ANNEXES
Annexe 1 : résultats de la
stationnarité sortis de STATA Tableau 1 : test de
stationnarité de DFA
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
Date: 04/27/13 Time: 09:33
|
|
|
|
|
Sample: 1 68
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA, FPSTA, FPN,
|
|
|
ETI, DPTD, DBANC, CBANC
|
|
|
|
|
Exogenous variables: None
|
|
|
|
|
Automatic selection of maximum lags
|
|
|
|
Automatic selection of lags based on SIC: 0 to 7
|
|
|
Total number of observations: 701
|
|
|
|
Cross-sections included: 11 (1 dropped)
|
|
|
|
Method
|
Statistic
|
|
Prob.**
|
|
ADF - Fisher Chi-square
|
55.9648
|
|
0.0001
|
|
ADF - Choi Z-stat
|
-0.56722
|
|
0.2853
|
|
** Probabilities for Fisher tests are computed using an asympotic
Chi
|
|
|
-square distribution. All other tests assume asymptotic
|
|
|
normality.
|
|
|
|
|
Intermediate ADF test results UNTITLED
|
|
|
|
Series
|
Prob.
|
Lag
|
Max Lag
|
Obs
|
|
TA
|
0.0733
|
6
|
|
10
|
61
|
|
RTCSTD
|
0.0512
|
1
|
|
10
|
66
|
|
RRSA
|
0.0654
|
1
|
|
10
|
66
|
|
ROE
|
0.0000
|
0
|
|
10
|
67
|
|
ROA
|
0.0063
|
1
|
|
10
|
66
|
|
FPSTA
|
0.0002
|
1
|
|
10
|
66
|
|
FPN
|
0.7557
|
6
|
|
10
|
61
|
|
ETI
|
0.6568
|
1
|
|
10
|
66
|
|
DPTD
|
0.0685
|
7
|
|
10
|
60
|
|
DBANC
|
0.9998
|
6
|
|
10
|
61
|
|
CBANC
|
0.0823
|
6
|
|
10
|
61
|
|
Null Hypothesis: Unit root (individual unit root process)
|
|
Date: 04/27/13 Time: 09:36
|
|
|
|
Sample: 1 68
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA, FPSTA, FPN,
|
|
ETI, DPTD, DBANC, CBANC
|
|
|
|
Exogenous variables: Individual effects
|
|
|
Automatic selection of maximum lags
|
|
|
Automatic selection of lags based on SIC: 0 to 6
|
|
Total number of observations: 722
|
|
|
Cross-sections included: 11 (1 dropped)
|
|
|
Method
|
Statistic
|
Prob.**
|
78
|
ADF - Fisher Chi-square
|
184.089
|
|
0.0000
|
|
ADF - Choi Z-stat
|
-7.46930
|
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an asympotic
Chi
|
|
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
Date: 04/27/13 Time: 09:36
|
|
|
|
|
Sample: 1 68
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA, FPSTA, FPN,
|
|
|
ETI, DPTD, DBANC, CBANC
|
|
|
|
Exogenous variables: Individual effects, individual linear
trends
|
|
|
Automatic selection of maximum lags
|
|
|
|
Automatic selection of lags based on SIC: 0 to 1
|
|
|
Total number of observations: 735
|
|
|
|
Cross-sections included: 11 (1 dropped)
|
|
|
|
Method
|
Statistic
|
|
Prob.**
|
|
ADF - Fisher Chi-square
|
248.869
|
|
0.0000
|
|
ADF - Choi Z-stat
|
-13.0593
|
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an asympotic
Chi
|
|
|
-square distribution. All other tests assume asymptotic
|
|
|
normality.
|
|
|
|
|
Intermediate ADF test results UNTITLED
|
|
|
|
Series
|
Prob.
|
Lag
|
Max Lag
|
|
Obs
|
|
TA
|
0.0000
|
0
|
10
|
|
67
|
|
RTCSTD
|
0.0000
|
0
|
10
|
|
67
|
|
RRSA
|
0.0237
|
1
|
10
|
|
66
|
|
ROE
|
0.0013
|
0
|
10
|
|
67
|
|
ROA
|
0.0035
|
0
|
10
|
|
67
|
|
FPSTA
|
0.0000
|
0
|
10
|
|
67
|
|
FPN
|
0.0000
|
0
|
10
|
|
67
|
|
ETI
|
0.2974
|
1
|
10
|
|
66
|
|
DPTD
|
0.0001
|
0
|
10
|
|
67
|
|
DBANC
|
0.0000
|
0
|
10
|
|
67
|
|
CBANC
|
0.0000
|
0
|
10
|
|
67
|
|
-square distribution. All other tests assume asymptotic
|
|
|
normality.
