The floor became a ceiling

We publish these at roughly the same hour each morning, which makes them easy to grade. Here's what every major did between yesterday's 10:26am print and last night's close:

Every index that was green at 10:26am closed red. That is what a 0.2× cushion looks like when it gets tested — QQQ had less than a quarter of a normal session's room above its flip, and the session used all of it before lunch. Nothing dramatic happened; there was no headline. The structure simply stopped holding price up, and price went where it was already leaning.

The cushion is how far spot sits from the zero-gamma flip, measured in that index's own expected move for the day. Above 2× is stable. Under 1× the tripwire is inside a single session. Below zero, there is no tripwire left to trip — you're already on the other side of it.

Drawn out — the solid line is zero (the flip itself), the grey mark is 1× expected move above it:

Look at DIA. Yesterday it was the one index whose cushion widened — 2.6× and the safest place to be long. Twenty-four hours later it's below its flip. That's a 3.3× swing in a single session, and it happened on a Dow that only fell 0.07%. The flip rose to meet price; price didn't have to fall to it.

Here's the part that most gamma commentary misses. SPY's net dealer gamma is roughly zero this morning — our three pulls read +$496M, +$123M and −$42M, which for a $700B ETF is indistinguishable from nothing. But "net zero" doesn't mean "no gamma." It means the positive and negative sides cancel. What matters is where each side sits relative to spot.

Positive dealer gamma above the market and nothing below it is not support — it's resistance. Dealers long gamma at 775+ sell into every rally that approaches it, which caps upside. And with no offsetting positive gamma underneath, a move down doesn't run into anyone mechanically obliged to buy. Same total number, completely inverted meaning. That's the whole regime in one sentence: the hedging flow that spent two weeks catching dips is now sitting overhead selling rips instead.

New to this? Start with what gamma exposure actually is →, or check any ticker yourself on the free gamma exposure map.

Wednesday's story was implied vol spiking 35–57% on a green tape — someone paying up for protection into strength. That protection got monetised on the reversal, and this morning the vol is gone:

Normally falling IV is a comfort signal. Today it's the opposite, and the reason is mechanical. Positive dealer gamma is what suppresses realised volatility — it's the reason a 16 IV was justified for the last two weeks. That suppression is now gone from three of four indices, but the options are still priced as though it's there. SPY is pricing a 0.59% day. It has delivered a bigger range than that on four of the last five sessions.

You don't have to believe a selloff is coming to see the mismatch. Cheap options into a structure that no longer damps movement is simply a better price for optionality than it was on Monday, regardless of direction.

In a broken index regime the index is the worst thing to trade — you're taking directional risk with no structural edge. The edge moves down to single names, because dealer positioning is name-by-name and plenty of stocks still have an intact cushion even when the index doesn't. So we scanned the 24 most liquid names on the board. Fourteen are still above their own flip. Seven are below it. Two have run past their call wall.

▲ Still cushioned — spot above the flip, room to the call wall. Dealers here still buy weakness for you. R:R = distance up to the wall ÷ distance down to the flip.

▼ Below their flip — dealers here amplify moves in both directions. These are not dip-buys. A chart that looks like support in this bucket behaves like a trapdoor.

■ Past the wall — QCOM at 159.64 is 2.9% above its 155 call wall and SMCI at 30.11 is through 30. Both are up today. Gamma caps names here; chasing green candles into a call wall is how a good read becomes a bad fill. PLTR is its own category — −3.16% at 153.53, third straight down day unwinding Monday's +29% earnings gap, with no clean flip anywhere in its chain.

One genuine contradiction worth flagging: SPY's intraday net options premium is +$7.7M this morning, having climbed steadily since the open. Yesterday at this hour it was −$5.3M. So the flow is bullish into a broken structure — the exact inverse of Wednesday, when flow was bearish into an intact one.

The single-name sweeps line up with the board rather than against it. The biggest directional bets on the tape this morning:

Two of those are worth sitting with. Somebody has now spent close to $1M over two sessions buying AMD calls struck at the reclaim level, and AMD has gone down both days. And the MSFT 500 calls are a clean, well-defined expression: MSFT is up 1.91% and still hasn't got back above 498.38. Reclaim it and the largest company in the index converts from amplifier to cushion. Fail it, and a 1.9% up day into a broken flip is exactly the kind of rally this regime is built to sell.

The one level worth writing on your screen: IWM 298.49. It's the last index flip on the right side of spot. If small caps lose it, all four majors are below their flip simultaneously — and the difference between "two of four" and "four of four" is the difference between a rotation and a de-risking.

A structure with no cushion doesn't need a big catalyst to produce a big move — that's the entire point. This is what's scheduled:

Note what isn't on that list: none of the seven cushioned names in Section 5 report inside 48 hours. That's deliberate — a gamma setup you can't hold through the night isn't a gamma setup, it's a coin flip. Run any of tonight's names through the free earnings analyzer to see whether the implied move is cheap or rich versus how the stock has actually reacted historically.

Something did break, and it broke on schedule. The Nasdaq's dealer gamma is negative, SPY's is net zero with all of the remaining mass sitting overhead as resistance rather than underneath as support, the Dow lost a 2.6× cushion in one session, and only small caps are still on the right side of a flip — by less than one day's range. The tape isn't crashing; it's up slightly and the flow is bullish. But the mechanism that has quietly kept every dip shallow for two weeks is switched off, and implied volatility got cheaper on the day it switched off. The trade isn't to get short. It's to stop trading the index, run every long through the "is it above its own flip" screen, and notice that optionality just went on sale into the least supportive structure of the month — one session before the jobs report.