crises financières et contagion: cas de subprime

ANNEXES

Statistiques descriptives

Test de stationnarité : test ADF

Null Hypothesis: LOG(DOW) has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 2 (Automatic based on SIC, MAXLAG=2)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-2.441149	0.3580
Test critical values:	1% level		-3.958914
	5% level		-3.410233
Date: 03/25/08 Time: 11:07
Sample (adjusted): 12/20/1967 12/17/2007
Included observations: 10434 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
LOG(DOW(-1))	-0.000789	0.000323	-2.441149	0.0147
D(LOG(DOW(-1)))	0.053757	0.009786	5.493515	0.0000
D(LOG(DOW(-2)))	-0.031709	0.009787	-3.240044	0.0012
C	0.004895	0.001987	2.463897	0.0138
@TREND(12/15/1967)	2.90E-07	1.09E-07	2.660624	0.0078
Null Hypothesis: D(LOG(DOW)) has a unit root
Exogenous: Constant
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-24.38133	0.0000
Test critical values:	1% level		-3.430803
	5% level		-2.861625
Date: 03/25/08 Time: 11:41
Sample (adjusted): 1/10/1968 12/17/2007
Included observations: 10419 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000256	9.67E-05	2.641640	0.0083
Null Hypothesis: LOG(MDAX) has a unit root
Exogenous: None
Lag Length: 1 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			3.109105	0.9996
Test critical values:	1% level		-2.565406
	5% level		-1.940885
Dependent Variable: D(LOG(MDAX))
Method: Least Squares
Date: 03/25/08 Time: 11:52
Sample (adjusted): 1/01/1988 12/17/2007
Included observations: 5207 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
LOG(MDAX(-1))	4.89E-05	1.57E-05	3.109105	0.0019
D(LOG(MDAX(-1)))	0.059925	0.013846	4.328000	0.0000
Null Hypothesis: D(LOG(MDAX)) has a unit root
Exogenous: Constant
Lag Length: 15 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-16.06369	0.0000
Test critical values:	1% level		-3.431428
	5% level		-2.861901
Date: 03/25/08 Time: 12:13
Sample (adjusted): 1/22/1988 12/17/2007
Included observations: 5192 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000348	0.000129	2.702353	0.0069
Null Hypothesis: LOG(CAC) has a unit root
Exogenous: None
Lag Length: 1 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			1.301714	0.9517
Test critical values:	1% level		-2.565396
	5% level		-1.940884
Date: 03/25/08 Time: 16:04
Sample (adjusted): 7/13/1987 12/17/2007
Included observations: 5331 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
LOG(CAC(-1))	2.90E-05	2.23E-05	1.301714	0.1931
D(LOG(CAC(-1)))	0.015095	0.013699	1.101882	0.2706
Null Hypothesis: D(LOG(CAC)) has a unit root
Exogenous: None
Lag Length: 15 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-17.04505	0.0000
Test critical values:	1% level		-2.565397
	5% level		-1.940884
	10% level		-1.616660
Date: 03/25/08 Time: 16:03
Sample (adjusted): 8/03/1987 12/17/2007
Included observations: 5316 after adjustments
Null Hypothesis: LOG(FTSE) has a unit root
Exogenous: Constant
Lag Length: 1 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-2.269817	0.1820
Test critical values:	1% level		-3.432097
	5% level		-2.862197
Date: 03/25/08 Time: 19:06
Sample (adjusted): 1/03/1995 12/17/2007
Included observations: 3380 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
LOG(FTSE(-1))	-0.001212	0.000534	-2.269817	0.0233
D(LOG(FTSE(-1)))	0.109198	0.017103	6.384664	0.0000
C	0.008407	0.003571	2.354485	0.0186
Null Hypothesis: D(LOG(FTSE)) has a unit root
Exogenous: Constant
Lag Length: 15 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-13.59434	0.0000
Test critical values:	1% level		-3.432106
	5% level		-2.862201
	10% level		-2.567166
Date: 03/25/08 Time: 19:26
Sample (adjusted): 1/24/1995 12/17/2007
Included observations: 3365 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000328	0.000158	2.077054	0.0379
Null Hypothesis: LOG(NIKK) has a unit root
Exogenous: Constant
Lag Length: 2 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-2.148151	0.2259
Test critical values:	1% level		-3.431682
	5% level		-2.862014
	10% level		-2.567065
Date: 03/25/08 Time: 19:16
Sample (adjusted): 6/05/1991 12/17/2007
Included observations: 4314 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
LOG(NIKK(-1))	-0.001692	0.000788	-2.148151	0.0318
D(LOG(NIKK(-1)))	-0.032611	0.015218	-2.142883	0.0322
D(LOG(NIKK(-2)))	-0.034651	0.015217	-2.277067	0.0228
C	0.016198	0.007603	2.130523	0.0332
Null Hypothesis: D(LOG(NIKK)) has a unit root
Exogenous: None
Lag Length: 17 (Fixed)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-15.09390	0.0000
Test critical values:	1% level		-2.565500
	5% level		-1.940898
	10% level		-1.616650
Date: 03/25/08 Time: 19:21
Sample (adjusted): 6/27/1991 12/17/2007
Included observations: 4298 after adjustments

