NVDA institutional options flow analysis — multi-leg block trades, dominant direction, and gamma analysis from the public options tape for August 7, 2026. Articles older than 60 days are public; sign in to read flow within the past month, upgrade to AIme Premium for today's unusual options trades without the delay.

NVDA Unusual Options Activity — 2026-08-07

Institutional flow on 2026-08-07

Multi-leg block trades, dominant direction, and gamma analysis

$227.7M8 trades
Deep-ITM Financing Spread

Trade Details

BUY$200 CALL2026-08-07$60.8MDeep-ITM Financing Spread
SELL$207.5 CALL2026-08-07$40.5MDeep-ITM Financing Spread
BUY$197.5 CALL2026-08-07$33.5MDeep-ITM Financing Spread
SELL$205 CALL2026-08-07$23.4MDeep-ITM Financing Spread
SELL$202.5 CALL2026-08-07$21.5MDeep-ITM Financing Spread
SELL$202.5 CALL2026-08-07$21.4MDeep-ITM Financing Spread
BUY$210 CALL2026-08-07$13.4MDeep-ITM Financing Spread
BUY$210 CALL2026-08-07$13.3MDeep-ITM Financing Spread

Full Analysis

🟢 NVDA — $227M of Call Trades on Expiry Day, and Not One of Them Has a View

NVIDIA Corporation builds accelerated computing and AI infrastructure silicon. Sector: Technology / Semiconductors. Market cap $5.39T — up 23.9% — with the stock at $222.52, up 1.61% today (StockAnalysis). Follow it on the NVIDIA ticker page.

🤝 Four Packages in Nine Minutes — All Expiring Today

Between 14:51 and 15:00, four separate two-leg packages crossed as multi-leg floor trades and auctions — negotiated, not swept. Every leg expires today.

TimeBuy/SellC/PExpirationStrikeSizeVolumeOI (prior)Option PricePremiumSpotOption Symbol
15:00:01BUYCALL2026-08-07$20027,06635,84942,925$22.45$60,763,170$222.32NVDA20260807C200
15:00:01SELLCALL2026-08-07$207.5027,06635,89439,626$14.95$40,463,670$222.32NVDA20260807C207.5
14:51:36BUYCALL2026-08-07$197.5013,50015,40625,159$24.80$33,480,000$222.28NVDA20260807C197.5
14:51:36SELLCALL2026-08-07$20513,50017,96423,378$17.30$23,355,000$222.28NVDA20260807C205
14:59:50SELLCALL2026-08-07$202.5010,81023,03525,275$19.90$21,511,900$222.14NVDA20260807C202.5
14:59:50BUYCALL2026-08-07$21010,81026,93353,372$12.40$13,404,400$222.14NVDA20260807C210
14:54:48SELLCALL2026-08-07$202.5010,79023,03525,275$19.83$21,396,570$223.30NVDA20260807C202.5
14:54:48BUYCALL2026-08-07$21010,79026,93353,372$12.33$13,304,070$223.30NVDA20260807C210

Gross premium: $227,678,780. Net: a $14,224,500 debit. Combined package delta: ≈−7,358 shares.

⭐ Check the Arithmetic — All Four Are Financing

Here is the number that decides this entire page. Every one of these pairs is $7.50 apart in strike, and every one nets to exactly $7.50.

PackageStrikesWidthPricesNet
15:00:01$200 / $207.50$7.50$22.45 − $14.95$7.50
14:51:36$197.50 / $205$7.50$24.80 − $17.30$7.50
14:59:50$202.50 / $210$7.50$19.90 − $12.40$7.50
14:54:48$202.50 / $210$7.50$19.83 − $12.33$7.50

Paying or collecting exactly the strike width is the signature of a financing trade. The maximum a spread can ever be worth at expiry is its width — so when the net equals the width, the position cannot make or lose meaningfully on direction.

The Greeks confirm it beyond argument. Every strike here — $197.50 through $210 — carries a delta between 0.9795 and 0.9922 against a stock at $222.52. These are not options in any meaningful sense; they are stock substitutes with hours left to live. Pair one long against one short at nearly identical deltas and the whole book of four packages nets to ≈−7,358 shares of delta.

That is on $227.7 million of gross premium. Roughly $1.6 million of stock-equivalent exposure from a quarter of a billion dollars of trading. These are capital and carry transactions executed in options, and they carry no opinion about NVIDIA whatsoever.

⚠️ And the Earnings Date Does Not Line Up Either

NVIDIA reports on August 26, 2026 — confirmed (StockAnalysis).

Every leg here expires August 7. None of these packages carries the print. Anyone reading $227M of NVIDIA call trading as earnings positioning has the calendar wrong by nineteen days.

⛔ RESOLVED AS UNRESOLVABLE — Confirmed August 10

We said there would be nothing for the next snapshot to resolve. There was nothing. We checked all six strikes against the August 10 snapshot anyway, and every one returned no data — the contracts expired on August 7 and no longer exist.

LegBaseline (Aug-7)Resolving (Aug-10)Print sizeDay volVerdict
$200 call (bought)42,925expired — no snapshot27,06636,186Permanently unprovable
$207.50 call (sold)39,626expired — no snapshot27,06636,102Permanently unprovable
$197.50 call (bought)25,159expired — no snapshot13,50015,465Permanently unprovable
$205 call (sold)23,378expired — no snapshot13,50018,169Permanently unprovable
$202.50 call (sold)25,275expired — no snapshot21,60023,099Permanently unprovable
$210 call (bought)53,372expired — no snapshot21,60028,180Permanently unprovable

Both routes to an answer were closed from the start: every leg traded below its existing open interest, so the size test was silent, and every leg expired the same afternoon, so the snapshot test never got a chance.

