The author makes 2 fundamental mistakes. The first is small-sample error and the second is confusing correlation with causality. in every noise you can see some patterns, but this does not mean that they are permanent,
and will be repeated in the future and mean something more than fortune-telling from tea leaves or fingerprints and that's what we see here. In my opinion, there is no regularity here, only a temporary correlation, if at all.

CME volumens are too low compared to large crypto exchanges to matter.