Modèle VAR

VAR Lag Order Selection Criteria
Endogenous variables: LOG(DOW) LOG(CAC) LOG(MDAX) LOG(NIKK) LOG(FTSE)
Date: 05/04/08 Time: 02:33
Sample: 12/15/1967 12/17/2007
Included observations: 3378
Lag	LogL	LR	FPE	AIC	SC	HQ
0	7674.940	NA	7.34e-09	-4.541113	-4.532047	-4.537871
1	56738.78	97953.38	1.80e-21	-33.57536	-33.52096	-33.55591
2	57229.55	978.3573	1.37e-21	-33.85113	-33.75140*	-33.81547
3	57282.13	104.6506	1.35e-21	-33.86745	-33.72240	-33.81559*
4	57311.12	57.62619*	1.34e-21*	-33.86982*	-33.67943	-33.80175
* indicates lag order selected by the criterion

Test de Johanson

Date: 04/16/08 Time: 23:15
Sample (adjusted): 1/05/1995 12/17/2007
Included observations: 3378 after adjustments
Series: LOG(DOW) LOG(CAC) LOG(MDAX) LOG(NIKK) LOG(FTSE)
Lags interval (in first differences): 1 to 3
Unrestricted Cointegration Rank Test (Trace)
Hypothesized		Trace	0.05
No. of CE(s)	Eigenvalue	Statistic	Critical Value	Prob.**
None *	0.008680	77.66879	76.97277	0.0442
At most 1	0.007136	48.21862	54.07904	0.1503
At most 2	0.003719	24.02741	35.19275	0.4610
At most 3	0.002424	11.43993	20.26184	0.5003
At most 4	0.000960	3.243380	9.164546	0.5362

Modèle du modèle à correction des erreurs vectoriel

Vector Error Correction Estimates
Date: 05/04/08 Time: 02:36
Sample (adjusted): 1/06/1995 12/17/2007
Included observations: 3377 after adjustments
Standard errors in ( ) & t-statistics in [ ]
Error Correction:	D(LOG(DOW))	D(LOG(CAC))	D(LOG(MDAX))	D(LOG(NIKK))	D(LOG(FTSE))
CointEq1	-0.010419	-0.008908	-0.000339	-0.000825	-0.009245
	(0.00236)	(0.00292)	(0.00201)	(0.00295)	(0.00208)
	[-4.42333]	[-3.04839]	[-0.16812]	[-0.27989]	[-4.44568]

Modélisation ARMA

Dependent Variable: D(LOG(CAC))
Method: Least Squares
Date: 04/03/08 Time: 23:14
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000247	0.000167	1.485076	0.1376
AR(1)	0.695699	0.218390	3.185580	0.0015
MA(1)	-0.682641	0.218485	-3.124431	0.0018
MA(2)	-0.031566	0.014250	-2.215082	0.0268
R-squared	0.001225	Mean dependent var		0.000245
Adjusted R-squared	0.000662	S.D. dependent var		0.012952
S.E. of regression	0.012947	Akaike info criterion		-5.855085
Sum squared resid	0.893000	Schwarz criterion		-5.850147
Log likelihood	15610.73	F-statistic		2.177278
Durbin-Watson stat	1.997988	Prob(F-statistic)		0.088547
Dependent Variable: D(LOG(CAC))
Method: Least Squares
Date: 04/03/08 Time: 23:17
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000248	0.000166	1.492873	0.1355
AR(1)	0.726477	0.200030	3.631844	0.0003
AR(2)	-0.033764	0.014168	-2.383054	0.0172
MA(1)	-0.711826	0.199887	-3.561136	0.0004
R-squared	0.001342	Mean dependent var		0.000245
Adjusted R-squared	0.000780	S.D. dependent var		0.012953
S.E. of regression	0.012948	Akaike info criterion		-5.855015
Sum squared resid	0.892895	Schwarz criterion		-5.850076
Log likelihood	15607.61	F-statistic		2.386231
Durbin-Watson stat	2.001364	Prob(F-statistic)		0.067131
Dependent Variable: D(LOG(DOW))
Method: Least Squares
Date: 04/03/08 Time: 23:18
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000258	9.34E-05	2.765274	0.0057
AR(1)	0.716355	0.146850	4.878122	0.0000
AR(2)	-0.062078	0.010176	-6.100513	0.0000
MA(1)	-0.663828	0.146974	-4.516620	0.0000
R-squared	0.004082	Mean dependent var		0.000259
Adjusted R-squared	0.003796	S.D. dependent var		0.009830
S.E. of regression	0.009811	Akaike info criterion		-6.410184
Sum squared resid	1.004005	Schwarz criterion		-6.407403
Log likelihood	33445.93	F-statistic		14.25131
Durbin-Watson stat	1.999011	Prob(F-statistic)		0.000000

Dependent Variable: D(LOG(DOW))
Method: Least Squares
Date: 04/03/08 Time: 23:18
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000259	9.32E-05	2.774526	0.0055
AR(1)	0.674693	0.145744	4.629312	0.0000
MA(1)	-0.621502	0.145618	-4.268032	0.0000
MA(2)	-0.062989	0.010222	-6.161833	0.0000
R-squared	0.004128	Mean dependent var		0.000259
Adjusted R-squared	0.003842	S.D. dependent var		0.009829
S.E. of regression	0.009811	Akaike info criterion		-6.410325
Sum squared resid	1.003960	Schwarz criterion		-6.407545
Log likelihood	33449.87	F-statistic		14.41310
Durbin-Watson stat	2.000318	Prob(F-statistic)		0.000000

