Garch迭代精度问题

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

我使用命令arch ex if id==9,ar(1) ma(1) arch(1) garch(1) difficult后995次迭代还是没有结果，是数据太平稳了吗？是否有什么办法得到结果呢？ 能否提前设定迭代的精度呢？多谢各位大侠!

Iteration 980: log likelihood = -37.569687  
BFGS stepping has contracted, resetting BFGS Hessian (98)
Iteration 981: log likelihood = -37.569664  (backed up)
Iteration 982: log likelihood = -37.569662  (backed up)
Iteration 983: log likelihood = -37.567699  (backed up)
Iteration 984: log likelihood = -37.567289  (backed up)
Iteration 985: log likelihood = -37.567289  (backed up)
Iteration 986: log likelihood = -37.567286  (backed up)
Iteration 987: log likelihood = -37.567286  (backed up)
Iteration 988: log likelihood = -37.567262  (backed up)
BFGS stepping has contracted, resetting BFGS Hessian (99)
Iteration 989: log likelihood = -37.566539  
(switching optimization to BHHH)
Iteration 990: log likelihood = -37.566531  (not concave)
Iteration 991: log likelihood = -37.566515  (not concave)
Iteration 992: log likelihood = -37.566513  (not concave)
Iteration 993: log likelihood = -37.566513  (not concave)
Iteration 994: log likelihood = -37.566513  (not concave)
(switching optimization to BFGS)
Iteration 995: log likelihood = -37.566513  
flat log likelihood encountered, cannot find uphill direction
r(430);

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：GARCH ARCH 迭代精度 ARC RCH difficult

建议用eviews软件进行garch模型的估计，用stata的运算速度很慢，而且需要设定合适的种子才会得出结果。很麻烦