想做GARCH模型,通过信息准则对股票指数收益率做了ARMA(1,1)模型后,检验ARCH 效应的时候无论用哪个检验方法都不存在ARCH效应该怎么办?
自相关和偏自相关图
Autocorrelation Partial Correlation AC PAC Q-Stat Prob
.| | .| | 1 0.025 0.025 0.5497 0.458
.| | .| | 2 -0.038 -0.039 1.8312 0.400
.| | .| | 3 0.004 0.006 1.8453 0.605
.| | .| | 4 -0.009 -0.011 1.9200 0.750
.| | .| | 5 0.024 0.025 2.4206 0.788
.| | .| | 6 -0.005 -0.007 2.4399 0.875
.| | .| | 7 0.065 0.068 6.2078 0.516
.| | .| | 8 -0.007 -0.012 6.2524 0.619
.| | .| | 9 -0.001 0.005 6.2541 0.714
.| | .| | 10 0.003 0.001 6.2620 0.793
.| | .| | 11 -0.022 -0.020 6.6871 0.824
.| | .| | 12 0.041 0.039 8.1635 0.772
.| | .| | 13 -0.001 -0.004 8.1654 0.833
.| | .| | 14 0.009 0.009 8.2454 0.876
.| | .| | 15 0.023 0.023 8.7228 0.892
Arma(1,1)模型结果
Variable Coefficient Std. Error t-Statistic Prob.
C -0.000871 0.000536 -1.625884 0.1043
AR(1) -0.768006 0.178474 -4.303176 0.0000
MA(1) 0.801852 0.166346 4.820391 0.0000
R-squared 0.003485 Mean dependent var -0.000874
Adjusted R-squared 0.001197 S.D. dependent var 0.015559
S.E. of regression 0.015550 Akaike info criterion -5.486083
Sum squared resid 0.210610 Schwarz criterion -5.469700
Log likelihood 2400.418 Hannan-Quinn criter. -5.479816
F-statistic 1.523044 Durbin-Watson stat 2.010287
Prob(F-statistic) 0.218627
LM检验
F-statistic 0.118137 Prob. F(2,869) 0.8886
Obs*R-squared 0.237566 Prob. Chi-Square(2) 0.8880
Test Equation:
Dependent Variable: RESID
Method: Least Squares
Date: 04/05/20 Time: 00:25
Sample: 2/22/2010 9/26/2013
Included observations: 874
Presample missing value lagged residuals set to zero.
Variable Coefficient Std. Error t-Statistic Prob.
C 6.80E-07 0.000537 0.001268 0.9990
AR(1) -0.019651 0.229786 -0.085518 0.9319
MA(1) 0.011618 0.196170 0.059222 0.9528
RESID(-1) 0.002260 0.058553 0.038604 0.9692
RESID(-2) -0.020029 0.047298 -0.423465 0.6721
R-squared 0.000272 Mean dependent var 7.96E-07
Adjusted R-squared -0.004330 S.D. dependent var 0.015532
S.E. of regression 0.015566 Akaike info criterion -5.481779
Sum squared resid 0.210553 Schwarz criterion -5.454473
Log likelihood 2400.537 Hannan-Quinn criter. -5.471333
F-statistic 0.059068 Durbin-Watson stat 1.999314
Prob(F-statistic) 0.993535
ARCH检验
Heteroskedasticity Test: ARCH
F-statistic 2.730239 Prob. F(1,871) 0.0988
Obs*R-squared 2.727957 Prob. Chi-Square(1) 0.0986
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 04/05/20 Time: 00:25
Sample (adjusted): 2/23/2010 9/26/2013
Included observations: 873 after adjustments
Variable Coefficient Std. Error t-Statistic Prob.
C 0.000228 1.85E-05 12.30889 0.0000
RESID^2(-1) 0.055896 0.033828 1.652344 0.0988
R-squared 0.003125 Mean dependent var 0.000241
Adjusted R-squared 0.001980 S.D. dependent var 0.000491
S.E. of regression 0.000491 Akaike info criterion -12.39930
Sum squared resid 0.000210 Schwarz criterion -12.38836
Log likelihood 5414.293 Hannan-Quinn criter. -12.39511
F-statistic 2.730239 Durbin-Watson stat 2.003296
Prob(F-statistic) 0.098825
残差自相关图
Q-statistic probabilities adjusted for 2 ARMA terms
Autocorrelation Partial Correlation AC PAC Q-Stat Prob
.| | .| | 1 -0.006 -0.006 0.0280
.| | .| | 2 -0.014 -0.014 0.2053
.| | .| | 3 -0.013 -0.013 0.3469 0.556
.| | .| | 4 0.003 0.003 0.3561 0.837
.| | .| | 5 0.014 0.014 0.5322 0.912
.| | .| | 6 0.002 0.002 0.5350 0.970
.| | .| | 7 0.058 0.059 3.5227 0.620
.| | .| | 8 -0.002 -0.001 3.5262 0.740
.| | .| | 9 -0.006 -0.005 3.5629 0.829
.| | .| | 10 0.008 0.009 3.6229 0.889
.| | .| | 11 -0.027 -0.028 4.2654 0.893
.| | .| | 12 0.045 0.044 6.1000 0.807
.| | .| | 13 -0.008 -0.008 6.1578 0.863
.| | .| | 14 0.016 0.014 6.3949 0.895
.| | .| | 15 0.014 0.016 6.5775 0.923