悬赏 20 个论坛币 已解决
代码如下:
specs = ugarchspec(variance.model=list(model="fGARCH", garchOrder=c(1,1),
submodel = "TGARCH"),
mean.model=list(armaOrder=c(1,0), arfima=FALSE, include.mean=TRUE,
archm = FALSE, archpow = 1), distribution.model="ged")
garch<-ugarchfit(specs,data=rets[,1], solver.control = list(trace=0))
garch
输出结果:
*---------------------------------*
* GARCH Model Fit *
*---------------------------------*
Conditional Variance Dynamics
-----------------------------------
GARCH Model : fGARCH(1,1)
fGARCH Sub-Model : TGARCH
Mean Model : ARFIMA(1,0,0)
Distribution : ged
Optimal Parameters
------------------------------------
Estimate Std. Error t value Pr(>|t|)
mu 0.000917 0.000000 2197.0 0
ar1 0.276470 0.000126 2196.9 0
omega 0.000019 0.000000 2197.7 0
alpha1 0.050000 0.000021 2339.5 0
beta1 0.900000 0.000055 16512.5 0
eta11 0.050000 0.000007 6671.1 0
shape 2.000000 0.000415 4822.3 0
Robust Standard Errors:
Estimate Std. Error t value Pr(>|t|)
mu 0.000917 NaN NaN NaN
ar1 0.276470 NaN NaN NaN
omega 0.000019 NaN NaN NaN
alpha1 0.050000 NaN NaN NaN
beta1 0.900000 NaN NaN NaN
eta11 0.050000 NaN NaN NaN
shape 2.000000 NaN NaN NaN
LogLikelihood : 72.697
Information Criteria
------------------------------------
Akaike -0.047797
Bayes -0.032725
Shibata -0.047810
Hannan-Quinn -0.042351
这是哪里出错了呀?求大佬解答!
最佳答案
719812133 查看完整内容
你TGARCH的参数估计值很奇怪,连小数点后几位都没有,感觉和默认的初始值是一样的,同时你每一个参数估计值的t值都特别大,感觉不太合理,应该是参数估计出了问题导致这个现象,结果肯定是不能用的。上面Optimal Parameters的标准误是用hessian inverse estimator去求的,下面的Robust Standard Errors里的标准误是用sandwich estimator计算的,算不出来说明robust variance-covariance matrix主对角线上元素全是负的,矩阵不是正 ...