2.3.2. Le taux d'inflation
(TINFL)
a) Analyse graphique
Graphique N°6 : Taux d'inflation
Le graphique présente cette variabilité autour
de n constante signe de la stationnarité de la série.
c) Test de la racine unitaire (Test ADF)
|
Null Hypothesis: TINFL has a unit root
|
|
|
Exogenous: Constant, Linear Trend
|
|
|
Lag Length: 5 (Automatic - based on SIC, maxlag=13)
|
|
|
|
|
|
|
|
|
t-Statistic
|
Prob.*
|
|
|
|
|
|
|
Augmented Dickey-Fuller test statistic
|
-3.066660
|
0.1180
|
|
Test critical values:
|
1% level
|
|
-4.016064
|
|
|
5% level
|
|
-3.437977
|
|
|
10% level
|
|
-3.143241
|
|
|
|
|
|
|
|
*MacKinnon (1996) one-sided p-values.
|
|
|
Augmented Dickey-Fuller Test Equation
|
|
|
Dependent Variable: D(TINFL)
|
|
|
|
Method: Least Squares
|
|
|
|
Date: 08/22/15 Time: 15:33
|
|
|
|
Sample (adjusted): 2000M08 2013M12
|
|
|
Included observations: 161 after adjustments
|
|
|
|
|
|
|
|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
Prob.
|
|
|
|
|
|
|
TINFL(-1)
|
-0.234049
|
0.076321
|
-3.066660
|
0.0026
|
|
D(TINFL(-1))
|
-0.452098
|
0.098292
|
-4.599534
|
0.0000
|
|
D(TINFL(-2))
|
-0.648177
|
0.101151
|
-6.407990
|
0.0000
|
|
D(TINFL(-3))
|
-0.272864
|
0.103400
|
-2.638920
|
0.0092
|
|
D(TINFL(-4))
|
-0.136312
|
0.086283
|
-1.579823
|
0.1162
|
|
D(TINFL(-5))
|
0.153236
|
0.071555
|
2.141518
|
0.0338
|
|
C
|
0.560033
|
0.844359
|
0.663264
|
0.5082
|
|
@TREND("2000M02")
|
-0.002929
|
0.007703
|
-0.380191
|
0.7043
|
|
|
|
|
|
|
R-squared
|
0.503465
|
Mean dependent var
|
-0.086211
|
|
Adjusted R-squared
|
0.480748
|
S.D. dependent var
|
5.478952
|
|
S.E. of regression
|
3.948087
|
Akaike info criterion
|
5.632752
|
|
Sum squared resid
|
2384.871
|
Schwarz criterion
|
5.785865
|
|
Log likelihood
|
-445.4365
|
Hannan-Quinn criter.
|
5.694922
|
|
F-statistic
|
22.16220
|
Durbin-Watson stat
|
1.984417
|
|
Prob(F-statistic)
|
0.000000
|
|
|
|
|
|
|
|
|
Le test sur le modèle avec tendance et constante
indique que la statistique ADF est non significatif (probabilité
critique=11,8%) mais le trend dans l'équation du test est
également non significatif par conséquent nous passons au
modèle sans tendance pour la décision du test.
|
Null Hypothesis: TINFL has a unit root
|
|
|
Exogenous: Constant
|
|
|
|
Lag Length: 5 (Automatic - based on SIC, maxlag=13)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
t-Statistic
|
Prob.*
|
|
|
|
|
|
|
|
|
|
|
|
Augmented Dickey-Fuller test statistic
|
-3.308541
|
0.0161
|
|
Test critical values:
|
1% level
|
|
-3.471192
|
|
|
5% level
|
|
-2.879380
|
|
|
10% level
|
|
-2.576361
|
|
|
|
|
|
|
|
|
|
|
|
|
*MacKinnon (1996) one-sided p-values.
|
|
|
Augmented Dickey-Fuller Test Equation
|
|
|
Dependent Variable: D(TINFL)
|
|
|
|
Method: Least Squares
|
|
|
|
Date: 08/22/15 Time: 15:33
|
|
|
|
Sample (adjusted): 2000M08 2013M12
|
|
|
Included observations: 161 after adjustments
|
|
|
|
|
|
|
|
|
|
|
|
|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
Prob.
|
|
|
|
|
|
|
|
|
|
|
|
TINFL(-1)
|
-0.219915
|
0.066469
|
-3.308541
|
0.0012
|
|
D(TINFL(-1))
|
-0.465542
|
0.091456
|
-5.090330
|
0.0000
|
|
D(TINFL(-2))
|
-0.659862
|
0.096101
|
-6.866352
|
0.0000
|
|
D(TINFL(-3))
|
-0.281612
|
0.100526
|
-2.801386
|
0.0057
|
|
D(TINFL(-4))
|
-0.141870
|
0.084799
|
-1.673023
|
0.0964
|
|
D(TINFL(-5))
|
0.150282
|
0.070934
|
2.118621
|
0.0357
|
|
C
|
0.269484
|
0.358046
|
0.752652
|
0.4528
|
|
|
|
|
|
|
|
|
|
|
|
R-squared
|
0.502996
|
Mean dependent var
|
-0.086211
|
|
Adjusted R-squared
|
0.483632
|
S.D. dependent var
|
5.478952
|
|
S.E. of regression
|
3.937106
|
Akaike info criterion
|
5.621274
|
|
Sum squared resid
|
2387.124
|
Schwarz criterion
|
5.755248
|
|
Log likelihood
|
-445.5125
|
Hannan-Quinn criter.
|
5.675673
|
|
F-statistic
|
25.97610
|
Durbin-Watson stat
|
1.984769
|
|
Prob(F-statistic)
|
0.000000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Le test sur le modèle avec constante sans tendance nous
indique que la statistique ADF est significatif (probabilité
critique=1,61%) la série est donc stationnaire à nouveau.
| 