Dependent Variable: D(LOG(FTSE))
Method: Least Squares
Date: 04/03/08 Time: 23:19
Sample (adjusted): 1/03/1995 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000352	0.000152	2.309437	0.0210
AR(1)	0.720484	0.162080	4.445243	0.0000
MA(1)	-0.612252	0.161609	-3.788475	0.0002
MA(2)	-0.114658	0.019465	-5.890439	0.0000
R-squared	0.015076	Mean dependent var		0.000349
Adjusted R-squared	0.014201	S.D. dependent var		0.009123
S.E. of regression	0.009058	Akaike info criterion		-6.569147
Sum squared resid	0.276994	Schwarz criterion		-6.561898
Log likelihood	11105.86	F-statistic		17.22574
Durbin-Watson stat	1.993941	Prob(F-statistic)		0.000000
Dependent Variable: D(LOG(FTSE))
Method: Least Squares
Date: 04/03/08 Time: 23:19
Sample (adjusted): 1/04/1995 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000352	0.000152	2.320476	0.0204
AR(1)	0.811782	0.143817	5.644546	0.0000
AR(2)	-0.119852	0.018550	-6.460947	0.0000
MA(1)	-0.700211	0.144254	-4.854006	0.0000
R-squared	0.015630	Mean dependent var		0.000350
Adjusted R-squared	0.014755	S.D. dependent var		0.009124
S.E. of regression	0.009057	Akaike info criterion		-6.569444
Sum squared resid	0.276830	Schwarz criterion		-6.562192
Log likelihood	11103.07	F-statistic		17.86273
Durbin-Watson stat	2.000970	Prob(F-statistic)		0.000000

Dependent Variable: D(LOG(NIKK))
Method: Least Squares
Date: 04/03/08 Time: 23:19
Sample (adjusted): 6/05/1991 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	-0.000119	0.000196	-0.605751	0.5447
AR(1)	0.067715	0.388465	0.174314	0.8616
AR(2)	-0.031280	0.020580	-1.519911	0.1286
MA(1)	-0.101040	0.388710	-0.259937	0.7949
R-squared	0.002271	Mean dependent var		-0.000120
Adjusted R-squared	0.001577	S.D. dependent var		0.013835
S.E. of regression	0.013824	Akaike info criterion		-5.723828
Sum squared resid	0.823708	Schwarz criterion		-5.717922
Log likelihood	12350.30	F-statistic		3.270523
Durbin-Watson stat	1.999445	Prob(F-statistic)		0.020331

Dependent Variable: D(LOG(NIKK))
Method: Least Squares
Date: 04/03/08 Time: 23:21
Sample (adjusted): 6/04/1991 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	-0.000122	0.000200	-0.609179	0.5424
AR(1)	-0.740684	0.167818	-4.413615	0.0000
MA(1)	0.707506	0.168016	4.210950	0.0000
MA(2)	-0.049049	0.015265	-3.213198	0.0013
R-squared	0.002568	Mean dependent var		-0.000123
Adjusted R-squared	0.001874	S.D. dependent var		0.013835
S.E. of regression	0.013822	Akaike info criterion		-5.724131
Sum squared resid	0.823650	Schwarz criterion		-5.718226
Log likelihood	12353.81	F-statistic		3.700290
Durbin-Watson stat	1.998320	Prob(F-statistic)		0.011260

Dependent Variable: D(LOG(MDAX))
Method: Least Squares
Date: 04/06/08 Time: 12:50
Sample (adjusted): 5/15/1991 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000330	0.000191	1.727807	0.0841
AR(1)	1.015754	0.037623	26.99830	0.0000
AR(2)	-0.052929	0.016715	-3.166538	0.0016
MA(1)	-0.947258	0.034152	-27.73659	0.0000
R-squared	0.007141	Mean dependent var		0.000337
Adjusted R-squared	0.006452	S.D. dependent var		0.008835
S.E. of regression	0.008806	Akaike info criterion		-6.625751
Sum squared resid	0.335416	Schwarz criterion		-6.619862
Log likelihood	14345.44	F-statistic		10.36839
Durbin-Watson stat	2.000350	Prob(F-statistic)		0.000001

Dependent Variable: D(LOG(MDAX))
Method: Least Squares
Date: 04/06/08 Time: 12:52
Sample (adjusted): 5/14/1991 12/17/2007
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000331	0.000190	1.742084	0.0816
AR(1)	0.958656	0.029600	32.38746	0.0000
MA(1)	-0.892365	0.033498	-26.63929	0.0000
MA(2)	-0.049133	0.016304	-3.013511	0.0026
R-squared	0.007000	Mean dependent var		0.000336
Adjusted R-squared	0.006312	S.D. dependent var		0.008834
S.E. of regression	0.008806	Akaike info criterion		-6.625782
Sum squared resid	0.335483	Schwarz criterion		-6.619895
Log likelihood	14348.82	F-statistic		10.16561
Durbin-Watson stat	1.995773	Prob(F-statistic)		0.000001

Modélisation ARMA(1,2)-GARCH(1,1)

