0.044288
1.000000
-0.003174
WIG
0.375169
0.017652
0.536316
0.497050
-0.003174
1.000000
The correlation analysis
 Verify the short-term interaction between returns using Granger causality test:
Yt  1  i 1i Yt 1  i 1 i X t 1   t
k
k
X t  1  i 1i X t 1  i 1 i Yt 1   t
k
Dependent
variable
ATX
WIG
BUX
PX
BET
SOFIX
k
ATX
0.5379
0.0954
0.6382
0.6461
0.3594
Granger causality test for returns
Lags of variable
WIG
BUX
PX
BET
0.8864
0.7603
0.0710 0.9771
0.3632
0.7832 0.7919
0.4349
0.7776 0.8995
0.5534
0.5316
0.1943
0.0311
0.8891
0.2188
0.2729
0.1126
0.8735 0.4195
SOFIX
0.3061
0.8956
0.8116
0.3178
0.5890
-
 Lag length criteria suggests a specification including 1 lag
 Choosing the order of the variables using the F-test, market capitalization
and the efficiency of the market: ATX, WIG, PX, BUX, BET, SOFIX.
 Verify the sign and proportion of the spillover between the returns using
the impulse response and variance decomposition.
The response of returns to shocks applied on the other markets
Response to Cholesky One S.D. Innovations ± 2 S.E.
Response of D_ATX to D_ATX
Response of D_ATX to D_WIG
Response of D_ATX to D_BUX
Response of D_ATX to D_PX
Response of D_ATX to D_BET
Response of D_ATX to D_SOFIX
.012
.012
.012
.012
.012
.012
.010
.010
.010
.010
.010
.010
.008
.008
.008
.008
.008
.008
.006
.006
.006
.006
.006
.006
.004
.004
.004
.004
.004
.004
.002
.002
.002
.002
.002
.000
.000
.000
.000
.000
.000
-.002
-.002
-.002
-.002
-.002
-.002
1
2
3
1
Response of D_WIG to D_ATX
2
3
1
Response of D_WIG to D_WIG
2
3
1
Response of D_WIG to D_BUX
2
3
.002
1
Response of D_WIG to D_PX
2
3
Response of D_WIG to D_BET
1
.016
.016
.016
.016
.016
.016
.012
.012
.012
.012
.012
.012
.008
.008
.008
.008
.008
.008
.004
.004
.004
.004
.004
.004
.000
.000
.000
.000
.000
-.004
-.004
1
2
3
-.004
1
Response of D_BUX to D_ATX
2
3
-.004
1
Response of D_BUX to D_WIG
2
3
Response of D_BUX to D_BUX
2
3
-.004
1
Response of D_BUX to D_PX
2
3
Response of D_BUX to D_BET
1
.012
.012
.012
.012
.012
.008
.008
.008
.008
.008
.008
.004
.004
.004
.004
.004
.004
.000
1
2
3
.000
1
Response of D_PX to D_ATX
2
3
.000
1
Response of D_PX to D_WIG
2
3
.000
1
Response of D_PX to D_BUX
2
3
2
3
1
Response of D_PX to D_BET
.012
.012
.012
.012
.012
.010
.010
.010
.010
.010
.010
.008
.008
.008
.008
.008
.008
.006
.006
.006
.006
.006
.006
.004
.004
.004
.004
.004
.004
.002
.002
.002
.002
.002
.000
.000
.000
.000
.000
.000
-.002
-.002
-.002
-.002
-.002
-.002
2
3
1
Response of D_BET to D_ATX
2
3
1
Response of D_BET to D_WIG
2
3
1
Response of D_BET to D_BUX
2
3
2
3
Response of D_BET to D_BET
1
.016
.016
.016
.016
.016
.012
.012
.012
.012
.012
.012
.008
.008
.008
.008
.008
.008
.004
.004
.004
.004
.004
.004
.000
.000
.000
.000
.000
-.004
1
2
3
Response of D_SOFIX to D_ATX
-.004
1
2
3
Response of D_SOFIX to D_WIG
-.004
1
2
3
Response of D_SOFIX to D_BUX
2
3
Response of D_SOFIX to D_PX
-.004
1
2
3
Response of D_SOFIX to D_BET
1
.024
.024
.024
.024
.024
.020
.020
.020
.020
.020
.020
.016
.016
.016
.016
.016
.016
.012
.012
.012
.012
.012
.012
.008
.008
.008
.008
.008
.008
.004
.004
.004
.004
.004
.000
.000
.000
.000
.000
.000
-.004
-.004
-.004
-.004
-.004
-.004
2
3
1
2
3
1
2
3
1
2
3
2
3
Response of D_SOFIX to D_SOFIX
