I guess it is a stupid question with an easy answer, namely: no. But I have discussed with someone claiming it's possible, so I am confused.
Consider a vector of random variables $(X_1, X_2, ..., X_n)$, each coming from the same distribution (if anyone is familiar with the process time models - it's process time model A).
Is it possible that the autocovariance function equal variance? In other words, I was told to consider the situation that autocovariance is constant and equal to variance.
Best regards!
Hint: Assume that it is, and calculate the correlation coefficient for any two samples, say $X_s,X_t$.