macrocephalopod

Some of those might be a partial answer? Most of them are either irrelevant, too small to matter, would act in the wrong direction, or flat out contradict each other. No attempt made to explain the mechanism or size of the effect, just screaming stupid buzzwords into the void.

The first question I’d ask is “are we actually seeing a surprising number of intraday reversals?” — it definitely feels like we are but I’d want to see some stats on that to see if there’s actually anything to explain or whether this could just be seeing patterns in noise.

Then I’d throw out the explanations that can’t possibly work eg negative gamma exposure, CTAs, risk parity, target vol funds... these would all act in the wrong direction (they are momentum-reinforcing) so can’t be a useful part of the explanation.

Then throw out the nonsense explanations (front running, “the algos” without any eleboration, “passive flows” without an explanation of why they would be counter-trend) or the irrelevant ones (eg charm flows given we are not near an opex)

That leaves us with one possible mechanical explanation, dealer hedging due to positive gamma exposure which *could* cause an intraday reversal although unlikely to cause one as big as what we saw on Friday. A partial answer, not the whole answer.

But look, we don’t need (nor should we expect) mechanical explanations for every market move! Sometimes the market moves just because a big buyer came along for reasons completely unrelated to whatever else was going on at the time. Sometimes it reverses because it looked ...

undervalued to someone and they stepped in to buy at a discount. Sometimes it moves for no goddamn reason whatsoever. If market moves were predictable for mechanical reasons there would be a lot more people keeping quiet and making money off it, not coming onto twitter...

and shouting about it to anyone who will listen. Searching for an explanation for every market move is a game for fools and charlatans. The best most of us can do is stay focused, stick to your process and keep searching for that 1% signal in the noise.