I am wondering if there are classical (or not so classical) examples of negative autocorrelation in a regression model.
To expalin the context : I am explaining OLS limitations, heterosckedasticity and autocorrelation for the moment, and I want to showcase as many examples as possible, but I have only encountered positive autocorrelation in my practical work.
And also, is there any particular theoretical results in case of such an autocorrelation, as the errors tend to "cancel each other" ?