2. Analisis Model ARCHGARCH
2.1. Plot ACF dan PACF Residual Kuadrat ARIMA 0,1,1
Date: 080312 Time: 14:28 Sample: 2000M02 2011M12
Included observations: 143 Q-statistic
probabilities adjusted for 1 ARMA terms
Autocorrelation Partial Correlation AC
PAC Q-Stat
Prob .| |
.| | 1
0.245 0.245
8.7859 .|. |
|. | 2 -0.008 -0.073
8.7963 0.003
.|. | .|. |
3 0.041
0.065 9.0398
0.011 .|. |
|. | 4 -0.041 -0.074
9.2925 0.026
.|. | .|. |
5 -0.007 0.029
9.3005 0.054
.|. | .|. |
6 0.058
0.048 9.8027
0.081 .|. |
.|. | 7
0.029 0.008
9.9292 0.128
.|. | .|. |
8 -0.051 -0.063 10.325
0.171 .|. |
.|. | 9
0.000 0.028
10.325 0.243
.|. | .|. |
10 -0.021 -0.033
10.396 0.319
.|. | .|. |
11 -0.031 -0.007
10.546 0.394
.|. | .|. |
12 -0.019 -0.024
10.604 0.477
.|. | .|. |
13 0.012
0.025 10.628
0.561 .|. |
|. | 14
-0.065 -0.078 11.301
0.586 |. |
.|. | 15
-0.073 -0.035 12.175
0.592 .|. |
.|. | 16
-0.025 -0.011 12.281
0.658 .|. |
.|. | 17
-0.032 -0.014 12.445
0.713 .|. |
.|. | 18
-0.035 -0.029 12.643
0.760 .|. |
.|. | 19
-0.043 -0.039 12.948
0.795 .|. |
.|. | 20
-0.058 -0.040 13.519
0.811
2.2. Evaluasi Model ARCH-GACH GARCH 1
Dependent Variable: DLO Method: ML - ARCH Marquardt - Normal distribution
Date: 080312 Time: 14:25 Sample adjusted: 2000M02 2011M12
Included observations: 143 after adjustments Convergence achieved after 16 iterations
MA Backcast: 2000M01 Presample variance: backcast parameter = 0.7
GARCH = C3 + C4RESID-12 + C5GARCH-1 Variable
Coefficient Std. Error
z-Statistic Prob.
C 0.011636 0.005442
2.138218 0.0325
MA1 0.244200 0.102438
2.383871 0.0171
Variance Equation
C 0.001661 0.000434
3.825624 0.0001
RESID-12 0.664333 0.216595
3.067161 0.0022
GARCH-1 0.067585 0.091784
0.736349 0.4615
R-squared 0.096168 Mean dependent var
0.005349 Adjusted R-squared
0.069969 S.D. dependent var 0.068164
S.E. of regression 0.065736 Akaike info criterion
-2.771253 Sum squared resid
0.596326 Schwarz criterion -2.667657
Log likelihood 203.1446 Hannan-Quinn criter.
-2.729157 F-statistic
3.670791 Durbin-Watson stat 1.767035
ProbF-statistic 0.007127 Inverted MA Roots
-.24
ARCH 1
Dependent Variable: DLO Method: ML - ARCH Marquardt - Normal distribution
Date: 080312 Time: 14:25 Sample adjusted: 2000M02 2011M12
Included observations: 143 after adjustments Convergence achieved after 19 iterations
MA Backcast: 2000M01 Presample variance: backcast parameter = 0.7
GARCH = C3 + C4RESID-12 Variable
Coefficient Std. Error
z-Statistic Prob.
C 0.012148 0.005076
2.393280 0.0167
MA1 0.229005 0.070439
3.251089 0.0011
Variance Equation
C 0.001921 0.000404
4.751247 0.0000
RESID-12 0.688593 0.216864
3.175231 0.0015
R-squared 0.092045 Mean dependent var
0.005349 Adjusted R-squared
0.072449 S.D. dependent var 0.068164
S.E. of regression 0.065648 Akaike info criterion
-2.779447 Sum squared resid
0.599046 Schwarz criterion -2.696570
Log likelihood 202.7304 Hannan-Quinn criter.
-2.745769 F-statistic
4.697081 Durbin-Watson stat 1.735504
ProbF-statistic 0.003724 Inverted MA Roots
-.23
3. Estimasi Persamaan VAR 1
3.1. Uji Lag Optimal