Dependent Variable: D(LOG(MDAX))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/06/08 Time: 13:19
Sample (adjusted): 5/14/1991 12/17/2007
GARCH = C(5) + C(6)*RESID(-1)^2 + C(7)*GARCH(-1)
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000874	0.000161	5.436838	0.0000
AR(1)	0.996938	0.000771	1292.484	0.0000
MA(1)	-0.904335	0.015275	-59.20353	0.0000
MA(2)	-0.093016	0.015374	-6.049992	0.0000
	Variance Equation
C	5.31E-06	5.61E-07	9.466094	0.0000
RESID(-1)^2	0.157596	0.010911	14.44425	0.0000
GARCH(-1)	0.778500	0.015346	50.72966	0.0000
R-squared	-0.010362	Mean dependent var		0.000336
Adjusted R-squared	-0.011765	S.D. dependent var		0.008834
S.E. of regression	0.008886	Akaike info criterion		-6.860999
Sum squared resid	0.341349	Schwarz criterion		-6.850695
Log likelihood	14861.06	Durbin-Watson stat		2.017654
Dependent Variable: D(LOG(CAC))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000544	0.000134	4.043721	0.0001
AR(1)	0.845703	0.115981	7.291753	0.0000
MA(1)	-0.825988	0.115486	-7.152267	0.0000
MA(2)	-0.034525	0.015243	-2.264996	0.0235
	Variance Equation
C	3.35E-06	3.44E-07	9.720807	0.0000
RESID(-1)^2	0.087733	0.005131	17.09699	0.0000
GARCH(-1)	0.891909	0.006406	139.2249	0.0000
R-squared	0.000125	Mean dependent var		0.000245
Adjusted R-squared	-0.001002	S.D. dependent var		0.012952
S.E. of regression	0.012958	Akaike info criterion		-6.105239
Sum squared resid	0.893983	Schwarz criterion		-6.096597
Log likelihood	16280.51	F-statistic		0.110894
Durbin-Watson stat	2.008991	Prob(F-statistic)		0.995204

Dependent Variable: D(LOG(DOW))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000494	8.46E-05	5.835302	0.0000
AR(1)	-0.423187	0.672958	-0.628847	0.5294
MA(1)	0.507802	0.673383	0.754105	0.4508
MA(2)	0.026955	0.061774	0.436343	0.6626
	Variance Equation
C	1.11E-06	9.52E-08	11.65722	0.0000
RESID(-1)^2	0.067885	0.001462	46.44196	0.0000
GARCH(-1)	0.922571	0.002428	379.9038	0.0000
R-squared	0.001942	Mean dependent var		0.000259
Adjusted R-squared	0.001367	S.D. dependent var		0.009829
S.E. of regression	0.009823	Akaike info criterion		-6.661176
Sum squared resid	1.006164	Schwarz criterion		-6.656311
Log likelihood	34761.69	F-statistic		3.380995
Durbin-Watson stat	2.059405	Prob(F-statistic)		0.002479

Dependent Variable: D(LOG(NIKK))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000228	0.000180	1.265985	0.2055
AR(1)	-0.975227	0.018710	-52.12427	0.0000
MA(1)	0.956207	0.025935	36.86921	0.0000
MA(2)	-0.022283	0.017122	-1.301457	0.1931
	Variance Equation
C	3.73E-06	4.25E-07	8.776883	0.0000
RESID(-1)^2	0.073469	0.005685	12.92267	0.0000
GARCH(-1)	0.908725	0.006811	133.4233	0.0000
R-squared	0.001266	Mean dependent var		-0.000123
Adjusted R-squared	-0.000125	S.D. dependent var		0.013835
S.E. of regression	0.013836	Akaike info criterion		-5.846792
Sum squared resid	0.824726	Schwarz criterion		-5.836458
Log likelihood	12621.45	F-statistic		0.909862
Durbin-Watson stat	2.024012	Prob(F-statistic)		0.486518

Dependent Variable: D(LOG(FTSE))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000656	0.000114	5.737062	0.0000
AR(1)	0.826285	0.092548	8.928134	0.0000
MA(1)	-0.712654	0.091854	-7.758557	0.0000
MA(2)	-0.125761	0.018191	-6.913323	0.0000
	Variance Equation
C	4.88E-07	1.05E-07	4.664111	0.0000
RESID(-1)^2	0.060624	0.005668	10.69560	0.0000
GARCH(-1)	0.934517	0.006126	152.5498	0.0000
R-squared	0.013380	Mean dependent var		0.000349
Adjusted R-squared	0.011625	S.D. dependent var		0.009123
S.E. of regression	0.009070	Akaike info criterion		-6.840866
Sum squared resid	0.277471	Schwarz criterion		-6.828180
Log likelihood	11568.06	F-statistic		7.624003
Durbin-Watson stat	2.001128	Prob(F-statistic)		0.000000