|
|
|
|
|
Intermediate ADF test results UNTITLED
|
|
|
|
Series
|
Prob.
|
Lag
|
Max Lag
|
|
Obs
|
|
TA
|
0.0535
|
1
|
10
|
|
66
|
|
RTCSTD
|
0.0000
|
0
|
10
|
|
67
|
|
RRSA
|
0.0401
|
1
|
10
|
|
66
|
|
ROE
|
0.0004
|
0
|
10
|
|
67
|
|
ROA
|
0.0010
|
0
|
10
|
|
67
|
|
FPSTA
|
0.0000
|
0
|
10
|
|
67
|
|
FPN
|
0.0000
|
0
|
10
|
|
67
|
|
ETI
|
0.3694
|
1
|
10
|
|
66
|
|
DPTD
|
0.0000
|
0
|
10
|
|
67
|
|
DBANC
|
0.9993
|
6
|
10
|
|
61
|
|
CBANC
|
0.9801
|
6
|
10
|
|
61
|
79
Tableau 2 : test de stationnarité de Phillips
Perron
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
|
|
Date: 04/27/13 Time: 09:37
|
|
|
|
|
|
Sample: 1 68
|
|
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA,
|
|
|
|
|
FPSTA, FPN, ETI, DPTD, DBANC, CBANC
|
|
|
|
|
Exogenous variables: None
|
|
|
|
|
Newey-West bandwidth selection using Bartlett kernel
|
|
|
|
|
Total (balanced) observations: 737
|
|
|
|
|
Cross-sections included: 11 (1 dropped)
|
|
|
|
|
Method
|
Statistic
|
|
|
Prob.**
|
|
PP - Fisher Chi-square
|
163.411
|
|
|
0.0000
|
|
PP - Choi Z-stat
|
-7.77469
|
|
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an
|
|
|
|
|
asympotic Chi-square distribution. All other tests
|
|
|
|
|
assume asymptotic normality.
|
|
|
|
|
Intermediate Phillips-Perron test results UNTITLED
|
|
|
|
|
Series
|
Prob.
|
Bandwidth
|
|
|
Obs
|
|
TA
|
0.0678
|
|
3.0
|
|
|
67
|
|
RTCSTD
|
0.1077
|
|
11.0
|
|
|
67
|
|
RRSA
|
0.0283
|
|
9.0
|
|
|
67
|
|
ROE
|
0.0000
|
|
2.0
|
|
|
67
|
|
ROA
|
0.0000
|
|
3.0
|
|
|
67
|
|
FPSTA
|
0.0000
|
|
4.0
|
|
|
67
|
|
FPN
|
0.0000
|
|
4.0
|
|
|
67
|
|
ETI
|
0.5973
|
|
2.0
|
|
|
67
|
|
DPTD
|
0.0005
|
|
5.0
|
|
|
67
|
|
DBANC
|
0.0139
|
|
13.0
|
|
|
67
|
|
CBANC
|
0.0249
|
|
0.0
|
|
|
67
|
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
Date: 04/27/13 Time: 09:41
|
|
|
|
Sample: 1 68
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA,
|
|
|
FPSTA, FPN, ETI, DPTD, DBANC, CBANC
|
|
|
Exogenous variables: Individual effects
|
|
|
Newey-West bandwidth selection using Bartlett kernel
|
|
|
Total (balanced) observations: 737
|
|
|
Cross-sections included: 11 (1 dropped)
|
|
|
Method
|
Statistic
|
|
Prob.**
|
|
PP - Fisher Chi-square
|
228.630
|
|
0.0000
|
|
PP - Choi Z-stat
|
-11.7321
|
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an
|
|
|
asympotic Chi-square distribution. All other tests
|
|
|
assume asymptotic normality.
