If I assume that $X(t)$ is a compound poisson process, how can it be found what $Cov(X(s), X(t))$ is? I have seen this over and over in books, but they only state it as fact. It is stated to be $\lambda E[Y]min(s,t)$, but I cannot quite seem to get the computation down. Many thanks if someone could guide me in the right direction.

  • 1
    $\begingroup$ What is $Y$? If it refers to the distribution of the size of the jumps, then your formula should involve the expectation of $Y^2$. After all, for $s=t$ it must reduce to the variance of $X(t)$, which is $\lambda t E[Y^2]$. $\endgroup$ – whuber Apr 16 '15 at 15:14

$Cov(X(s),X(t))$ are overlapping in timeline...

So now, we need to get it non-overlapping in timeline.

For $s<t$, $Cov(X(s),X(t))=Cov(X(s),X(t)-X(s)+X(s))=Cov(X(s),X(t)-X(s))+Cov(X(s),X(s))=\lambda{s}E(Y^2)$

Now here, "$Cov(X(s),X(t)-X(s))+Cov(X(s),X(s))$" see first term is independent so equals zero, second term is variance at timeline 's'.

For $t<s$.... It will be $Cov(X(t),X(s))=Cov(X(t),X(s)-X(t)+X(t))=Cov(X(t),X(s)-X(t))+Cov(X(t),X(t))=\lambda{t}E(Y^2)$

Hence, consider $min(s,t)$

Note:-as @whuber's comment says y is not defined here, Y considered as size of individual output.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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