Modélisation ARMA(1,2)E-GARCH

Dependent Variable: D(LOG(FTSE))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/06/08 Time: 20:15
Sample (adjusted): 1/03/1995 12/17/2007
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1))
	Coefficient	Std. Error	z-Statistic	Prob.
C	-0.000158	0.000613	-0.257897	0.7965
AR(1)	0.996702	0.002797	356.3754	0.0000
MA(1)	-0.870517	0.017800	-48.90432	0.0000
MA(2)	-0.117193	0.017561	-6.673539	0.0000
	Variance Equation
C(5)	-0.139763	0.019822	-7.050841	0.0000
C(6)	0.080565	0.009796	8.224362	0.0000
C(7)	-0.079220	0.006451	-12.28041	0.0000
C(8)	0.991905	0.001595	621.7916	0.0000
R-squared	0.009925	Mean dependent var		0.000349
Adjusted R-squared	0.007869	S.D. dependent var		0.009123
S.E. of regression	0.009087	Akaike info criterion		-6.860387
Sum squared resid	0.278443	Schwarz criterion		-6.845889
Log likelihood	11602.05	F-statistic		4.828804
Durbin-Watson stat	2.019154	Prob(F-statistic)		0.000020
Dependent Variable: D(LOG(DOW))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000244	0.000103	2.374792	0.0176
AR(1)	0.997223	0.000791	1260.495	0.0000
MA(1)	-0.920639	2.95E-05	-31172.45	0.0000
MA(2)	-0.076298	0.000378	-201.7813	0.0000
	Variance Equation
C(5)	-0.131403	0.007266	-18.08566	0.0000
C(6)	0.062011	0.002450	25.31559	0.0000
C(7)	-0.025455	0.002093	-12.16462	0.0000
C(8)	0.991168	0.000663	1494.805	0.0000
R-squared	0.002466	Mean dependent var		0.000259
Adjusted R-squared	0.001797	S.D. dependent var		0.009829
S.E. of regression	0.009821	Akaike info criterion		-6.657012
Sum squared resid	1.005635	Schwarz criterion		-6.651452
Log likelihood	34740.96	F-statistic		3.682861
Durbin-Watson stat	2.044455	Prob(F-statistic)		0.000557

Dependent Variable: D(LOG(MDAX))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000328	0.000145	2.252701	0.0243
AR(1)	0.553010	0.195592	2.827362	0.0047
MA(1)	-0.461546	0.196232	-2.352040	0.0187
MA(2)	-0.003541	0.028723	-0.123288	0.9019
	Variance Equation
C(5)	-0.635444	0.057311	-11.08758	0.0000
C(6)	0.145575	0.008599	16.92946	0.0000
C(7)	-0.085175	0.007918	-10.75721	0.0000
C(8)	0.945329	0.005499	171.8960	0.0000
R-squared	0.005976	Mean dependent var		0.000336
Adjusted R-squared	0.004367	S.D. dependent var		0.008834
S.E. of regression	0.008815	Akaike info criterion		-6.862758
Sum squared resid	0.335829	Schwarz criterion		-6.840983
Log likelihood	14864.22	F-statistic		3.712214
Durbin-Watson stat	2.043936	Prob(F-statistic)		0.000520

Dependent Variable: D(LOG(NIKK))
	Coefficient	Std. Error	z-Statistic	Prob.
C	-8.10E-05	0.000175	-0.462430	0.6438
AR(1)	-0.951570	0.037303	-25.50914	0.0000
MA(1)	0.931769	0.042222	22.06837	0.0000
MA(2)	-0.026501	0.016995	-1.559318	0.1189
	Variance Equation
C(5)	-0.322514	0.027099	-11.90134	0.0000
C(6)	0.136560	0.009809	13.92142	0.0000
C(7)	-0.076946	0.005740	-13.40617	0.0000
C(8)	0.974714	0.002621	371.8839	0.0000
R-squared	0.002162	Mean dependent var		-0.000123
Adjusted R-squared	0.000540	S.D. dependent var		0.013835
S.E. of regression	0.013832	Akaike info criterion		-5.869641
Sum squared resid	0.823986	Schwarz criterion		-5.857831
Log likelihood	12671.75	F-statistic		1.333210
Durbin-Watson stat	2.023802	Prob(F-statistic)		0.229971
Dependent Variable: D(LOG(CAC))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000303	0.000148	2.043747	0.0410
AR(1)	-0.888330	0.014354	-61.88906	0.0000
MA(1)	0.909546	0.001833	496.1335	0.0000
MA(2)	0.029151	0.014192	2.054048	0.0400
	Variance Equation
C(5)	-0.274801	0.017413	-15.78157	0.0000
C(6)	0.113312	0.005239	21.62936	0.0000
C(7)	-0.074317	0.005312	-13.99143	0.0000
C(8)	0.978941	0.001800	543.9941	0.0000
R-squared	0.000719	Mean dependent var		0.000245
Adjusted R-squared	-0.000595	S.D. dependent var		0.012952
S.E. of regression	0.012956	Akaike info criterion		-6.116993
Sum squared resid	0.893452	Schwarz criterion		-6.107116
Log likelihood	16312.84	F-statistic		0.547043
Durbin-Watson stat	2.010317	Prob(F-statistic)		0.799168