And it does not matter here. The conclusion in this article rests on the net premium matching the strike width — four times out of four, on deltas above 0.979. That arithmetic identifies a financing package without needing to know whether anything opened or closed. An open-versus-close verdict would not have changed a word of it.

🤓 What This Actually Means — Plain English

A deep in-the-money call is not really an option. When the strike sits far below the price — here $200 against $222.52 — there is essentially no chance it expires worthless. Its price is almost entirely the gap between strike and spot, and it moves dollar for dollar with the shares. Delta of 0.985 means exactly that.

So a spread built from two of them is not a bet on anything. You own one stock-like instrument and are short another $7.50 away. The payoff is bounded, known in advance, and independent of what NVIDIA does.

Why do institutions do this? To move capital, manage margin and balance-sheet usage, or capture small differences in financing rates. It is plumbing. It shows up on a flow scanner looking like enormous conviction because the premium numbers are genuinely enormous — $60.8M on a single leg here — but the economics are close to a wash.

The check that reveals it takes five seconds: compare the net premium to the strike width. Four for four today. When those numbers match, stop looking for a market view — there is not one.

📊 The Charts

One-Year Price Action

NVIDIA 1-year price and volume

NVIDIA is +23.1% over the past year and trades near the upper end of a $164.07–$236.54 range — about 6% below its high. Fiscal 2026 delivered revenue of $215.94B (+65.47%) and earnings of $120.07B (+64.75%) (StockAnalysis). At a PE of 33.54 against that growth, the multiple is not the obvious stretch the headline market cap suggests.

Gamma Support and Resistance

NVIDIA gamma exposure

Dealer gamma sits close around spot: support at $222.50, $220, $215, $210 and $200, resistance at $225 and $230. The stock is sitting right on the $222.50 support shelf.

Every strike traded today — $197.50 to $210 — falls below the current price and inside or under that support structure. That is consistent with deep in-the-money financing rather than any attempt to position for a move.

Implied Move

NVIDIA implied move

The chain prices ±2.19% by August 10 ($217.66–$227.40), ±6.70% by August 21 ($207.61–$237.45), ±13.18% by September 18 ($193.21–$251.85), and ±39.06% out to June 2027 ($135.61–$309.45).

Note that the August 21 range bottoms at $207.61 — meaning the market considers the $210 strike traded today reachable two weeks out, but these contracts expire tonight. With hours left, the chain's own weekly range of $217.66–$227.40 sits entirely above every strike in this article. None of them had a realistic chance of finishing out of the money, which is precisely why they trade as stock substitutes.

📅 Catalysts

  • ⭐ Earnings: August 26, 2026 — confirmed (StockAnalysis). After every expiry traded today. These packages do not carry it.
  • SpaceX said it would "exclusively" use NVIDIA chips for AI infrastructure (StockAnalysis) — with further AI-infrastructure partnerships including Zayo and AZIO AI.
  • Consensus Strong Buy, average target $302.83 (+36.09%) across 61 analysts (StockAnalysis). Deep coverage; this average carries weight.
  • Index context: NVIDIA is 7.37% of SPY — the second-largest holding (StockAnalysis) — and 21.70% of SMH. Its August 26 print is an index-level event, not just a single-stock one.

👥 Four Ways to Read This

🎲 The YOLO trader — nothing to chase, and that is the whole point. The largest NVIDIA options activity of the day is plumbing. If you want a view on the August 26 print, these expiries are useless to you — they died tonight.

📈 The swing trader — ignore this print entirely when reading NVIDIA sentiment. The useful levels are the gamma shelves: $222.50 support, $225–$230 resistance. The date that matters is August 26.

💰 The premium collector — worth understanding how much of the notional you see on any tape is financing rather than volatility. A quarter of a billion dollars here produced roughly $1.6M of directional exposure.

🌱 The beginner — today's single most valuable habit: compare the net premium to the strike width before you read anything into a two-strike trade. It was $7.50 on $7.50 four separate times. That one check turns "$227M of NVIDIA call buying" into "no view at all" — and it is the difference between a signal and a mirage.

⚠️ Honest Risk and Limits — What the Tape Cannot Prove

  • Per-leg buy/sell direction is not resolvable on a negotiated multi-leg package — the pricing engine allocates a net price across legs, so per-leg readings are unreliable. What is proven is the structural fact: every pair nets to exactly its strike width.
  • Open versus close can never be established. All eight legs traded below existing open interest and expire today; no future snapshot resolves them.
  • We cannot see what these financing packages are attached to. Something else almost certainly sits alongside them, invisible here.
  • "Financing" is an inference from the arithmetic and the deltas, not a stated fact. It is a strong inference — four for four on the width test, with deltas above 0.979 — but an inference.
  • ⛔ Open versus close will never be settled for any of these six legs. All expired August 7 with no next-day snapshot, and all traded below prior open interest. This was flagged on the day and the August 10 check confirms it.

Nothing here is investment advice.


Last updated: August 10, 2026 — ⏳ provisional flags checked against the August 10 OPRA open-interest snapshot. All six legs expired August 7 and are permanently unresolvable, exactly as flagged on the day; no narrative change.