|
|
|
Intermediate Phillips-Perron test results UNTITLED
|
|
|
Series
|
Prob.
|
Bandwidth
|
Obs
|
|
TA
|
0.0002
|
|
4.0
|
67
|
80
|
RTCSTD
|
0.0000
|
2.0
|
67
|
|
RRSA
|
0.0916
|
1.0
|
67
|
|
ROE
|
0.0004
|
2.0
|
67
|
|
ROA
|
0.0009
|
3.0
|
67
|
|
FPSTA
|
0.0000
|
0.0
|
67
|
|
FPN
|
0.0000
|
2.0
|
67
|
|
ETI
|
0.3164
|
2.0
|
67
|
|
DPTD
|
0.0000
|
4.0
|
67
|
|
DBANC
|
0.0590
|
1.0
|
67
|
|
CBANC
|
0.0000
|
4.0
|
67
|
|
Null Hypothesis: Unit root (individual unit root process)
|
|
Date: 04/27/13 Time: 09:42
|
|
|
Sample: 1 68
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA,
|
|
FPSTA, FPN, ETI, DPTD, DBANC, CBANC
|
|
Exogenous variables: Individual effects, individual linear
|
|
Trends
|
|
|
|
Newey-West bandwidth selection using Bartlett kernel
|
|
Total (balanced) observations: 737
|
|
Cross-sections included: 11
|
|
Method
|
Statistic
|
Prob.**
|
|
PP - Fisher Chi-square
|
251.193
|
0.0000
|
|
PP - Choi Z-stat
|
-13.3844
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an
|
|
asympotic Chi-square distribution. All other tests
|
|
assume asymptotic normality.
|
|
Intermediate Phillips-Perron test results UNTITLED
|
|
Series
|
Prob.
|
Bandwidth
|
Obs
|
|
TA
|
0.0000
|
4.0
|
67
|
|
RTCSTD
|
0.0000
|
2.0
|
67
|
|
RRSA
|
0.0000
|
3.0
|
67
|
|
ROE
|
0.0011
|
2.0
|
67
|
|
ROA
|
0.0028
|
3.0
|
67
|
|
FPSTA
|
0.0000
|
0.0
|
67
|
|
FPN
|
0.0000
|
17.0
|
67
|
|
ETI
|
0.2526
|
1.0
|
67
|
|
DPTD
|
0.0001
|
5.0
|
67
|
|
DBANC
|
0.0000
|
3.0
|
67
|
|
CBANC
|
0.0000
|
2.0
|
67
|
81
Tableau 3 : TEST DE DFA EN DIFFERENCE
PREMIERE
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
|
Date: 05/01/13 Time: 07:38
|
|
|
|
|
|
Sample: 1 68
|
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA, FPSTA, FPN,
|
|
|
|
ETI, DPTD, DBANC, CBANC
|
|
|
|
|
|
Exogenous variables: None
|
|
|
|
|
|
Automatic selection of maximum lags
|
|
|
|
|
Automatic selection of lags based on SIC: 0 to 5
|
|
|
|
Total number of observations: 715
|
|
|
|
|
Cross-sections included: 11
|
|
|
|
|
Method
|
Statistic
|
|
|
Prob.**
|
|
ADF - Fisher Chi-square
|
2421.85
|
|
|
0.0000
|
|
ADF - Choi Z-stat
|
-47.1423
|
|
|
0.0000
|
|
** Probabilities for Fisher tests are computed using an asympotic
Chi
|
|
|
|
-square distribution. All other tests assume asymptotic
|
|
|
|
normality.
|
|
|
|
|
|
Intermediate ADF test results D(UNTITLED)
|
|
|
|
|
Series
|
Prob.
|
Lag
|
Max Lag
|
|
Obs
|
|
D(TA)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(RTCSTD)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(RRSA)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(ROE)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(ROA)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(FPSTA)
|
0.0000
|
1
|
|
10
|
|
65
|
|
D(FPN)
|
0.0000
|
5
|
|
10
|
|
61
|
|
D(ETI)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(DPTD)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(DBANC)
|
0.0000
|
0
|
|
10
|
|
66
|
|
D(CBANC)
|
0.0000
|
5
|
|
10
|
|
61
|
Tableau 4 : TEST DE Phillips Perron EN DIFFERENCE
PREMIERE
|
Null Hypothesis: Unit root (individual unit root process)
|
|
|
Date: 05/01/13 Time: 07:40
|
|
|
|
Sample: 1 68
|
|
|
|
|
Series: TA, RTCSTD, RRSA, ROE, ROA,
|
|
|
FPSTA, FPN, ETI, DPTD, DBANC, CBANC
|
|
|
Exogenous variables: None
|
|
|
|
Newey-West bandwidth selection using Bartlett kernel
|
|
|
Total (balanced) observations: 726
|
|
|
Cross-sections included: 11
|
|
|
Method
|
Statistic
|
Prob.**
|
|
|
PP - Fisher Chi-square
|
2897.30
|
0.0000
|
|
|
PP - Choi Z-stat
|
-53.0660
|
0.0000
|
|
|
** Probabilities for Fisher tests are computed using an
|
|
|
asympotic Chi-square distribution. All other tests
|
|
|
assume asymptotic normality.
|
|
|
Intermediate Phillips-Perron test results D(UNTITLED)
|
|
|
Series
|
Prob.
|
Bandwidth
|
|
Obs
|
|
D(TA)
|
0.0000
|
19.0
|
|
66
|
82
|
D(RTCSTD)
|
0.0000
|
12.0
|
66
|
|
D(RRSA)
|
0.0000
|
10.0
|
66
|
|
D(ROE)
|
0.0000
|
5.0
|
66
|
|
D(ROA)
|
0.0000
|
4.0
|
66
|
|
D(FPSTA)
|
0.0000
|
47.0
|
66
|
|
D(FPN)
|
0.0000
|
65.0
|
66
|
|
D(ETI)
|
0.0000
|
2.0
|
66
|
|
D(DPTD)
|
0.0000
|
38.0
|
66
|
|
D(DBANC)
|
0.0000
|
16.0
|
66
|
|
D(CBANC)
|
0.0000
|
65.0
|
66
|
Annexe 2 : résultats des régressions
par les MCO et des tests d'autocorrélation et
d'hétéroscédasticité
Tableau 1 : résultats des
régressions par les MCO, des tests d'autocorrélation et
d'hétéroscédasticité et la régression par
les MCG du modèle ROE.
|
Equation de ROE
|
df MS
8 1474.6534
59 121.141
67 282.75442
|
|
Number of obs = 68
F( 8, 59) = 12.17
Prob > F = 0.0000
R-squared = 0.6227
Adj R-squared = 0.5716
Root MSE = 11.006
|
|
Regression par les MCO
|
|
Source | SS
+
Model | 11797.2272
Residual | 7147.319
+
Total | 18944.5462
|
|
roe | Coef.
|
Std. Err.
|
t
|
P>|t|
|
[95% Conf.
|
Interval]
|
|
fpsta |
|
.3291626
|
.3588413
|
0.92
|
0.363
|
-.3888772
|
1.047203
|
|
rrsa |
|
2.294563
|
.8906147
|
2.58
|
0.013
|
.5124471
|
4.076679
|
|
ta |
|
.0000238
|
.0000236
|
1.01
|
0.318
|
-.0000235
|
.0000711
|
|
dptd |
|
.0700445
|
.1692866
|
0.41
|
0.681
|
-.2686971
|
.4087861
|
|
fpn |
|
.0004475
|
.0001753
|
2.55
|
0.013
|
.0000968
|
.0007982
|
|
rtcstd |
|
-126.1769
|
75.76681
|
-1.67
|
0.101
|
-277.7859
|
25.43218
|
|
cbanc |
|
-.0001542
|
.0000866
|
-1.78
|
0.080
|
-.0003275
|
.0000191
|
|
dbanc |
|
-.0000273
|
.0000465
|
-0.59
|
0.559
|
-.0001203
|
.0000657
|
|
_ cons | 17.36273
|
13.66144
|
1.27
|
0.209
|
-9.973761
|
44.69921
|
Test d'heteroscedasticité
|
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Constant variance
|
|
Variables:
|
fitted values of roe
|
|
chi2(1)
|
=
|
19.09
|
|
Prob > chi2
|
=
|
0.0000
|
Test d'heteroscedasticité à travers le VIF
|
Variable
|
|
|
VIF
|
1/VIF
|
|
+
|
|
|
|
|
cbanc
|
|
|
60.38
|
0.016561
|
|
dbanc
|
|
|
32.87
|
0.030427
|
|
fpn
|
|
|
24.92
|
0.040120
|
|
ta |
|
|
16.63
|
0.060120
|
|
rrsa
|
|
|
6.35
|
0.157357
|
|
rtcstd
|
|
|
4.15
|
0.241239
|
|
dptd
|
|
|
1.82
|
0.549482
|
|
fpsta
|
|
|
1.15
|
0.870648
|
|
+
|
|
|
|
|
Mean VIF
|
|
|
18.53
|
|
Test d'heteroscedasticité conditionnelle
LM test for autoregressive conditional heteroskedasticity
(ARCH)
lags(p) | chi2 df Prob > chi2
+
1 | 38.981 1 0.0000
H0: no ARCH effects vs. H1: ARCH(p) disturbance
Test d'autocorrelation de Breusch godfrey
83
84
Breusch-Godfrey LM test for autocorrelation
lags(p) | chi2 df Prob > chi2
+
1 | 47.769 1 0.0000
H0: no serial correlation
Test d'autocorrelation de durbin Watson
Durbin's alternative test for autocorrelation
lags(p) | chi2 df Prob > chi2
+
1 | 136.947 1 0.0000
H0: no serial correlation
Determination de la valeur de AIC
Model | Obs ll(null) ll(model) df AIC BIC
+
. | 68 -287.8998 -254.7573 9 527.5146 547.4902
Note: N=Obs used in calculating BIC; see [R] BIC note
Correction de l'heteroscedasticité par la procedure de white
Linear regression Number of obs = 68
F( 8, 59) = 12.57
Prob > F = 0.0000
R-squared = 0.6227
Root MSE = 11.006
|
|
roe |
fpsta | rrsa |
ta |
|
+
Robust
Coef. Std. Err. t
.3291626 .1498602 2.20
2.294563 .9915612 2.31
.0000238 .0000135 1.77
|
P>|t| [95% Conf.
0.032 .029293
0.024 .3104536
0.082 -3.14e-06
|
Interval]
.6290323
4.278672
.0000508
|
|
dptd |
|
.0700445 .3119628 0.22
|
0.823 -.5541917
|
|
.6942806
|
|
fpn |
|
.0004475 .0001957 2.29
|
0.026 .0000559
|
|
.0008391
|
|
rtcstd |
|
-126.1769 76.12553 -1.66
|
0.103 -278.5037
|
|
26.14999
|
|
cbanc |
|
-.0001542 .0001275 -1.21
|
0.231 -.0004093
|
|
.0001009
|
|
dbanc |
|
-.0000273 .00004 -0.68
|
0.498 -.0001074
|
|
.0000528
|
|
_ cons | 17.36273 11.40456 1.52
|
0.133 -5.457752
|
|
40.18321
|
|
Estimation par la method moindres carrées
generalisées
|
|
|
|
Iteration 0: log likelihood = -254.75731
|
|
|
|
|
Generalized linear models
|
No. of obs
|
=
|
68
|
|
Optimization : ML
|
Residual df
|
=
|
59
|
|
Scale parameter
|
=
|
121.141
|
|
Deviance = 7147.319001
|
(1/df) Deviance
|
=
|
121.141
|
|
Pearson = 7147.319001
|
(1/df) Pearson
|
=
|
121.141
|
85
Variance function: V(u) = 1 [Gaussian]
Link function : g(u) = u [Identity]
AIC = 7.757568
Log likelihood = -254.7573067 BIC = 6898.368
| OIM
roe | Coef. Std. Err. z P>|z| [95% Conf. Interval]
+
fpsta | .3291626 .3588413 0.92 0.359 -.3741535 1.032479
rrsa | 2.294563 .8906147 2.58 0.010 .5489903 4.040136
ta | .0000238 .0000236 1.01 0.314 -.0000225 .0000701
dptd | .0700445 .1692866 0.41 0.679 -.2617511 .40184
fpn | .0004475 .0001753 2.55 0.011 .000104 .000791
rtcstd | -126.1769 75.76681 -1.67 0.096 -274.6771 22.32336
cbanc | -.0001542 .0000866 -1.78 0.075 -.0003239 .0000156
dbanc | -.0000273 .0000465 -0.59 0.557 -.0001184 .0000638
cons | 17.36273 13.66144 1.27 0.204 -9.413213 44.13867
_
Test de normalité des residus
Shapiro-Wilk W test for normal data
Variable | Obs W V z Prob>z
+
resid | 68 0.99270 0.439 -1.787 0.96300
86
Annexe 3: résultats des
régressions par les MCO, des tests d'autocorrélation et
d'hétéroscédasticité et la régression par
les MCG du modèle ROA.
