Help us double down on what's working, instead of guessing. Takes 5 seconds, totally optional.
SPY on August 5, 2024 - Dealer Positioning During the Yen-Carry Unwind
On August 5, 2024 the yen-carry unwind sent the VIX into the 60s before the open and SPX closed -3.0%. Replayed from minute-level data: net GEX printed -$10.6B at 9:45 ET - roughly four times deeper than the worst day of the Covid crash - with dealers needing to trade ~21 million SPY shares per 1% move, and the 0DTE share of gamma doubling into the close. Every number is one API call.
The August 2024 vol spike is the modern reference event for one reason: it was a positioning crash, not a news crash. No recession printed, no default, no war headline - a crowded funding trade unwound and the hedging mechanics did the rest. Which makes it the perfect day to replay through a dealer-positioning lens, because the mechanics ARE the story. One call per timestamp:
Three things in this snapshot deserve a slow read:
Net GEX -$10.6B. For scale: at 15:30 ET on March 16, 2020 - the -12% day of the Covid crash - net GEX was -$2.8B. August 5 opened almost four times deeper into short-gamma territory. The mechanical read: dealers needed to trade about 20.7 million SPY shares against every 1% move, selling weakness, chasing strength.
The flip at 544, spot at 514. Spot stood 5.5% below the gamma flip - the entire local price range sat in the amplification regime. No stabilizing hedge flow anywhere nearby.
Vanna at +$126B. The one shock absorber. With positive VEX, the pre-open vol explosion mechanically pushed dealer deltas up - forced buying into the panic. This is the counterforce that "short gamma amplifies everything" narratives skip, and on August 5 it was enormous.
15:30 ET - The Close Approaches, Positioning Gets Worse
Price had stabilized - spot was actually up slightly from 9:45 - but positioning deteriorated: net GEX deepened to -$11.5B, and the required hedge flow grew to ~22.2M shares per 1%. Two details worth noticing:
The 0DTE share doubled intraday, from 7.4% of total gamma at 9:45 to 14.3% by 15:30 (-$787M to -$1.64B). Same-day options piled into the stress and made the last hour twitchier - a dynamic that simply did not exist in 2020, when daily expirations were not yet listed.
The vanna cushion shrank from +$126B to +$86B as vol came off its highs - the morning's mechanical buying support faded into the afternoon.
The End-of-Day Board
From the full summary replay at the close: VIX finished at 38.57, up 64.9% on the day (after printing into the 60s pre-open), VVIX at 173, the VIX term structure in backwardation (VIX9D 42.8 over VIX 38.6 over VIX3M 33.7), and the 1DTE 25-delta skew at 6.8 vol points. SPY closed at 517.42 with ATM IV at 32.6 against a 20-day realized of just 10.5 - a 22-point volatility risk premium, the options market refusing to believe the calm of the prior weeks would return. The put wall sat at 500 with 643,000 puts of open interest - the largest single-strike position on the board.
What This Day Teaches
Depth of negative GEX ≠ size of the price move. August 5 printed 4x deeper GEX than Covid's worst day, but closed -3% versus -12%. The difference was the +$126B vanna offset and the speed of the vol round-trip. GEX sets the amplification coefficient; the shock still has to arrive through it.
Positioning can worsen while price recovers. The -$10.6B to -$11.5B deepening during a stabilizing tape is why the following sessions stayed violent even as headlines faded.
0DTE is now a stress accelerant. The doubling of same-day gamma share into the close is a structural feature of every modern vol event - visible only at minute resolution.
Replay It Yourself
Timestamp
What you will see
2024-08-05T09:45:00
The -$10.6B opening print, flip 5.5% overhead
2024-08-05T15:30:00
Deeper GEX into a stabilizing close, 0DTE share doubled
2024-08-06T09:45:00
The morning after - how fast did the regime heal?
2024-07-31T15:30:00
Three sessions earlier: the calm before, for contrast
At 9:45 ET, SPY traded at 514.13 with the gamma flip at 544.03 (5.5% overhead), net GEX at -$10.6B (dealers short gamma, needing to trade ~20.7 million shares per 1% move), net DEX at -$135B, and a +$126B vanna offset from the vol spike. By 15:30 net GEX had deepened to -$11.5B and the 0DTE share of gamma had doubled from 7.4% to 14.3%. VIX closed at 38.57, up 64.9% on the day.
Because GEX measures the amplifier, not the shock. The +$126B positive vanna exposure meant the vol explosion itself forced dealer buying, cushioning the decline, and the underlying shock (a funding-trade unwind) faded within days rather than compounding like a pandemic. Deeply negative GEX made every move bigger than it otherwise would have been - it did not require a 2020-sized catastrophe to be flowing through it.
Call GET /v1/exposure/summary/SPY?at=2024-08-05T09:45:00 on historical.flashalpha.com with an Alpha-tier key - any minute from 09:30 to 16:00 ET works, and every other analytics endpoint (GEX by strike, 0DTE, max pain, vol surface, full stock summary) replays the same timestamps with the same response shapes as the live API.
August 5, 2024 is the cleanest specimen of a positioning event in the archive: the deepest short-gamma print of the modern era, a nine-figure vanna cushion doing exactly what the textbook says it should, and 0DTE gamma doubling into the close - all measurable, minute by minute, in the same JSON your production code already parses. The Covid replay shows what negative gamma does to a real catastrophe; this day shows what it does to a purely mechanical one. Both are one at= parameter away.