Modélisation : ARMA(1,2)-GARCH(1,1)-M

Dependent Variable: D(LOG(DOW))
Date: 04/28/08 Time: 12:57
Included observations: 3905 after adjustments
Convergence achieved after 144 iterations
MA backcast: 12/25/1992 12/28/1992, Variance backcast: ON
GARCH = C(6) + C(7)*RESID(-1)^2 + C(8)*GARCH(-1)
	Coefficient	Std. Error	z-Statistic	Prob.
GARCH	4.903921	2.617447	1.873551	0.0610
C	0.000254	0.000200	1.268935	0.2045
AR(1)	-0.647980	0.255892	-2.532240	0.0113
MA(1)	0.653715	0.255901	2.554561	0.0106
MA(2)	0.000386	0.017420	0.022147	0.9823
	Variance Equation
C	9.22E-07	1.23E-07	7.506099	0.0000
RESID(-1)^2	0.068143	0.005026	13.55868	0.0000
GARCH(-1)	0.923635	0.005681	162.5833	0.0000
R-squared	0.000212	Mean dependent var		0.000352
Adjusted R-squared	-0.001583	S.D. dependent var		0.009811
S.E. of regression	0.009819	Akaike info criterion		-6.667024
Sum squared resid	0.375712	Schwarz criterion		-6.654179
Log likelihood	13025.36	F-statistic		0.118266
Durbin-Watson stat	2.031544	Prob(F-statistic)		0.997144
Dependent Variable: D(LOG(CAC))
	Coefficient	Std. Error	z-Statistic	Prob.
GARCH	4.274848	1.758515	2.430943	0.0151
C	-7.10E-05	0.000240	-0.295895	0.7673
AR(1)	0.355481	0.571300	0.622232	0.5338
MA(1)	-0.340661	0.570525	-0.597102	0.5504
MA(2)	-0.024345	0.015434	-1.577381	0.1147
	Variance Equation
C	3.50E-06	3.69E-07	9.480571	0.0000
RESID(-1)^2	0.088359	0.005217	16.93606	0.0000
GARCH(-1)	0.890069	0.006587	135.1247	0.0000
R-squared	-0.003138	Mean dependent var		0.000245
Adjusted R-squared	-0.004457	S.D. dependent var		0.012952
S.E. of regression	0.012981	Akaike info criterion		-6.104082
Sum squared resid	0.896901	Schwarz criterion		-6.094206
Log likelihood	16278.43	Durbin-Watson stat		1.987431

Dependent Variable: D(LOG(NIKK))
	Coefficient	Std. Error	z-Statistic	Prob.
GARCH	2.625754	2.247081	1.168518	0.2426
C	-0.000169	0.000384	-0.439334	0.6604
AR(1)	-0.975269	0.018697	-52.16283	0.0000
MA(1)	0.955806	0.025987	36.78000	0.0000
MA(2)	-0.022708	0.017202	-1.320105	0.1868
	Variance Equation
C	3.79E-06	4.33E-07	8.762121	0.0000
RESID(-1)^2	0.073429	0.005708	12.86449	0.0000
GARCH(-1)	0.908394	0.006870	132.2358	0.0000
R-squared	0.001877	Mean dependent var		-0.000123
Adjusted R-squared	0.000255	S.D. dependent var		0.013835
S.E. of regression	0.013834	Akaike info criterion		-5.846661
Sum squared resid	0.824221	Schwarz criterion		-5.834851
Log likelihood	12622.17	F-statistic		1.157107
Durbin-Watson stat	2.023721	Prob(F-statistic)		0.324132

Dependent Variable: D(LOG(FTSE))
	Coefficient	Std. Error	z-Statistic	Prob.
GARCH	0.085594	3.139956	0.027260	0.9783
C	0.000652	0.000197	3.302720	0.0010
AR(1)	0.825377	0.097521	8.463620	0.0000
MA(1)	-0.711748	0.097278	-7.316665	0.0000
MA(2)	-0.125681	0.018339	-6.853219	0.0000
	Variance Equation
C	4.88E-07	1.06E-07	4.611566	0.0000
RESID(-1)^2	0.060635	0.005684	10.66706	0.0000
GARCH(-1)	0.934504	0.006166	151.5682	0.0000
R-squared	0.013375	Mean dependent var		0.000349
Adjusted R-squared	0.011327	S.D. dependent var		0.009123
S.E. of regression	0.009071	Akaike info criterion		-6.840274
Sum squared resid	0.277472	Schwarz criterion		-6.825776
Log likelihood	11568.06	F-statistic		6.530439
Durbin-Watson stat	2.001091	Prob(F-statistic)		0.000000
Dependent Variable: D(LOG(MDAX))
	Coefficient	Std. Error	z-Statistic	Prob.
GARCH	9.715800	2.965677	3.276082	0.0011
C	0.000470	0.000294	1.600729	0.1094
AR(1)	0.996686	0.001757	567.3282	0.0000
MA(1)	-0.904248	0.017957	-50.35680	0.0000
MA(2)	-0.091624	0.017636	-5.195323	0.0000
	Variance Equation
C	5.01E-06	4.61E-07	10.86957	0.0000
RESID(-1)^2	0.158989	0.011339	14.02158	0.0000
GARCH(-1)	0.779478	0.014484	53.81704	0.0000
R-squared	0.003582	Mean dependent var		0.000336
Adjusted R-squared	0.001968	S.D. dependent var		0.008834
S.E. of regression	0.008825	Akaike info criterion		-6.861429
Sum squared resid	0.336638	Schwarz criterion		-6.849653
Log likelihood	14862.99	F-statistic		2.219498
Durbin-Watson stat	2.004154	Prob(F-statistic)		0.029908

Transmission en moyenne

Dependent Variable: D(LOG(CAC))
Method: Least Squares
Date: 04/06/08 Time: 21:02
Sample (adjusted): 7/10/1987 12/17/2007
Included observations: 5332 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000157	0.000173	0.908820	0.3635
D(LOG(DOW(-1)))	0.282519	0.016171	17.47110	0.0000

