Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

So, I have this function: $$ y_t=v_t+\rho v_{t-1}+\rho^2 v_{t-2}+\dots+\rho^{t-1}v_1+\rho^ty_0 $$

And I want to find the variance (and after that the covariance, but I should be able to do that..). I know that the variance is: $ Var[y_t]=\frac{var[v]}{1-\rho^2}=\frac{\sigma^2_v}{1-\rho^2} $

But somewhere I mess everything up : (

share|cite|improve this question
You used the F-word! Not that it bothers me much, but others here are more easily offended... (suggestion: replace it with mess). Also, try to fix your latex (you probably want to write v_{t-1} to get $v_{t-1}$). – t.b. Apr 4 '11 at 10:45
This is to second Theo's comments and to signal that the question has no simple answer unless one specifies the dependence structure of $(v_t)$. Is $(v_t)$ an independent sequence? Also you say you want to find the variance but you already know what the variance is? Maybe you want to understand how one can find the variance. – Did Apr 4 '11 at 10:57
@user9079: I tried to fix your displayed equation, but I don't know if the result is the intended result. To get a compound exponent/index, enclose the entire exponent/index in curly brackets. That is, to get $v_{t-2}$, type v_{t-2}, and not v_t_-_2. – Arturo Magidin Apr 4 '11 at 15:23
Oh, I am sorry! I am used to another version of LaTeX where you can use _ everytime. And about the swearing, I am even more sorry. I come from a culture where swearing is very common. To @Didier: You are spot on in every comment. This is what happens with too much coffee, too little IQ and too hard classes. Sorry. To @Arturo Magidin thank you for correcting it! You guys are nice! I guess this is why I usually do social science, if you don't know the answer, you just write something that sounds smart.. – Eirik Apr 6 '11 at 11:16
@Eirik What a piece of advertisement for social science... :-) What is your specific field? – Did Apr 6 '11 at 22:19

You probably assume that the sequence $(v_t)$ is i.i.d. and that $(v_t)$ and $y_0$ are independent. Write $s_t$ for the variance of $y_t$ and $u$ for the variance of $v_t$. The variance of an independent sum is the sum of the variances, hence $$ s_t=u+\rho^2u+\cdots+\rho^{2t-2}u+\rho^{2t}s_0=\frac{1-\rho^{2t}}{1-\rho^2}u+\rho^{2t}s_0, $$ where the second formula assumes that $|\rho|<1$. To get your formula, one can either assume that $t\to+\infty$, or that $(y_t)$ is stationary. In the first case, $\rho^{2t}\to0$ hence $s_t\to u/(1-\rho^2)$. In the second case, $s_t=s_0$ hence plugging this value of $s_0$ in the formula above, one gets $s_t=u/(1-\rho^2)$ for every $t$.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.