.024
1
3
.000
-.004
1
2
Response of D_BET to D_SOFIX
.016
-.004
3
.002
1
Response of D_BET to D_PX
2
Response of D_PX to D_SOFIX
.012
1
3
.000
1
Response of D_PX to D_PX
2
Response of D_BUX to D_SOFIX
.012
.000
3
.000
-.004
1
2
Response of D_WIG to D_SOFIX
.004
1
2
3
1
2
3
Variance decomposition for the returns
 The initial shock in the returns
works through the system in about 3
days
 None of the emergent markets
influence the Austrian returns, but
changes in returns on the three
larger emergent stock markets are
due to changes in the Austrian
returns
The three larger emergent
markets: Poland, Czech Republic
and Hungary are correlated between
themselves in terms of returns
 BET and SOFIX returns seem
uninfluenced by the movements of
the other returns.
Days ahead
1
2
3
Days ahead
1
2
3
Days ahead
1
2
3
Days ahead
1
2
3
Days ahead
1
2
3
Days ahead
1
2
3
Variance decomposition for ATX returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
100.0000 0.000000 0.000000 0.000000 0.000000
99.59105 0.044013 0.286347 0.007451 3.33E-06
99.58768 0.044011 0.287208 0.007801 0.000799
Variance decomposition for WIG returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
14.17161 85.82839 0.000000 0.000000 0.000000
14.10401 85.81295 0.021019 0.056347 0.004508
14.10415 85.81132 0.021926 0.056702 0.004724
Variance decomposition for PX returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
23.69439 11.67964 64.62597 0.000000 0.000000
23.59788 11.63133 64.56473 0.023212 0.115084
23.59706 11.63074 64.56074 0.023219 0.120464
Variance decomposition for BUX returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
19.80737 15.93645 5.813288 58.44290 0.000000
20.07214 15.91336 5.791510 58.21816 0.000972
20.07180 15.91375 5.792217 58.21720 0.001113
Variance decomposition for BET returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
0.149843 0.007663 0.243116 0.016866 99.58251
0.156055 0.222483 0.295727 0.019366 99.28715
0.155925 0.237918 0.299797 0.019994 99.26612
Variance decomposition for SOFIX returns
D_ATX
D_WIG
D_PX
D_BUX
D_BET
0.020217 0.001275 0.311388 0.063885 0.043412
0.042319 0.011096 0.330570 0.218532 0.092179
0.044023 0.011096 0.330957 0.219286 0.092734
D_SOFIX
0.000000
0.071139
0.072503
D_SOFIX
0.000000
0.001168
0.001177
D_SOFIX
0.000000
0.067762
0.067778
D_SOFIX
0.000000
0.003862
0.003918
D_SOFIX
0.000000
0.019224
0.020247
D_SOFIX
99.55982
99.30530
99.30190
Methods in obtaining returns volatility
 Obtained variances series for returns using GARCH(1,1) method
- the mean equation:
y t    y t 1  t
- the conditional variance equation:
 t2     t21   t21
 Using a EGARCH(1,1,1) method to estimate variance for SOFIX returns
 Conditional variance equation for the EGARCH:
ln(  t2 )     ln(  t21 )  
  t 1
 t 1
2
 2 

2


 t 1
 t 1

 Advantages in using a EGARCH method:
- the coefficients can be negative because ln(  t2 ) is modeled.
- the asymmetry of the EGARCH model capture the leverage effect.
Volatility - the correlation analyze
 Verify stationarity of the variance using the ADF and PP tests, the series
are stationary at any significance level.
 Check the relation between the series using the matrix correlation and Granger
Causality test.
