This is of course true and your remarks about energy credits are interesting (although nothing new for me, I had analyzed the Riot case for another purpose). But the decisions to invest or disinvest are taken all the time. And the halving event was an event that for sure all serious mining companies would have taken into account previous to the business decisions they took in, roughly, from late 2023 on.

I had for example expected that some mining companies would have simply given up after the halving. They might have estimated their efficiency in comparison to other companies already before the halving and then decided if they buy their energy credits/hardware or not.

My expectation was also not a 30% drop like some other forum users expected, but something closer to 10% The 2-3% drop has also to be compared with the Moore's Law-adjusted hashrate/diff growth, so it might be actually a bit deeper (4-5%, without having done an exhaustive calculation).

Good observation, and yes, that might have flattened the curver already in anticipation. However, what I expected was that more companies would have decided to delay their market exit to the halving date, which would have caused a more substantial dip.

I also agree with @hugeblack that a more substantial bear market might cause a substantial drop too. However, in the past these drops were not that catastrophic. Hashrate most of the time did not even lower but grow slower.