Dependent Variable: D(LOG(NIKK))
Method: Least Squares
Date: 04/06/08 Time: 21:06
Sample (adjusted): 6/03/1991 12/17/2007
Included observations: 4316 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	-0.000260	0.000203	-1.282886	0.1996
D(LOG(DOW(-1)))	0.399382	0.021120	18.91024	0.0000

Dependent Variable: D(LOG(FTSE))
Method: Least Squares
Date: 04/06/08 Time: 21:06
Sample (adjusted): 1/02/1995 12/17/2007
Included observations: 3381 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000309	0.000156	1.985390	0.0472
D(LOG(DOW(-1)))	0.105979	0.015180	6.981607	0.0000

Dependent Variable: D(LOG(MDAX))
Method: Least Squares
Date: 04/06/08 Time: 21:06
Sample (adjusted): 5/13/1991 12/17/2007
Included observations: 4331 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000231	0.000127	1.816794	0.0693
D(LOG(DOW(-1)))	0.297763	0.013265	22.44764	0.0000

Transmission en variance

Dependent Variable: D(LOG(CAC))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/06/08 Time: 21:14
Sample (adjusted): 7/13/1987 12/17/2007
Included observations: 5331 after adjustments
Convergence achieved after 75 iterations
MA backcast: 7/09/1987 7/10/1987, Variance backcast: ON
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1)) +
C(9)*E(-1)^2
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000278	0.000150	1.859171	0.0630
AR(1)	-0.914986	0.066335	-13.79344	0.0000
MA(1)	0.940164	0.066468	14.14452	0.0000
MA(2)	0.037319	0.014791	2.523136	0.0116
	Variance Equation
C(5)	-0.314027	0.019850	-15.82036	0.0000
C(6)	0.107548	0.006314	17.03395	0.0000
C(7)	-0.072388	0.005644	-12.82584	0.0000
C(8)	0.974473	0.002009	485.0948	0.0000
C(9)	29.61134	6.324429	4.682057	0.0000

Dependent Variable: D(LOG(FTSE))
	Coefficient	Std. Error	z-Statistic	Prob.
C	-0.000242	0.000650	-0.371793	0.7100
AR(1)	0.996011	0.003444	289.1607	0.0000
MA(1)	-0.866870	0.018246	-47.50911	0.0000
MA(2)	-0.118056	0.017680	-6.677356	0.0000
	Variance Equation
C(5)	-0.174712	0.027020	-6.466112	0.0000
C(6)	0.070092	0.010532	6.655349	0.0000
C(7)	-0.082202	0.006709	-12.25234	0.0000
C(8)	0.987885	0.002521	391.9285	0.0000
C(9)	38.40408	19.53376	1.966036	0.0493

Dependent Variable: D(LOG(NIKK))
	Coefficient	Std. Error	z-Statistic	Prob.
C	-7.33E-05	0.000175	-0.418164	0.6758
AR(1)	-0.959303	0.029233	-32.81532	0.0000
MA(1)	0.940280	0.035092	26.79452	0.0000
MA(2)	-0.025184	0.017030	-1.478808	0.1392
	Variance Equation
C(5)	-0.395135	0.032038	-12.33345	0.0000
C(6)	0.133564	0.010352	12.90286	0.0000
C(7)	-0.075835	0.006021	-12.59492	0.0000
C(8)	0.967169	0.003104	311.6083	0.0000
C(9)	95.19737	14.06179	6.769931	0.0000

Dependent Variable: D(LOG(MDAX))
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000331	0.000148	2.242193	0.0249
AR(1)	0.588946	0.179652	3.278253	0.0010
MA(1)	-0.497582	0.179934	-2.765366	0.0057
MA(2)	-0.004393	0.028313	-0.155150	0.8767
	Variance Equation
C(5)	-0.981700	0.089610	-10.95527	0.0000
C(6)	0.164040	0.013923	11.78212	0.0000
C(7)	-0.075554	0.009955	-7.589596	0.0000
C(8)	0.912850	0.008388	108.8244	0.0000
C(9)	185.9741	28.42676	6.542222	0.0000

Test de contagion (avec une variable Dumy)

En moyenne

Dependent Variable: D(LOG(CAC))
Method: Least Squares
Date: 04/13/08 Time: 19:36
Sample (adjusted): 7/10/1987 12/17/2007
Included observations: 5332 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000160	0.000172	0.929442	0.3527
D(LOG(DOW(-1)))	0.285536	0.016101	17.73413	0.0000
D(LOG(DOW))*DUMMY(-1)	0.728506	0.102054	7.138411	0.0000

Dependent Variable: D(LOG(NIKK))
Method: Least Squares
Date: 04/13/08 Time: 19:40
Sample (adjusted): 6/03/1991 12/17/2007
Included observations: 4316 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	-0.000259	0.000203	-1.280414	0.2005
D(LOG(DOW(-1)))	0.400148	0.021130	18.93706	0.0000
D(LOG(DOW))*DUMMY(-1)	0.120825	0.108240	1.116271	0.2644

