+ Reply to Thread
Results 1 to 2 of 2

Thread: Time series expectation

  1. #1
    Points: 5,083, Level: 45
    Level completed: 67%, Points required for next Level: 67

    Thanked 3 Times in 3 Posts

    Time series expectation

    Maybe it's because i've been at it for 8 hours straight today, but i'm stumped trying to show this is true:

    Any help might spare my sanity.

  2. #2
    TS Contributor
    Points: 7,081, Level: 55
    Level completed: 66%, Points required for next Level: 69

    Copenhagen , Denmark
    Thanked 123 Times in 116 Posts

    Re: Time series expectation

    When you ask other people a question it is a good idea to provide them with the information needed for the question at hand to make sense....

    So one guess is that you have and AR(q) model...for simplicity assume it is AR(1) no constantterm:
    y_t = a_1 y_{t-1} + e_t and by definition \sigma^2 = var(e_t)
    then using the fact that E[y_t]=0 it follows that E[y_te_t]=cov(y_t,e_t)
    then E[y_te_t]=cov(y_t,e_t)=cov(a_1 y_{t-1} + e_t,e_t)

    =cov(a_1 y_{t-1} , e_t) + cov(e_t,e_t) = 0 + var(e_t) = \sigma^2

+ Reply to Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats