User:Hassani.stat

Hassani's -1/2 Theorem

The Hassani's -1/2 Theorem indicates that the sum of sample autocorrelation function is always -1/2 for any stationary time series with arbitrary length.

Using Hassani's -1/2 Theorem, it can be shown that the distribution of a set of the sample autocorrelation estimates is not independent and identically distributed. This finding implies that the result of diagnostic check and model building using the traditional assumption of iid can be quite misleading.