Annexes
~ 1 ~
A. ANALYSE DE LA STATIONARITE A.1 CSS (Croissance
économique)
CSS à niveau sans tendance ni constance
Null Hypothesis: CSS has a unit root
Exogenous: None
Lag Length: 9 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -0.959149 0.2973
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(CSS)
Method: Least Squares
Date: 09/17/19 Time: 02:05
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
CSS(-1) -0.010602 0.011054 -0.959149 0.3427
D(CSS(-1)) 0.703874 0.132994 5.292526 0.0000
D(CSS(-2)) 0.146384 0.139140 1.052061 0.2985
D(CSS(-3)) 0.042183 0.141077 0.299009 0.7663
D(CSS(-4)) -0.973929 0.146505 -6.647737 0.0000
D(CSS(-5)) 0.705037 0.178102 3.958602 0.0003
D(CSS(-6)) 0.073763 0.140533 0.524880 0.6023
D(CSS(-7)) 0.012824 0.141675 0.090517 0.9283
D(CSS(-8)) -0.711354 0.157091 -4.528291 0.0000
D(CSS(-9)) 0.537907 0.147490 3.647070 0.0007
R-squared 0.658426 Mean dependent var -0.084740
Adjusted R-squared 0.588559 S.D. dependent var 0.785433
S.E. of regression 0.503805 Akaike info criterion 1.632322
Sum squared resid 11.16807 Schwarz criterion 2.000653
Log likelihood -34.07270 Hannan-Quinn criter.
1.774373
Durbin-Watson stat 2.079698
CSS en différence première sans tendance ni
constance
Null Hypothesis: D(CSS) has a unit root
Exogenous: None
Lag Length: 8 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -1.963887 0.0482
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(CSS,2)
Method: Least Squares
Date: 09/17/19 Time: 02:08
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
D(CSS(-1)) -0.476197 0.242477 -1.963887 0.0557
D(CSS(-1),2) 0.185175 0.208146 0.889638 0.3784
D(CSS(-2),2) 0.327944 0.203201 1.613890 0.1135
D(CSS(-3),2) 0.363175 0.202131 1.796732 0.0791
D(CSS(-4),2) -0.618014 0.203596 -3.035495 0.0040
D(CSS(-5),2) 0.092137 0.142927 0.644643 0.5224
D(CSS(-6),2) 0.164409 0.142061 1.157314 0.2533
D(CSS(-7),2) 0.173627 0.143861 1.206910 0.2338
D(CSS(-8),2) -0.540790 0.147328 -3.670646 0.0006
R-squared 0.648513 Mean dependent var -0.011007
Adjusted R-squared 0.586027 S.D. dependent var 0.782330
S.E. of regression 0.503357 Akaike info criterion 1.615978
Sum squared resid 11.40158 Schwarz criterion 1.947475
Log likelihood -34.63141 Hannan-Quinn criter.
1.743824
Durbin-Watson stat 2.068400
A.2 LBNBCC (taux du Bon BCC)
LBNBCC à niveau sans tendance ni constance
Null Hypothesis: LBNBCC has a unit root
Exogenous: None
Lag Length: 9 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -1.168533 0.2185
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LBNBCC)
Method: Least Squares
Date: 09/17/19 Time: 02:13
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LBNBCC(-1) -0.025168 0.021538 -1.168533 0.2489
D(LBNBCC(-1)) 0.696219 0.135485 5.138701 0.0000
D(LBNBCC(-2)) 0.167748 0.142478 1.177360 0.2454
D(LBNBCC(-3)) 0.058949 0.144696 0.407396 0.6857
D(LBNBCC(-4)) -1.081963 0.144538 -7.485668 0.0000
D(LBNBCC(-5)) 0.756907 0.187162 4.044130 0.0002
D(LBNBCC(-6)) 0.103419 0.143303 0.721683 0.4743
D(LBNBCC(-7)) 0.029240 0.144028 0.203016 0.8401
D(LBNBCC(-8)) -0.615956 0.142829 -4.312548 0.0001
D(LBNBCC(-9)) 0.430058 0.136170 3.158253 0.0029
R-squared 0.693985 Mean dependent var -0.031418
~ 2 ~
Adjusted R-squared 0.631391 S.D. dependent var 0.615837
S.E. of regression 0.373894 Akaike info criterion 1.035886
Sum squared resid 6.151052 Schwarz criterion 1.404217
Log likelihood -17.96893 Hannan-Quinn criter.
1.177937
Durbin-Watson stat 2.053745
LBNBCC en différence première avec tendance et
constance
Null Hypothesis: D(LBNBCC) has a unit root
Exogenous: None
Lag Length: 8 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -2.348632 0.0195
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LBNBCC,2) Method: Least Squares
Date: 09/17/19 Time: 02:21
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
D(LBNBCC(-1)) -0.573021 0.243981 -2.348632 0.0233
D(LBNBCC(-1),2) 0.260374 0.208728 1.247431 0.2187
D(LBNBCC(-2),2) 0.414989 0.205627 2.018160 0.0496
D(LBNBCC(-3),2) 0.455124 0.206955 2.199141 0.0330
D(LBNBCC(-4),2) -0.648925 0.205470 -3.158245 0.0028
D(LBNBCC(-5),2) 0.095886 0.139867 0.685551 0.4965
D(LBNBCC(-6),2) 0.194367 0.139627 1.392043 0.1708
D(LBNBCC(-7),2) 0.214097 0.140790 1.520686 0.1353
D(LBNBCC(-8),2) -0.415180 0.136122 -3.050046 0.0038
R-squared 0.702631 Mean dependent var -0.012906
Adjusted R-squared 0.649765 S.D. dependent var 0.634345
S.E. of regression 0.375409 Akaike info criterion 1.029411
Sum squared resid 6.341940 Schwarz criterion 1.360908
Log likelihood -18.79410 Hannan-Quinn criter.
1.157257
Durbin-Watson stat 2.022600
A.3 LCRED (niveau des crédits accordés
à l'économie)
LCRED à niveau avec constance sans tendance
Null Hypothesis: LCRED has a unit root
Exogenous: Constant
Lag Length: 8 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -4.357775 0.0009
Test critical values: 1% level -3.555023
5% level -2.915522
10% level -2.595565 *MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation Dependent Variable: D(LCRED) Method:
Least Squares
Date: 09/17/19 Time: 02:26
Sample (adjusted): 2005Q2 2018Q4 Included observations: 55 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LCRED(-1) -0.036544 0.008386 -4.357775 0.0001
D(LCRED(-1)) 0.446521 0.128657 3.470638 0.0012
D(LCRED(-2)) 0.199485 0.144189 1.383496 0.1733
D(LCRED(-3)) 0.102486 0.145439 0.704666 0.4846
D(LCRED(-4)) -1.004966 0.137742 -7.295978 0.0000
D(LCRED(-5)) 0.314475 0.138369 2.272720 0.0279
D(LCRED(-6)) 0.139659 0.145536 0.959621 0.3424
D(LCRED(-7)) 0.076717 0.144429 0.531178 0.5979
D(LCRED(-8)) -0.447108 0.129838 -3.443576 0.0013
C 0.594038 0.130801 4.541552 0.0000
R-squared 0.732942 Mean dependent var 0.067449
Adjusted R-squared 0.679530 S.D. dependent var 0.079739
S.E. of regression 0.045140 Akaike info criterion -3.195125
Sum squared resid 0.091693 Schwarz criterion -2.830155
Log likelihood 97.86593 Hannan-Quinn criter. -3.053988
F-statistic 13.72252 Durbin-Watson stat 1.841220
Prob(F-statistic) 0.000000
A.4 LRO (taux du coefficient de réserve
obligatoire)
LRO à niveau avec tendance et constance
Null Hypothesis: LRO has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 9 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -3.493347 0.0502
Test critical values: 1% level -4.137279
5% level -3.495295
10% level -3.176618
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LRO)
Method: Least Squares
Date: 09/17/19 Time: 02:27
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LRO(-1) -0.111596 0.031945 -3.493347 0.0011
D(LRO(-1)) 0.662966 0.115537 5.738135 0.0000
D(LRO(-2)) 0.169449 0.099603 1.701245 0.0963
D(LRO(-3)) 0.116711 0.100497 1.161341 0.2521
D(LRO(-4)) -1.059319 0.102859 -10.29874 0.0000
D(LRO(-5)) 0.758620 0.148938 5.093517 0.0000
D(LRO(-6)) 0.094570 0.097681 0.968153 0.3385
~ 3 ~
D(LRO(-7)) 0.061263 0.098127 0.624321 0.5358
D(LRO(-8)) -0.673348 0.102087 -6.595801 0.0000
D(LRO(-9)) 0.495259 0.104962 4.718456 0.0000
C 0.135205 0.031029 4.357351 0.0001
@TREND("2003Q1"
) 0.002530 0.000961 2.631447 0.0118
R-squared 0.850510 Mean dependent var 0.036746
Adjusted R-squared 0.811358 S.D. dependent var 0.074937
S.E. of regression 0.032547 Akaike info criterion -3.819108
Sum squared resid 0.044492 Schwarz criterion -3.377112
Log likelihood 115.1159 Hannan-Quinn criter. -3.648648
F-statistic 21.72321 Durbin-Watson stat 2.251481
Prob(F-statistic) 0.000000
LRO en différence première sans tendance ni
constance
Null Hypothesis: D(LRO) has a unit root
Exogenous: None
Lag Length: 8 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -1.017734 0.2740
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LRO,2)
Method: Least Squares
Date: 09/17/19 Time: 02:29
