Similarly to my last question, for which I obtained very interesting and useful answers, I would like to know if there has been any study regarding heteroskedasticity and time-frames of the returns.
As an example could it be that the lower the time frame (take the 5 minute returns) the less heteroskedastic are returns?