Correlation matrix for volatiles
ATX
BET
BUX
PX
SOFIX
ATX
BET
BUX
PX
SOFIX
WIG
Dependent
variable
ATX
BET
BUX
SOFIX
PX
WIG
WIG
1.000000 -0.075787 0.461381 0.522018 -0.046720 0.179787
-0.075787 1.000000 0.062745 -0.035368 -0.033744 -0.166919
0.461381 0.062745 1.000000 0.675333 0.038740 0.332519
0.522018 -0.035368 0.675333 1.000000 0.106245 0.347898
-0.046720 -0.033744 0.038740 0.106245 1.000000 0.366838
0.179787 -0.166919 0.332519 0.347898 0.366838 1.000000
Granger causality test for variance
Lags of variable
ATX
BET
BUX
SOFIX
PX
0.5589
0.6707
-
WIG
0.4989
0.1782
0.8455
0.8701
0.0259
0.3360
0.1367
0.0946
0.0401
0.9509
0.3882
0.0285
0.3143
0.7626
0.6825
0.9919
0.3906
0.2474
-
0.1647
0.9934
0.1262
0.1319
0.7174
0.9941
0.2192
0.3314
0.4108
0.8961
-
Impulse response for volatility
Response to Cholesky One S.D. Innovations ± 2 S.E.
Response of D_ATXVARIANCE to D_ATXVARIANCE
Response of D_ATXVARIANCE to D_WIGVARIANCE
Response of D_ATXVARIANCE to D_BUXVARIANCE
Response of D_ATXVARIANCE to D_PXVARIANCE
Response of D_ATXVARIANCE to D_BETVARIANCE
Response of D_ATXVARIANCE to D_SOFIXVARIANCE
.00006
.00006
.00006
.00006
.00006
.00006
.00005
.00005
.00005
.00005
.00005
.00005
.00004
.00004
.00004
.00004
.00004
.00004
.00003
.00003
.00003
.00003
.00003
.00003
.00002
.00002
.00002
.00002
.00002
.00002
.00001
.00001
.00001
.00001
.00001
.00001
.00000
.00000
.00000
.00000
.00000
.00000
-.00001
-.00001
-.00001
-.00001
-.00001
-.00001
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Response of D_WIGVARIANCE to D_ATXVARIANCE
Response of D_WIGVARIANCE to D_WIGVARIANCE
Response of D_WIGVARIANCE to D_BUXVARIANCE
Response of D_WIGVARIANCE to D_PXVARIANCE
Response of D_WIGVARIANCE to D_BETVARIANCE
Response of D_WIGVARIANCE to D_SOFIXVARIANCE
.000016
.000016
.000016
.000016
.000016
.000016
.000012
.000012
.000012
.000012
.000012
.000012
.000008
.000008
.000008
.000008
.000008
.000008
.000004
.000004
.000004
.000004
.000004
.000004
.000000
.000000
.000000
.000000
.000000
.000000
-.000004
-.000004
-.000004
-.000004
-.000004
-.000004
-.000008
-.000008
25
50
75
100
125
150
175
200
-.000008
25
50
75
100
125
150
175
200
-.000008
25
50
75
100
125
150
175
200
-.000008
25
50
75
100
125
150
175
200
-.000008
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Response of D_BUXVARIANCE to D_ATXVARIANCE
Response of D_BUXVARIANCE to D_WIGVARIANCE
Response of D_BUXVARIANCE to D_BUXVARIANCE
Response of D_BUXVARIANCE to D_PXVARIANCE
Response of D_BUXVARIANCE to D_BETVARIANCE
Response of D_BUXVARIANCE to D_SOFIXVARIANCE
.000025
.000025
.000025
.000025
.000025
.000025
.000020
.000020
.000020
.000020
.000020
.000020
.000015
.000015
.000015
.000015
.000015
.000015
.000010
.000010
.000010
.000010
.000010
.000010
.000005
.000005
.000005
.000005
.000005
.000005
.000000
.000000
.000000
.000000
.000000
-.000005
-.000005
25
50
75
100
125
150
175
200
-.000005
25
50
75
100
125
150
175
200
-.000005
25
50
75
100
125
150
175
200
.000000
-.000005
25
50
75
100
125
150
175
200
-.000005
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Response of D_PXVARIANCE to D_ATXVARIANCE
Response of D_PXVARIANCE to D_WIGVARIANCE
Response of D_PXVARIANCE to D_BUXVARIANCE
Response of D_PXVARIANCE to D_PXVARIANCE
Response of D_PXVARIANCE to D_BETVARIANCE
Response of D_PXVARIANCE to D_SOFIXVARIANCE
.00005
.00005
.00005
.00005
.00005
.00005
.00004
.00004
.00004
.00004
.00004
.00004
.00003
.00003
.00003
.00003
.00003
.00003
.00002
.00002
.00002
.00002
.00002
.00002
.00001
.00001
.00001
.00001
.00001
.00001
.00000
.00000
.00000
.00000
.00000
-.00001
-.00001
25
50
75
100
125
150
175
200
-.00001
25
50
75
100
125
150
175
200
-.00001
25
50
75
100
125
150
175
200
.00000
-.00001
25
50
75
100
125
150
175
200
-.00001
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Response of D_BETVARIANCE to D_ATXVARIANCE
Response of D_BETVARIANCE to D_WIGVARIANCE
Response of D_BETVARIANCE to D_BUXVARIANCE
Response of D_BETVARIANCE to D_PXVARIANCE
Response of D_BETVARIANCE to D_BETVARIANCE
Response of D_BETVARIANCE to D_SOFIXVARIANCE
.00012
.00012
.00012
.00012
.00012
.00012
.00010
.00010
.00010
.00010
.00010
.00010
.00008
.00008
.00008
.00008
.00008
.00008
.00006
.00006
.00006
.00006
.00006
.00006
.00004
.00004
.00004
.00004
.00004
.00004
.00002
.00002
.00002
.00002
.00002
.00000
.00000
.00000
.00000
.00000
.00000
-.00002
-.00002
-.00002
-.00002
-.00002
-.00002
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
.00002
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Response of D_SOFIXVARIANCE to D_ATXVARIANCE
Response of D_SOFIXVARIANCE to D_WIGVARIANCE
Response of D_SOFIXVARIANCE to D_BUXVARIANCE
Response of D_SOFIXVARIANCE to D_PXVARIANCE
Response of D_SOFIXVARIANCE to D_BETVARIANCE
Response of D_SOFIXVARIANCE to D_SOFIXVARIANCE
.00030
.00030
.00030
.00030
.00030
.00030
.00025
.00025
.00025
.00025
.00025
.00025
.00020
.00020
.00020
.00020
.00020
.00020
.00015
.00015
.00015
.00015
.00015
.00015
.00010
.00010
.00010
.00010
.00010
.00010
.00005
.00005
.00005
.00005
.00005
.00005
.00000
.00000
.00000
.00000
.00000
-.00005
-.00005
25
50
75
100
125
150
175
200
-.00005
25
50
75
100
125
150
175
200
-.00005
25
50
75
100
125
150
175
200
.00000
-.00005
25
50
75
100
125
150
175
200
-.00005
25
50
75
100
125
150
175
200
25
50
75
100
125
150
175
200
Variance Decomposition for volatility
 The initial shock in the volatilities
works through the system in about
90 days, exception being WIG
volatility which affects the market for
about 5 months.
 Changes in ATX volatility have a
positive influence on Poland, Czech
Republic and Hungarian volatility.
 The three larger emergent
markets are correlated in terms of
volatilities between themselves.
 Romanian and Bulgarian
volatilities are correlated with
volatilities in Poland and Hungary.
