各位，我想用garch类方法估计企业每年的收益率条件方差，数据描述如下：大约有1000+企业以上，2004-2013年度上市企业的每日收益率。数据样本见附件
我用forvalues编写程序如下：

forvalues i = 1(1)1056{
bysort year: arch r if id==`i', arch(1) garch(1) nolog
predict h, variance
}
其中id表示企业代码，year表示不同年度。
可以运行，但是已经半天了还是没有结束，正在进行的程序显示如下：

-> year = 2004

Number of gaps in sample:29 (gap count includes panel changes)
(note: conditioning reset at each gap)


ARCH family regression

Sample: 20040628 - 20041231, but with gaps Number of obs = 129
Distribution: Gaussian Wald chi2(.) = .
Log likelihood =270.8619 Prob > chi2 = .

------------------------------------------------------------------------------
| OPG
r | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
r |
_cons |-.0045339 .0027907 -1.62 0.104 -.0100035 .0009357
-------------+----------------------------------------------------------------
ARCH |
arch |
L1. | .2879147 .1467449 1.96 0.050 .0003 .5755294
|
garch |
L1. |-.1687872 .2891155 -0.58 0.059 -.7354432 .3978688
|
_cons | .0008414 .0002682 3.14 0.002 .0003157 .001367
------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------------------------------------------------------
-> year = 2005

Number of gaps in sample:53 (gap count includes panel changes)
(note: conditioning reset at each gap)

。。。。。






请问我的代码有问题吗？应该怎么编写更好？
谢谢各位对我的帮助！！