# I read that for a stationary process, Cov(Xt+h, Xt) should be a constant. What does it actually mean?

According to wiki, covariance is a measure of the joint variability of two random variables.

Isnt Xt+h one particular value and Xt another value.. So what does it mean when you say covariance between these two values?﻿

• In this context, $X_t$ is a random variable. For example, $X_t$ may denote whether a coin flip lands heads or tails at time $t$. Or $X_t$ may denote the closing value of the S&P500 index on day $t$. Typically a random variable $X_t$ is defined so that the value of $X_t$ is known at time $t$. Feb 11, 2018 at 14:49
• What, if anything, do you understand by covariance? What you ciaim to have read as per the title of your question, is incorrect, and it would help if you would let us know what misconceptions you have about the notion of covariance. Feb 11, 2018 at 14:52
• Maybe this thread will help: stats.stackexchange.com/questions/126791.
– whuber
Feb 11, 2018 at 17:31