Period
1
30
90
Period
1
30
90
Period
1
30
90
Period
1
30
90
Period
1
30
90
Period
1
30
90
Variance Decomposition of ATX volatility
ATX
WIG
PX
BUX
BET
100.0000 0.000000 0.000000 0.000000 0.000000
98.28603 0.164898 0.025445 0.425067 0.343396
98.21085 0.181920 0.041278 0.452319 0.344078
Variance Decomposition of WIG volatility
ATX
WIG
PX
BUX
BET
6.809933 93.19007 0.000000 0.000000 0.000000
1.005268 87.43912 0.337155 1.220722 0.014976
9.585778 86.66214 0.287021 2.076229 0.031225
Variance Decomposition of PX volatility
ATX
WIG
PX
BUX
BET
21.79920 3.173114 75.02768 0.000000 0.000000
31.21845 4.910377 61.72163 1.784894 0.001183
31.42500 5.438533 60.84827 1.925888 0.001301
Variance Decomposition of BUX volatility
ATX
WIG
PX
BUX
BET
16.21773 4.740944 7.948067 71.09326 0.000000
29.01829 6.690134 13.64212 50.32515 0.205289
29.81494 7.025592 13.85145 48.98721 0.207251
Variance Decomposition of BET volatility
ATX
WIG
PX
BUX
BET
0.045655 0.001999 0.025464 0.005082 99.92180
0.236730 0.805854 0.233885 2.838162 95.52907
0.245924 1.730081 0.250796 3130038 94.27640
Variance Decomposition of SOFIX volatility
ATX
WIG
PX
BUX
BET
0.097580 0.002854 0.013009 0.029495 0.009514
0.131636 4.089252 0.035681 0.856636 0.048411
0.348825 8.306711 0.034418 1.198519 0.048603
SOFIX
0.000000
0.755163
0.769553
SOFIX
0.000000
0.935341
1.357606
SOFIX
0.000000
0.363466
0.361007
SOFIX
0.000000
0.119021
0.113559
SOFIX
0.000000
0.356304
0.366764
SOFIX
99.84755
94.83838
90.06292
Conclusions
 There are no cointegration relationships between the
countries under study
 None of the emerging markets has a significant influence on
the industrialized market
 There are unidirectional spillovers from Austria to Poland,
Hungary and the Czech Republic in term of returns and
volatility.
 Between the larger emerging markets there are correlations
relationships in returns and volatility.
 Romania and Bulgarian stock markets are driven mainly by
the developments at domestic level.
 Spillover effects between volatilities are stronger compared
to spillover effects between returns.
References
1) Angeloni, I., M. Flad, and F. P. Mongelli (2005), “Economic and Monetary Integration of the New
Member State. Helping to Chart the Route”, European Central Bank, Occasional Paper Series, no.36
2) Bekaert, G. and C.R., Harvey (1997), “Emerging Equity market volatility”, Journal of Financial
Economics 43
3) Bekaert,G., C.R. Harvey and A. Ng (2003), “Marketing Integration And Contagion”, NBER Working
Paper no.9510
4) Brooks, C (2002), “Introductory Econometrics for Finance”, Cambridge University Press
5) Cappiello, L., B. Gerard, A. Kadareja and S. Manganelli (2005), “Equity Market Integration of New EU
Member States”,
(2006), “Financial Integration of New EU Member States”, European Central Bank, Working Paper Series no. 683
6) Cerny, A., (2004), “Stock Market Integration and the Speed of Information Transmission”, Charles
University, Center for Economic Research and Graduate Education, Academy of Sciences of the
Czech Republic, Economic Institute, Working Paper Series 242
7) Cheung, Y.-L., Y.-W. Cheung and C.C. Ng (2006), “East Asian equity markets, financial crises, and the
Japanese currency”, Journal of The Japanese International Economies, 21, 138-152
8) Dabla-Norris, E. and H. Floerkemeier (2006), “Transmission Mechanisms of Monetary Policy in
Armenia: Evidence from VAR Analysis”, IMF Working Paper, 06/248
9) Danthine, J.-P., F. Giavazzi and E.L. Von Thadden (2000), “European Financial Markets After EMU: A
First Assessment”, NBER Working Paper no. 8044
10) Egert, B. and E. Kocenda (2005), “Contagion Across and Integration of Central and Eastern European
Stock Markets: Evidence from Intraday Data”, William Davidson Institute Working Paper, no. 798