|
Equation de ROA
|
df MS
|
|
Number of obs
F( 8, 59)
|
= 68
= 12.09
|
|
Regression par les MCO
|
|
Source |
+
|
SS
|
|
Model |
|
18.5175555
|
8 2.31469444
|
|
Prob > F
|
= 0.0000
|
|
Residual |
|
11.296186
|
59 .19146078
|
|
R-squared
|
= 0.6211
|
|
+
|
|
|
|
Adj R-squared
|
= 0.5697
|
|
Total |
|
29.8137416
|
67 .444981218
|
|
Root MSE
|
= .43756
|
|
roa |
|
Coef.
|
Std. Err. t
|
P>|t|
|
[95% Conf.
|
Interval]
|
|
+
fpsta |
|
.012863
|
.0142658 0.90
|
0.371
|
-.0156828
|
.0414089
|
|
rrsa |
|
.1006643
|
.0354066 2.84
|
0.006
|
.0298159
|
.1715128
|
|
ta |
|
-1.15e-07
|
9.39e-07 -0.12
|
0.903
|
-1.99e-06 1.76e-06
|
|
|
dptd |
|
.002146
|
.00673 0.32
|
0.751
|
-.0113208
|
.0156127
|
|
fpn |
|
.000012
|
6.97e-06 1.73
|
0.090
|
-1.92e-06
|
.000026
|
|
rtcstd |
|
-9.129399
|
3.012127 -3.03
|
0.004
|
-15.15665
|
-3.102146
|
|
cbanc |
|
-3.21e-06
|
3.44e-06 -0.93
|
0.356
|
-.0000101
|
3.68e-06
|
|
dbanc |
|
-1.55e-06
|
1.85e-06 -0.84
|
0.404
|
-5.25e-06
|
2.14e-06
|
|
_cons |
|
1.58716
|
.543114 2.92
|
0.005
|
.5003917
|
2.673929
|
Test d'heteroscedasticité
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Constant variance
Variables: fitted values of roa
chi2(1) = 13.37
Prob > chi2 = 0.0003
Test d'heteroscedasticité à travers le VIF
|
Variable
|
|
|
VIF
|
1/VIF
|
|
cbanc
|
|
|
60.38
|
0.016561
|
|
dbanc
|
|
|
32.87
|
0.030427
|
|
fpn
|
|
|
24.92
|
0.040120
|
|
ta |
|
|
16.63
|
0.060120
|
|
rrsa
|
|
|
6.35
|
0.157357
|
|
rtcstd |
|
4.15
|
0.241239
|
|
dptd |
|
1.82
|
0.549482
|
|
fpsta |
|
1.15
|
0.870648
|
|
+
|
|
|
|
Mean VIF |
|
18.53
|
|
Test d'heteroscedasticité conditionnelle
LM test for autoregressive conditional heteroskedasticity
(ARCH)
lags(p) | chi2 df Prob > chi2
+
1 | 40.732 1 0.0000
H0: no ARCH effects vs. H1: ARCH(p) disturbance
Test d'autocorrelation de Breusch godfrey
Breusch-Godfrey LM test for autocorrelation
lags(p) | chi2 df Prob > chi2
+
1 | 52.830 1 0.0000
H0: no serial correlation
Test d'autocorrelation de durbin Watson
Durbin's alternative test for autocorrelation
lags(p) | chi2 df Prob > chi2
+
1 | 201.978 1 0.0000
H0: no serial correlation
Determination de la valeur de AIC
Model | Obs ll(null) ll(model) df AIC BIC
+
. | 68 -287.8998 -254.7573 9 527.5146 547.4902
Note: N=Obs used in calculating BIC; see [R] BIC note
|
=
|
68
|
|
=
|
13.26
|
|
=
|
0.0000
|
|
=
|
0.6211
|
|
=
|
.43756
|
87
Correction de l'heteroscedasticité par la procedure de
white
Linear regression Number of obs
F( 8, 59) Prob > F R-squared Root MSE
88
|
roa
|
|
|
|
Coef.
