Deribit's Front-End Skew Repriced in Four Hours. Nov. 27 Didn't Follow.
Put premiums on the Nov. 19 and Nov. 20 expiries widened faster than Nov. 27 inside a single four-hour window on Deribit. That term structure split is the signal, not the headline vol print. Here's what the curve is actually pricing, and why buying protection now is a tax.
I check Deribit's term structure before I check spot. Old habit from the Zurich days. Tuesday it told me something price action didn't.
Over a four-hour window, the gap between put and call implied volatility widened faster on the Nov. 19 and Nov. 20 expiries than it did on Nov. 27. Same underlying. Same order book. Different urgency.
That's a divergence worth sitting with. It isn't about direction. It's about when.
The Four-Hour Tell
Start with mechanics. Skew is the spread between what the market charges for downside and what it charges for upside. Positive skew means puts cost more. That's ordinary in crypto. What matters is how fast the spread moves and where on the curve it moves first.
On a normal session the curve slopes. Front-dated options carry more implied vol because near-term uncertainty is concentrated. Roll out and vol decays. Event risk gets diluted across more days and more candles.
Tuesday didn't do that. The front end repriced and the back end didn't follow. Inside those four hours the Nov. 19 and Nov. 20 tenors added downside premium at a materially faster clip than Nov. 27, picking up roughly 200 basis points of extra put-call spread relative to the later date.
That's not a market forming a view on the next month. That's a market pricing a specific hour.
Fed minutes land inside that window. So the timing makes sense. Scheduled catalysts have a habit of pulling protection demand forward into the nearest liquid strike set, and Deribit is where that demand clears. The Nov. 19 and Nov. 20 contracts sit right on top of the event. Nov. 27 sits past it.
Which is exactly why the lag matters.
Four hours is a reactive window. It's not an allocation shift. It's not a fund quietly rotating a sleeve. Somebody with a book looked at a screen, looked at their delta, and paid up. Then somebody else did the same thing. Market makers absorbed that flow and repriced short-dated gamma to compensate. That repricing is a liquidity tax, not a forecast.
The data is unambiguous on that point. The near dates ran. The far date watched.
Why the Front End Blinks First
Here's the question that matters. If traders genuinely thought the next two weeks were dangerous, why didn't Nov. 27 steepen alongside them?
Because event risk decays. That's the whole answer.
A scheduled release is a known unknown. It has a timestamp. Once it passes, the uncertainty it carried is gone, and the vol that was priced for it collapses. Nobody wants to own protection that expires three days after the thing they were worried about already happened. So demand concentrates, the front end gaps wider, and the back end stays roughly where it was.
That's a term structure doing its job. It's also a warning about misreading it.
Front-end skew spikes are reflexes. Back-end skew moves are conviction. When the far dates start bidding, that's a portfolio manager deciding the regime has changed. When only the front dates bid, that's a desk managing a Tuesday. The two get conflated constantly and they shouldn't be.
According to on-chain flows, spot holders aren't the ones driving this. Exchange balances haven't moved in a way that suggests broad distribution. This is a derivatives book event sitting on top of a spot market that's been in consolidation. The aggregate open interest picture supports that read. Protection demand is showing up in the options market, not in the coins themselves.
So who wins here? Anyone who sold front-end vol before the repricing. They got paid the premium and the move happened after their fill. Market makers who were quick enough to lift their short-dated offers mid-window did fine too.
Who loses? The buyer who shows up now, after the spread has already widened. They're paying the tax that somebody else collected. And the spot holder with no options book who reads a headline about surging downside premiums and panic-hedges into the most expensive part of the curve.
History rhymes here. The front end almost always blinks first. It rarely tells you what happens next.
What I'd Actually Do
Nothing dramatic.
If you needed protection for the next seven days, you're late. That trade cleared Tuesday afternoon. Chasing it now means paying up for a position that's already been repriced against you. That's not risk management. That's buying insurance after the fire alarm.
If your horizon is a month or longer, the Nov. 27 expiry is where the market hasn't charged you yet. The relative cheapness on the back end is the actual opportunity in this print. You're buying the same underlying event exposure, just without the crowd's premium stacked on top.
And if you're a spot holder with no options mandate, the correct response is to do nothing. A 200 basis point skew move on two near-dated tenors doesn't change your thesis. It changes a market maker's inventory.
Here's my bigger concern. There's a persistent habit of treating options positioning as a directional oracle. It isn't. It's a price for uncertainty, and uncertainty has a maturity date. The Nov. 19 and Nov. 20 bid tells you people are nervous about a scheduled release. The Nov. 27 lag tells you they don't think the nervousness survives the week. Those are two different claims and the second one is the more interesting bet.
I'd rather be on the side of the curve that isn't crowded. If losses hold through the weekly close, the front-end buyers will look prescient for about 48 hours and then they'll be holding decayed premium with nothing left to hedge. That's the structural problem with buying reflex protection. You can be right on direction and still lose on the vol.
Not speculation. Arithmetic.