11) Engle, R.F. and C.W.J. Granger (1987), “Co-Integration and Error Correction: Representation,
Estimation, and Testing”, Econometrica, 55, pp.251-276
12) Engle, R.F. and H. White (1999), “Cointegration, Causality, and Forecasting”, Oxford University Press
13) Forbes, K. and R. Rigobon (1999), “No Contagion, Only Interdependence Measuring Stock Market CoMovements”, NBER Working Paper no.7267
(2001), “Contagion in Latin America: Definitions, Measurement, and Policy
Implications”
14) Gujarati, D.N., (1995), “Basic Econometrics”, McGraw-Hill, Inc
15) Hajalmarsson, E. and P. Osterholm (2007), “Testing for Cointegration Using the Johansen
Methodology when Variables are Near-Integrated”, IMF Working Paper, 07/141
16) Hall, S.G., G. Hondroyiannis (2006), “Measuring the Correlation of Stocks between the EU-15 and the
New Member Countries”, Bank of Greece, Working Paper no.31
17) Harris, R.I.D. (1995), “Using Cointegration Analysis in Econometric Modeling”, Prentice Hall
18) Janakiramanan, S. and A.S. Lamba (1998), “An empirical examination of linkages between PacificBasin stock markets”, Journal of International Financial Markets, Institutions and Money, 8, 155-173
19) Miyakoshi, T. (2002), “Spillovers of stock return volatility to Asian equity markets from Japan and the
US”, Journal of International Financial Markets, Institutions and Money, 13,383-399
20) Nelson, D.B. (1991), “Conditional Heteroskedasticity in Asset Returns: A New Approach”, Econometrica,
no. 2
21) Nicolo, de G. and A. Tieman (2006), “Economic Integration and Financial Stability: A European
Perspective”, IMF Working Paper, 06/296
22) Obstfeld, M. and A.M. Taylor (2002), “Globalization and Capital Markets”, NBER Working Paper
no.8846
23) Phylaktis, K., (1999), “Capital market integration in the Pacific Basin region: an impulse
response analysis”, Journal of International Money and Finance, 18, 267-287
24) Rockinger, M. and G. Urga (2000), “A Time Varying Parameter Model To Test For
Predictability And Integration In Stock Markets Of Transition Economies”, CEPR Discussion
Paper no. 2346
25) Syllignakis, M. (2006), “EMU`s Impact on the Correlation across the European Stock
Markets”, International Research Journal of Finance and Economics
26) Terasvirta, T. (2006), “An Introduction to Univariate GARCH Models”, SSE/EFI Working
Papers in Economics and Finance, No. 646
27) http://www.bse.hu/onlinesz/index_e.html
28) http://www.bse-sofia.bg
29) http://www.bvb.ro
30) http://www.gpw.pl/index.asp
31) http://www.pse.cz
32) http://en.wienerborse.at
Engle Granger residual base cointegration test
Dependent Variable: ATX
Method: Least Squares
Date: 06/20/07 Time: 02:42
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
ATX=C(1)*BET+C(2)*BUX+C(3)*SOFIX+C(4)*PX+C(5)*WIG
Coefficient Std. Error t-Statistic
Prob.
C(1)
0.160751
0.005633
28.53943
0.0000
C(2)
0.056116
0.004273
13.13232
0.0000
C(3)
-0.033904 0.043034 -0.787851
0.4309
C(4)
0.233037
0.080167
2.906890
0.0037
C(5)
0.369370
0.012866
28.70958
0.0000
R-squared
0.988895
Mean dependent var
2173.383
Adjusted R-squared 0.988864
S.D. dependent var
1150.403
S.E. of regression
121.3962
Akaike info criterion
12.43941
Sum squared resid
21471850
Schwarz criterion
12.45749
Log likelihood
-9088.209
Durbin-Watson stat
0.045170
Dependent Variable: BET
Method: Least Squares
Date: 06/20/07 Time: 02:38
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
BET=C(1)*ATX+C(2)*BUX+C(3)*SOFIX+C(4)*PX+C(5)*WIG
Coefficient Std. Error t-Statistic
Prob.
Dependent Variable: BUX
Method: Least Squares
Date: 06/20/07 Time: 02:43
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
BUX=C(1)*ATX+C(2)*BET+C(3)*SOFIX+C(4)*PX+C(5)*WIG
Coefficient Std. Error t-Statistic
Prob.
