3.3.1.2. Décalage optimale du modèle
VAR
Endogenous variables: DTPIB TPRIM TSEC
DDEPCAH
Exogenous variables: C
|
LogL
|
LR
|
FPE
|
AIC
|
SC
|
HQ
|
|
|
5888175
|
26.9398
|
27.1284
|
26.9989
|
|
-386.6279
|
NA*
|
.*
|
6*
|
5*
|
2*
|
|
|
1034454
|
|
|
|
|
-378.5986
|
13.28981
|
2
|
27.48956
|
28.43252
|
27.78489
|
|
|
1665357
|
|
|
|
|
-368.4773
|
13.96046
|
6
|
27.89499
|
29.59232
|
28.42657
|
|
|
2898476
|
|
|
|
|
-357.7653
|
11.82020
|
4
|
28.25967
|
30.71138
|
29.02752
|
* indicates lag order selected by the
criterion
LR: sequential modified LR test statistic (each test
at 5%
level)
FPE: Final prediction error
AIC: Akaike information
criterion
SC: Schwarz information
criterion
HQ: Hannan-Quinn information criterion
Pour trouver le décalage optimale nous allons faire
recours aux critères (AKAIKE ET SCHWARZ) ces deux critères nous
permettra de minimiser le nombre de paramètres dans le modèle et
c'est celui qui sera choisi, le décalage minimise les deux
critères à un Lag de 0 mais vu que le modèle exige un auto
régressif des équations nous considérons directement le
Lag 1 qui constitue notre première décalage.
L'Impact des dépenses publiques e
n Education sur l a croissance E c o n o m i q u e e n R D C
d e 1 9 8 0 à 2 0 1 2
Page 66
3.3.1.3. Estimation du modèle retenu de
l'estimation du
modèle vectoriel autorégressif VAR (-1)
avec la constante.
|
Vector Autoregression Estimates
Date: 09/15/13 Time: 16:55
Sample (adjusted): 1982
2012 Included observations: 31 after adjustments Standard errors in ( ) &
t-statistics in [ ]
|
|
|
|
|
DTPIB
|
TPRIM
|
TSEC
|
DDEPCAH
|
|
DTPIB(-1) -0.120691
|
0.554986
|
12.41863
|
0.006116
|
|
(0.19604)
|
(0.52364)
|
(14.6639)
|
(0.00990)
|
|
[-0.61566]
|
[ 1.05986]
|
[ 0.84688]
|
[ 0.61778]
|
|
TPRIM(-1) 0.056248
|
-0.262688
|
-6.080605
|
-0.001360
|
|
(0.06891)
|
(0.18407)
|
(5.15479)
|
(0.00348)
|
|
[ 0.81623]
|
[-1.42708]
|
[-1.17960]
|
[-0.39073]
|
|
TSEC(-1) 0.001398
|
-0.002767
|
0.205017
|
0.000163
|
|
(0.00251)
|
(0.00671)
|
(0.18783)
|
(0.00013)
|
|
[ 0.55661]
|
[-0.41245]
|
[ 1.09148]
|
[ 1.28486]
|
|
DDEPCAH(-1) 0.370240
|
13.60800
|
-59.77445
|
-0.375207
|
|
(3.53693)
|
(9.44768)
|
(264.572)
|
(0.17861)
|
|
[ 0.10468]
|
[ 1.44035]
|
[-0.22593]
|
[-2.10075]
|
|
C 0.348764
|
0.766530
|
-124.4712
|
0.034114
|
|
(0.82381)
|
(2.20051)
|
(61.6229)
|
(0.04160)
|
|
[ 0.42336]
|
[ 0.34834]
|
[-2.01989]
|
[ 0.82004]
|
|
R-squared 0.042198
|
0.155672
|
0.127589
|
0.208570
|
|
Adj. R-squared -0.105156
|
0.025776
|
-0.006628
|
0.086812
|
|
Sum sq. resids 408.0979
|
2911.805
|
2283486.
|
1.040652
|
|
S.E. equation 3.961827
|
10.58265
|
296.3552
|
0.200063
|
|
F-statistic 0.286370
|
1.198433
|
0.950619
|
1.712983
|
|
Log likelihood -83.93865
|
-114.3965
|
-217.6991
|
8.622072
|
|
Akaike AIC 5.737978
|
7.702999
|
14.36768
|
-0.233682
|
|
Schwarz SC 5.969266
|
7.934287
|
14.59897
|
-0.002394
|
|
Mean dependent 0.151613
|
1.070323
|
-162.3877
|
0.006474
|
|
S.D. dependent 3.768631
|
10.72173
|
295.3779
|
0.209356
|
|
Determinant resid covariance (dof adj.)
|
5357403.
|
|
|
|
Determinant resid covariance
|
2650946.
|
|
|
|
Log likelihood
|
-405.2000
|
|
|
|
Akaike information criterion
|
27.43226
|
|
|
|
Schwarz criterion
|
28.35741
|
|
|
|
|
|
|
C'est le modèle par principe de parcimonie qui comporte
moins des paramètres possibles.
L'Impact des dépenses publiques e
n Education sur l a croissance E c o n o m i q u e e n R D C
d e 1 9 8 0 à 2 0 1 2
Page 67
| 
