请问MA（1）下的dynamic forecast

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

请教大牛们一个可能stupid的问题哈
如果用了MA（1）来估计
arima return, ma(1)
然后利用1982q1之前的数据，对1982q1之后的数据做dynamic forecast
predict forecast1 if tin(1982q1,1986q4), dynamic(1982q1)
发现做出来的forecast1是一个常数，
这是否合理呢（是不是因为model是MA（1），所以dynamic forecast就是常数？）
还是我写的code有问题呢

提前谢过~

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：Forecast Dynamic forecas Dynam AMI dynamic return

There is no problem for a constant outcome if you are only using MA(1) and C in the fomular and choosing Dynamic in the forecast, the constant you see in the forecast should be equal to the C in your function. This is because Move Average is assuming the terms are all white noise, so their expectation in the future are 0.
But, if you choose Static in the forecast, it will give you different result, it is using the Theta multiply t-1 residual to estimate t time residual. The detail explaination can be found here:
http://forums.eviews.com/viewtopic.php?f=7&t=465