C(1)
1.886040 0.143618
13.13232
0.0000
C(2)
-0.089182 0.040705 -2.190921
0.0286
C(3)
-3.807791 0.228730 -16.64756
0.0000
C(4)
13.40919 0.306340
43.77223
0.0000
C(5)
-0.159655 0.093236 -1.712374
0.0870
R-squared
0.987852
Mean dependent var
13089.12
Adjusted R-squared 0.987819
S.D. dependent var
6376.682
S.E. of regression
703.7783
Akaike info criterion
15.95422
Sum squared resid
7.22E+08
Schwarz criterion
15.97230
Log likelihood
-11657.53
Durbin-Watson stat
0.065961
Dependent Variable: PX
Method: Least Squares
Date: 07/09/07 Time: 03:45
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
PX=C(1)*ATX+C(2)*BET+C(3)*BUX+C(4)*SOFIX+C(5)*WIG
Coefficient Std. Error t-Statistic
Prob.
C(1)
0.024744 0.008512
2.906890
0.0037
C(2)
0.004620 0.002288
2.018791
0.0437
C(3)
0.042362 0.000968
43.77223
0.0000
C(4)
0.289009 0.011806
24.47900
0.0000
C(5)
0.045756 0.005107
8.959594
0.0000
R-squared
0.991982 Mean dependent var
856.4309
Adjusted R-squared
0.991960 S.D. dependent var
441.1593
S.E. of regression
39.55700 Akaike info criterion
10.19678
Sum squared resid
2279850. Schwarz criterion
10.21486
Log likelihood
-7448.844 Durbin-Watson stat
0.059943
C(1)
C(2)
C(3)
C(4)
C(5)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
2.230612
-0.036820
2.854737
0.603770
-1.450192
0.973051
0.972977
452.2098
2.98E+08
-11010.87
0.078159
28.53943
0.016806 -2.190921
0.141828
20.12813
0.299075
2.018791
0.046399 -31.25499
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Durbin-Watson stat
0.0000
0.0286
0.0000
0.0437
0.0000
3540.710
2750.874
15.06958
15.08767
0.037582
Engle Granger residual base cointegration test
Dependent Variable: SOFIX
Method: Least Squares
Date: 06/20/07 Time: 02:45
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
SOFIX=C(1)*ATX+C(2)*BET+C(3)*BUX+C(4)*PX+C(5)*WIG
Coefficient Std. Error t-Statistic
Prob.
C(1)
-0.012560 0.015942 -0.787851
0.4309
C(2)
0.076213 0.003786
20.12813
0.0000
C(3)
-0.041970 0.002521 -16.64756
0.0000
C(4)
1.008339 0.041192
24.47900
0.0000
C(5)
-0.030242 0.009766 -3.096585
0.0020
R-squared
0.959420 Mean dependent var
497.5816
Adjusted R-squared
0.959309 S.D. dependent var
366.2858
S.E. of regression
73.88751 Akaike info criterion
11.44638
Sum squared resid
7954294 Schwarz criterion
11.46446
Log likelihood
-8362.303 Durbin-Watson stat
0.029330
Dependent Variable: WIG
Method: Least Squares
Date: 06/20/07 Time: 02:49
Sample(adjusted): 10/20/2000 5/29/2006
Included observations: 1462 after adjusting endpoints
WIG=C(1)*ATX+C(2)*BET+C(3)*BUX+C(4)*SOFIX+C(5)*PX
Coefficient Std. Error t-Statistic
Prob.
C(1)
0.978186 0.034072
28.70958
0.0000
C(2)
-0.276768 0.008855 -31.25499
0.0000
C(3)
-0.012580 0.007347 -1.712374
0.0870
C(4)
-0.216194 0.069817 -3.096585
0.0020
C(5)
1.141233 0.127376
8.959594
0.0000
R-squared
0.920975 Mean dependent var
1872.073
Adjusted R-squared
0.920758 S.D. dependent var
701.7897
S.E. of regression
197.5536 Akaike info criterion
13.41331
Sum squared resid
56862958 Schwarz criterion
13.43139
Log likelihood
-9800.130 Durbin-Watson stat
0.034693
Estimating the volatilities
Dependent Variable: D_ATX
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:07
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 17 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic
Prob.
