Rendement et volatilité en présence de noise traders

Annexe 10 : la stationnarité de la variable D_ARMS1( )

Null Hypothesis: D_ARMS1 has a unit root
Exogenous: Constant
Lag Length: 1 (Automatic - based on SIC, maxlag=18)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-15.05929	0.0000
Test critical values:	1% level		-3.442746
	5% level		-2.866900
	10% level		-2.569686
*MacKinnon (1996) one-sided p-values.

Annexe 11 : la stationnarité de la variable DARMS2 ( )

Null Hypothesis: D_ARMS2 has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic - based on SIC, maxlag=18)
			t-Statistic	Prob.*
Augmented Dickey-Fuller test statistic			-23.59193	0.0000
Test critical values:	1% level		-3.442722
	5% level		-2.866889
	10% level		-2.569680
*MacKinnon (1996) one-sided p-values.

Annexe 12 : estimation du modèle de base avant l'introduction les variables de sentiment

Dependent Variable: R
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 06/13/12 Time: 18:27
Sample: 1 520
Included observations: 520
Convergence achieved after 42 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(3) + C(4)*RESID(-1)^2 + C(5)*GARCH(-1)
Variable	Coefficient	Std. Error	z-Statistic	Prob.
@SQRT(GARCH)	3.798726	0.448970	8.460970	0.0000
C	-0.110163	0.010965	-10.04716	0.0000
	Variance Equation
C	7.26E-07	1.79E-06	0.406206	0.6846
RESID(-1)^2	0.026470	0.004159	6.364097	0.0000
GARCH(-1)	0.974994	0.004881	199.7345	0.0000
R-squared	0.283947	Mean dependent var		-0.017887
Adjusted R-squared	0.282564	S.D. dependent var		0.026763
S.E. of regression	0.022669	Akaike info criterion		-4.748160
Sum squared resid	0.266185	Schwarz criterion		-4.707257
Log likelihood	1239.522	Hannan-Quinn criter.		-4.732137
Durbin-Watson stat	1.787865

annexe 13 : estimation du modèle avec la variable dsent Dependent Variable: R
Method: ML - ARCH (Marquardt) - Generalized error distribution (GED)
Date: 06/13/12 Time: 18:41
Sample: 1 520
Included observations: 520
Convergence achieved after 73 iterations
Presample variance: backcast (parameter = 0.7)
GED parameter fixed at 1.5
GARCH = C(4) + C(5)*RESID(-1)^2 + C(6)*GARCH(-1) + C(7)*DT_DSENT2
+ C(8)*DT1_DSENT2
Variable	Coefficient	Std. Error		z-Statistic			Prob.
@SQRT(GARCH)	3.575849	0.342221		10.44896			0.0000
C	-0.098380	0.008214		-11.97737			0.0000
DSENT	0.001987	0.000212		9.383172			0.0000
	Variance Equation
C	-1.01E-06	1.21E-06		-0.838263			0.4019
RESID(-1)^2	0.020858	0.003144		6.633284			0.0000
GARCH(-1)	0.978774	0.004057		241.2722			0.0000
DT_DSENT2	-7.75E-08	3.89E-08		-1.993871			0.0462
DT1_DSENT2	6.67E-07	1.47E-07		4.522635			0.0000
R-squared	0.433196	Mean dependent var					-0.017887
Adjusted R-squared	0.431003	S.D. dependent var					0.026763
S.E. of regression	0.020188	Akaike info criterion					-5.010931
Sum squared resid	0.210703	Schwarz criterion					-4.945487
Log likelihood	1310.842	Hannan-Quinn criter.					-4.985294
Durbin-Watson stat	1.701980