Dependent Variable: D(LOG(FTSE))
Method: Least Squares
Date: 04/13/08 Time: 19:41
Sample (adjusted): 1/02/1995 12/17/2007
Included observations: 3381 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000314	0.000153	2.046390	0.0408
D(LOG(DOW(-1)))	0.111444	0.014941	7.459007	0.0000
D(LOG(DOW))*DUMMY(-1)	0.774697	0.072545	10.67885	0.0000

Dependent Variable: D(LOG(MDAX))
Method: Least Squares
Date: 04/13/08 Time: 19:42
Sample (adjusted): 12/31/1987 12/17/2007
Included observations: 5208 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	0.000314	0.000119	2.643713	0.0082
D(LOG(DOW(-1)))	0.332304	0.012275	27.07227	0.0000
D(LOG(DOW))*DUMMY(-1)	0.733479	0.069796	10.50897	0.0000

En variance

Dependent Variable: D(LOG(MDAX))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/13/08 Time: 19:56
Sample (adjusted): 1/01/1988 12/17/2007
Included observations: 5207 after adjustments
Convergence achieved after 153 iterations
MA backcast: 12/30/1987 12/31/1987, Variance backcast: ON
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1)) +
C(9)*E(-1)^2 + C(10)*DUMMY(-1)*E(-1)^2
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000439	0.000152	2.888360	0.0039
AR(1)	0.883926	0.050379	17.54555	0.0000
MA(1)	-0.781102	0.053346	-14.64219	0.0000
MA(2)	-0.056469	0.018763	-3.009686	0.0026
	Variance Equation
C(5)	-1.675837	0.094003	-17.82739	0.0000
C(6)	0.216894	0.018728	11.58144	0.0000
C(7)	-0.073034	0.010762	-6.786302	0.0000
C(8)	0.848336	0.008683	97.70180	0.0000
C(9)	468.2025	14.86657	31.49365	0.0000
C(10)	1035.459	256.2599	4.040662	0.0001
Dependent Variable: D(LOG(CAC))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/13/08 Time: 20:24
Sample (adjusted): 7/13/1987 12/17/2007
Included observations: 5331 after adjustments
Convergence achieved after 48 iterations
MA backcast: 7/09/1987 7/10/1987, Variance backcast: ON
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1)) +
C(9)*E(-1)^2 + C(10)*DUMMY(-1)*E(-1)^2
	Coefficient	Std. Error	z-Statistic	Prob.
C	0.000277	0.000150	1.853738	0.0638
AR(1)	-0.915036	0.066353	-13.79045	0.0000
MA(1)	0.940178	0.066495	14.13916	0.0000
MA(2)	0.037277	0.014794	2.519786	0.0117
	Variance Equation
C(5)	-0.314083	0.019880	-15.79927	0.0000
C(6)	0.107449	0.006318	17.00748	0.0000
C(7)	-0.072437	0.005654	-12.81281	0.0000
C(8)	0.974458	0.002012	484.2545	0.0000
C(9)	29.93935	6.324655	4.733751	0.0000
C(10)	-19.34045	83.37548	-0.231968	0.8166
Dependent Variable: D(LOG(NIKK))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/13/08 Time: 20:26
Sample (adjusted): 6/04/1991 12/17/2007
Included observations: 4315 after adjustments
Convergence achieved after 22 iterations
MA backcast: 5/31/1991 6/03/1991, Variance backcast: ON
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1)) +
C(9)*E(-1)^2 + C(10)*DUMMY(-1)*E(-1)^2
	Coefficient	Std. Error	z-Statistic	Prob.
C	-7.10E-05	0.000176	-0.403677	0.6865
AR(1)	-0.959903	0.028603	-33.55963	0.0000
MA(1)	0.940777	0.034547	27.23208	0.0000
MA(2)	-0.025246	0.017025	-1.482858	0.1381
	Variance Equation
C(5)	-0.393796	0.031971	-12.31721	0.0000
C(6)	0.133549	0.010369	12.87938	0.0000
C(7)	-0.075264	0.006099	-12.34035	0.0000
C(8)	0.967328	0.003095	312.5291	0.0000
C(9)	94.33492	13.96116	6.756955	0.0000
C(10)	34.74344	79.86350	0.435035	0.6635

Dependent Variable: D(LOG(FTSE))
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/13/08 Time: 20:29
Sample (adjusted): 1/03/1995 12/17/2007
Included observations: 3380 after adjustments
Convergence achieved after 24 iterations
MA backcast: 12/30/1994 1/02/1995, Variance backcast: ON
LOG(GARCH) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(GARCH(-1))) +
C(7)*RESID(-1)/@SQRT(GARCH(-1)) + C(8)*LOG(GARCH(-1)) +
C(9)*E(-1)^2 + C(10)*DUMMY(-1)*E(-1)^2
	Coefficient	Std. Error	z-Statistic	Prob.
C	-0.000226	0.000630	-0.359339	0.7193
AR(1)	0.995898	0.003486	285.6815	0.0000
MA(1)	-0.866165	0.018328	-47.26019	0.0000
MA(2)	-0.118582	0.017734	-6.686801	0.0000
	Variance Equation
C(5)	-0.175133	0.026804	-6.533775	0.0000
C(6)	0.067880	0.010560	6.427941	0.0000
C(7)	-0.081668	0.006655	-12.27211	0.0000
C(8)	0.987705	0.002502	394.7233	0.0000
C(9)	40.28242	19.37862	2.078705	0.0376
C(10)	43.05302	55.43153	0.776688	0.0373