C
0.001524 0.000226 6.734204
0.0000
Variance Equation
C
1.29E-05 2.00E-06 6.481563
0.0000
ARCH(1)
0.125142 0.018366 6.813744
0.0000
GARCH(1)
0.765879 0.030141 25.40960
0.0000
R-squared
-0.002485
Mean dependent var 0.000999
Adjusted R-squared -0.004549
S.D. dependent var 0.010543
S.E. of regression
0.010567
Akaike info criterion -6.351087
Sum squared resid
0.162692
Schwarz criterion
-6.336612
Dependent Variable: D_BET
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:09
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 13 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic Prob.
C
0.001599 0.000305 5.247555
0.0000
Variance Equation
C
1.09E-05 1.69E-06 6.428384
0.0000
ARCH(1)
0.202481 0.021477 9.428005
0.0000
GARCH(1)
0.760199 0.021575 35.23453
0.0000
R-squared
-0.000293
Mean dependent var 0.001845
Adjusted R-squared -0.002352
S.D. dependent var 0.014388
S.E. of regression
0.014405
Akaike info criterion -5.866433
Sum squared resid
0.302325
Schwarz criterion
-5.851958
Log likelihood
4289.429
Durbin-Watson stat 1.639734
Dependent Variable: D_BUX
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:11
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 9 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic
Prob.
C
0.001038 0.000357 2.908883
0.0036
Variance Equation
Variance Equation
C
9.80E-06 2.83E-06 3.465744
0.0005
ARCH(1)
0.070711 0.013044 5.420837
0.0000
GARCH(1)
0.881748 0.022236 39.65325
0.0000
R-squared
-0.000434
Mean dependent var 0.000739
Adjusted R-squared -0.002494
S.D. dependent var
0.014391
S.E. of regression
0.014409
Akaike info criterion -5.704374
Sum squared resid
0.302492
Schwarz criterion
-5.689899
Log likelihood
4171.045
Durbin-Watson stat 1.945470
Dependent Variable: D_PX
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:15
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 18 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic
C
0.001361 0.000290 4.698530
Variance Equation
C
1.35E-05 2.83E-06 4.776848
ARCH(1)
0.113726 0.017111 6.646257
GARCH(1)
0.811543 0.030036 27.01892
R-squared
-0.001571 Mean dependent var
Adjusted R-squared -0.003633 S.D. dependent var
S.E. of regression
0.013225 Akaike info criterion
Sum squared resid
0.254829 Schwarz criterion
Log likelihood
4324.850 Durbin-Watson stat
Prob.
0.0000
0.0000
0.0000
0.0000
0.000837
0.013201
-5.914921
-5.900447
1.892869
Estimating the volatilities
Dependent Variable: D_SOFIX
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:13
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 41 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic
C
0.001427 0.000254 5.619383
Variance Equation
C
6.03E-07 1.34E-07 4.511134
ARCH(1)
0.106533 0.006070 17.55001
GARCH(1)
0.908129 0.003494 259.9401
R-squared
-0.000280 Mean dependent var
Adjusted R-squared -0.002339 S.D. dependent var
S.E. of regression
0.019327 Akaike info criterion
Sum squared resid
0.544256 Schwarz criterion
Log likelihood
4185.264 Durbin-Watson stat
Prob.
0.0000
0.0000
0.0000
0.0000
0.001750
0.019305
-5.723838
-5.709364
2.129854
Dependent Variable: D_WIG
Method: ML - ARCH (Marquardt)
Date: 06/20/07 Time: 03:16
Sample(adjusted): 10/23/2000 5/29/2006
Included observations: 1461 after adjusting endpoints
Convergence achieved after 12 iterations
Variance backcast: ON
Coefficient Std. Error z-Statistic
C
0.000860 0.000381 2.258107
Variance Equation
C
2.56E-06 1.02E-06 2.525784
ARCH(1)
0.036995 0.007936 4.661851
GARCH(1)
0.952114 0.010197 93.36855
R-squared
-0.000454 Mean dependent var
Adjusted R-squared -0.002514 S.D. dependent var
S.E. of regression
0.015624 Akaike info criterion
Sum squared resid
0.355682 Schwarz criterion
Log likelihood
4053.524 Durbin-Watson stat
Prob.
0.0239
0.0115
0.0000
0.0000
0.000527
0.015605
-5.543496
-5.529021
1.883032