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
D(LRO(-1)) -0.126813 0.124603 -1.017734 0.3142
D(LRO(-1),2) 0.051206 0.128605 0.398163 0.6924
D(LRO(-2),2) 0.129846 0.127159 1.021135 0.3126
D(LRO(-3),2) 0.146242 0.127933 1.143115 0.2590
D(LRO(-4),2) -1.026712 0.129354 -7.937200 0.0000
D(LRO(-5),2) 0.030763 0.101845 0.302051 0.7640
D(LRO(-6),2) 0.079529 0.101208 0.785804 0.4361
D(LRO(-7),2) 0.087908 0.102246 0.859767 0.3945
D(LRO(-8),2) -0.659730 0.103499 -6.374284 0.0000
R-squared 0.790180 Mean dependent var -0.000154
Adjusted R-squared 0.752879 S.D. dependent var 0.076791
S.E. of regression 0.038174 Akaike info criterion -3.542319
Sum squared resid 0.065576 Schwarz criterion -3.210822
Log likelihood 104.6426 Hannan-Quinn criter. -3.414474
Durbin-Watson stat 2.219603
LRO en différence seconde sans tendance ni constance
Null Hypothesis: D(LRO,2) has a unit root
Exogenous: None
Lag Length: 7 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -6.768039 0.0000
Test critical values: 1% level -2.608490
5% level -1.946996
10% level -1.612934
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LRO,3)
Method: Least Squares
Date: 09/17/19 Time: 02:31
Sample (adjusted): 2005Q3 2018Q4 Included observations: 54 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
D(LRO(-1),2) -2.665969 0.393906 -6.768039 0.0000
D(LRO(-1),3) 1.629761 0.364543 4.470693 0.0001
D(LRO(-2),3) 1.672494 0.319914 5.227955 0.0000
D(LRO(-3),3) 1.730753 0.266811 6.486807 0.0000
D(LRO(-4),3) 0.614808 0.216756 2.836413 0.0068
D(LRO(-5),3) 0.607945 0.190346 3.193890 0.0025
D(LRO(-6),3) 0.649365 0.150760 4.307267 0.0001
D(LRO(-7),3) 0.698734 0.096179 7.264933 0.0000
R-squared 0.917909 Mean dependent var -0.002715
Adjusted R-squared 0.905417 S.D. dependent var 0.124173
S.E. of regression 0.038189 Akaike info criterion -3.556600
Sum squared resid 0.067085 Schwarz criterion -3.261936
Log likelihood 104.0282 Hannan-Quinn criter. -3.442959
Durbin-Watson stat 2.259729
A.5 LTXDIR (taux directeur)
LTXDIR à niveau sans tendance ni constance
Null Hypothesis: LTXDIR has a unit root
Exogenous: None
Lag Length: 5 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -1.012808 0.2762
Test critical values: 1% level -2.605442
5% level -1.946549
10% level -1.613181
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LTXDIR)
Method: Least Squares
Date: 09/17/19 Time: 02:33
Sample (adjusted): 2004Q3 2018Q4 Included observations: 58 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
LTXDIR(-1) -0.008530 0.008422 -1.012808 0.3158
D(LTXDIR(-1)) 0.705277 0.128163 5.502965 0.0000
D(LTXDIR(-2)) 0.176527 0.138348 1.275965 0.2076
~ 4 ~
D(LTXDIR(-3)) 0.052633 0.140325 0.375079 0.7091
D(LTXDIR(-4)) -0.587642 0.138541 -4.241661 0.0001
D(LTXDIR(-5)) 0.389492 0.129765 3.001521 0.0041
R-squared 0.591583 Mean dependent var -0.006143
Adjusted R-squared 0.552313 S.D. dependent var 0.257431
S.E. of regression 0.172245 Akaike info criterion -0.582097
Sum squared resid 1.542759 Schwarz criterion -0.368948
Log likelihood 22.88081 Hannan-Quinn criter. -0.499071
Durbin-Watson stat 2.010728
LTXDIR en différence sans tendance ni constance
Null Hypothesis: D(LTXDIR) has a unit root
Exogenous: None
Lag Length: 4 (Automatic - based on SIC, maxlag=10)
t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -2.259952 0.0242
Test critical values: 1% level -2.605442
5% level -1.946549
10% level -1.613181
*MacKinnon (1996) one-sided p-values. Augmented Dickey-Fuller
Test Equation Dependent Variable: D(LTXDIR,2)
Method: Least Squares
Date: 09/17/19 Time: 02:35
Sample (adjusted): 2004Q3 2018Q4 Included observations: 58 after
adjustments
Variable Coefficient Std. Error t-Statistic Prob.
D(LTXDIR(-1)) -0.293286 0.129775 -2.259952 0.0280
D(LTXDIR(-1),2) -0.000615 0.132686 -0.004634 0.9963
D(LTXDIR(-2),2) 0.174680 0.132134 1.321988 0.1918
D(LTXDIR(-3),2) 0.221553 0.132890 1.667198 0.1014
D(LTXDIR(-4),2) -0.376137 0.129124 -2.912978 0.0052
R-squared 0.400110 Mean dependent var -0.008671
Adjusted R-squared 0.354835 S.D. dependent var 0.214495
S.E. of regression 0.172287 Akaike info criterion -0.597045
Sum squared resid 1.573193 Schwarz criterion -0.419421
Log likelihood 22.31431 Hannan-Quinn criter. -0.527857
Durbin-Watson stat 1.989906
~ 5 ~
| 