|
Robust
Std. Err.
|
t
|
P>|t|
|
[95% Conf.
|
Interval]
|
|
fpsta
|
|
|
.012863
|
.0056025
|
2.30
|
0.025
|
.0016524
|
.0240737
|
|
rrsa
|
|
|
.1006643
|
.0371351
|
2.71
|
0.009
|
.0263571
|
.1749716
|
|
ta |
|
-1.15e-07
|
5.05e-07
|
-0.23
|
0.820
|
-1.13e-06
|
8.95e-07
|
|
dptd
|
|
|
.002146
|
.0118866
|
0.18
|
0.857
|
-.0216391
|
.0259311
|
|
fpn
|
|
|
.000012
|
7.57e-06
|
1.59
|
0.117
|
-3.11e-06
|
.0000272
|
|
rtcstd
|
|
|
-9.129399
|
2.750259
|
-3.32
|
0.002
|
-14.63265
|
-3.626143
|
|
cbanc
|
|
|
-3.21e-06
|
4.85e-06
|
-0.66
|
0.511
|
-.0000129
|
6.50e-06
|
|
dbanc
|
|
|
-1.55e-06
|
1.65e-06
|
-0.94
|
0.351
|
-4.86e-06
|
1.75e-06
|
|
_cons
|
|
|
1.58716
|
.4185245
|
3.79
|
0.000
|
.7496946
|
2.424626
|
Estimation par la method moindres carrées
generalisées
|
Iteration 0: log likelihood = -35.456373
Generalized linear models
Optimization : ML
|
No. of obs =
Residual df =
|
68
59
|
|
|
|
|
|
Scale parameter =
|
.1914608
|
|
Deviance
|
|
= 11.29618604
|
|
(1/df) Deviance =
|
.1914608
|
|
Pearson
|
|
= 11.29618604
|
|
(1/df) Pearson =
|
.1914608
|
|
Variance function: V(u) =
|
1
|
|
[Gaussian]
|
|
|
Link function
|
: g(u) =
|
u
|
|
[Identity]
|
|
|
|
|
|
AIC =
|
1.30754
|
|
Log likelihood
|
= -35.4563734
|
|
BIC =
|
-237.6548
|
|
|
|
|
OIM
|
|
|
|
|
roa
|
|
|
Coef.
|
Std. Err.
|
z
|
P>|z| [95% Conf.
|
Interval]
|
|
fpsta
|
|
|
.012863
|
.0142658
|
0.90
|
0.367 -.0150975
|
.0408235
|
|
rrsa
|
|
|
.1006643
|
.0354066
|
2.84
|
0.004 .0312687
|
.17006
|
|
ta |
|
|
-1.15e-07
|
9.39e-07
|
-0.12
|
0.902 -1.96e-06
|
1.73e-06
|
|
dptd
|
|
|
.002146
|
.00673
|
0.32
|
0.750 -.0110446
|
.0153366
|
|
fpn
|
|
|
.000012
|
6.97e-06
|
1.73
|
0.084 -1.63e-06
|
.0000257
|
|
rtcstd
|
|
|
-9.129399
|
3.012127
|
-3.03
|
0.002 -15.03306
|
-3.225738
|
|
cbanc
|
|
|
-3.21e-06
|
3.44e-06
|
-0.93
|
0.352 -9.95e-06
|
3.54e-06
|
|
dbanc
|
|
|
-1.55e-06
|
1.85e-06
|
-0.84
|
0.400 -5.17e-06
|
2.07e-06
|
|
_cons
|
|
|
1.58716
|
.543114
|
2.92
|
0.003 .5226764
|
2.651644
|
Test de normalité des residus
Shapiro-Wilk W test for normal data
Variable | Obs W V z Prob>z
+
.
resid2 | 68 0.98853 0.690 -0.807 0.79007
89
| 
