Macroeconometrics
Module 5: Practice Problems
1. Consider the ARMA(1,1) process
yt = φ 1yt-1 + εt — θ 1εt-1 .
Use iterative backwards substitution to determine the conditions that φ 1 and θ 1 must satisfy in order for this process to be invertible. More specifically, what you should do here is firste move from ARMA to pure AR in order to establish conditions for φ 1 and then move from ARMA to pure MA in order to establish conditions for θ 1 .
2. Consider the ARMA(1,0), i.e., AR(1) process:
yt = φ 1yt-1 + εt.
where jφ1 j < 1 so that the process is invertible. Show that
E (yt ) = 0
and also that the variance of this process is constant and finite. In particular:
3. Consider the AR(p) process
yt = φ1yt-1 + ... + φpyt-p + εt ,
and assume invertibility. Show that E (yt ) = 0 and that the variance of yt is constant and finite.
4. Consider the process yt = α + βt + εt. Show that yt is not stationary.
5. Consider the process
yt = et . yt(φ)-1 . eη t ,
where ηt isa random walk process such that
ηt = ηt-1 + εt ,
where εt is white noise, and jφj < 1. Show that this process is not stationary by setting it up in logs (recall that for any variable x, ln (ex ) = x).
版权所有:留学生编程辅导网 2020 All Rights Reserved 联系方式:QQ:821613408 微信:horysk8 电子信箱:[email protected]
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。