🤝 NVDA $6.3M LEAP Call Cross — A Bet on the Rubin Era Through Jan 2028 (Direction Unproven)

📅 July 9, 2026 | 🔥 Unusual Activity Detected

✏️✅ Corrected & updated July 10, 2026: this article originally reported prior open interest of 100 contracts. The real figure was 35,052 — a scanner data error on our side. The next-day OPRA snapshot shows OI rising 35,052 → 36,460 (+1,408), meaning the trade was a partial net open (≈70% new contracts, ≈30% transfer), not the clean "size ≫ OI" fresh open we originally described. The Unusual Score has been revised down. Direction remains unproven. See the corrected resolution below.

---

🎯 The Quick Take

A desk just crossed 2,000 NVDA January 21, 2028 $250 calls at $31.45 this morning at 11:39:46 ET — about $6.3 million of premium, struck ≈24% above today's $201.47 spot. This printed as a block cross 🤝 (a known counterparty matched off the open book, not aggressive lit buying), and it landed just a hair below the ask ($31.60), which is only a weak lean toward "someone paid up." Translation: this is a large, multi-year, out-of-the-money call position getting opened today — but because it's a negotiated cross, we genuinely can't prove whether the buyer or the seller initiated it. What we can say: it's fresh (open interest was just 100 contracts before this), and it's a long runway bet spanning roughly six more earnings reports and NVIDIA's entire Rubin product cycle.

---

📊 Company Overview

NVIDIA Corporation (NVDA) is the dominant designer of GPUs that power the global AI build-out, plus the CUDA software platform and data-center networking (NVLink/InfiniBand/Spectrum-X) that ties it all together:

• Market Cap: ≈$4.9 trillion — the world's most valuable public company
• Industry: Semiconductors & Related Devices (Information Technology sector)
• Current Price: ≈$201-204 (trade printed at $201.47; latest close ≈$204.12)
• Primary Business: AI training/inference GPUs (Data Center segment, the overwhelming majority of revenue), gaming graphics, professional visualization, automotive/robotics
• What they do, plainly: if a hyperscaler is building an AI data center, NVIDIA's chips and networking gear are the default choice — Blackwell today, Rubin arriving later this year, Rubin Ultra in 2027.

---

💰 The Option Flow Breakdown

📊 What Just Happened

The Tape (July 9, 2026 @ 11:39:46 ET):

✏️ Correction (July 10, 2026): the original version of this table listed prior open interest as 100 contracts, sourced from our intraday scanner feed. That figure was wrong. The actual OPRA open-interest baseline for this contract going into July 9 was 35,052 — this is a heavily-populated LEAP strike, not an empty one. The correction materially changes the open/close read below, and we have rewritten that section accordingly. We regret the error.

• 💸 Premium: ≈$6.3M ($31.45 × 2,000 contracts × 100 shares)
• 🎯 Strike distance: $250 is ≈24% above the $201.47 spot at print time — a genuinely out-of-the-money strike
• ⏰ Expiration: January 21, 2028 — that's ≈561 days (≈1.5 years) from today. NVIDIA LEAP series like this are typically first listed roughly 2.5 years before they expire, so this contract has already "aged" — what matters now is the ≈1.5 years of runway actually left
• 🖨️ Print location: $31.45, just inside the $31.05 bid / $31.60 ask — close to the ask, but on a cross, the usual "where did it print in the spread" aggressor read does not apply (there's no lit order book being taken; a broker matched two sides off-market)

✅ RESOLVED — A Partial Open, and a Correction We Owe You (updated July 10, 2026)

First, the correction. Yesterday this section claimed "Size (2,000) is dramatically larger than the pre-trade open interest (100 contracts) — that part is straightforward: this is a fresh open." That claim was built on a bad number and it was wrong. The scanner feed reported prior OI of 100; the actual OPRA baseline was 35,052 contracts. This is one of the most heavily-populated LEAP strikes on the NVIDIA board, not a sleepy one.

That inverts the logic entirely. When size (2,000) is far smaller than prior OI (35,052), the trade size proves nothing about open vs. close — a 2,000-lot print could be opening, closing, or transferring, and the tape alone cannot tell you which. The honest label yesterday should have been "unprovable, pending next-day OI." It wasn't. Here is what the OI actually says:

Verdict: a partial, net open — not the clean 100% fresh open we implied. Open interest rose by 1,408 contracts against a 2,000-lot cross (and 2,213 contracts of total volume in the strike that day). Roughly 70% of the block created genuinely new contracts; the remaining ≈30% was a transfer — existing holders handing their position to someone else, creating no new open interest at all.

This matters, and here's the plain-English version: a big premium number does not mean a big new position. Of the $6.3M headline, only about $4.4M represents newly-created risk. The rest was one party stepping out of a position while another stepped in. We have seen this exact pattern before (AMAT in June, where a size-beats-OI "open" resolved as mostly transfer) — the difference is that here we never had the size-vs-OI cushion in the first place.

What next-day OI still cannot tell us: OI resolves open vs. close. It does not resolve direction. A cross has a known counterparty by definition, so there is no aggressor signal to lean on. Treat this as a partially-opened, long-dated position with a mild buy-lean (it printed near the ask) — not a confirmed bullish bet, and not the 21x-prior-OI whale the original numbers made it look like.

🤓 What This Actually Means — Plain English

Let's decode this one carefully, because the mechanism matters as much as the trade itself:

• 🤝 This is a "block cross," not a sweep. Somebody wanted to move 2,000 contracts of a fairly illiquid, far-dated LEAP option, and instead of hitting the lit market (which would have moved the price and taken forever to fill), a broker pre-arranged the trade with a counterparty who already agreed to take the other side. That's normal, common institutional plumbing — it is not urgent, panicked, or aggressive by nature.
• 🎲 Direction is a coin we can't fully flip. In a lit trade, if you buy at the ask, that's strong evidence someone was willing to pay up to get filled now. On a cross, both sides already agreed on the price beforehand — printing near the ask ($31.45 vs. $31.05/$31.60) is a soft hint the buy-side pushed the negotiation slightly, but it is genuinely consistent with either:
  1. A buyer accumulating long-dated, out-of-the-money upside exposure to NVIDIA's next two product cycles (Rubin, Rubin Ultra) without wanting to move the market, or
  2. A seller — someone writing (selling) these calls against an existing long stock or long-call position as an overwrite/financing structure, who was simply willing to accept a price near the top of the negotiated range because the premium collected (≈$6.3M) served their own goal (income, hedging, or unwinding delta).
• 📅 The order type remains genuinely unprovable. Our internal classifier flags this trade LOW confidence and explicitly can't assign a clean BTO/STO because it's a cross, not a clean single-sided electronic print. Next-day OI (+1,408) tells us a position was partially opened; it still says nothing about which side our desk was on. Read this as "a LEAP call position was partly opened and partly transferred, direction leans slightly toward a buyer, but treat that lean as a hint, not a fact."
• 🗓️ Why the date matters: a Jan 2028 expiry captures roughly six more quarterly earnings reports (Aug 2026, Nov 2026, Feb 2027, May 2027, Aug 2027, Nov 2027) plus the entire Rubin → Rubin Ultra → Feynman product roadmap. Whoever is on the long side of this contract (if it is a buyer) isn't betting on one earnings print — they're betting on the multi-year AI-capex arc.

Unusual Score: 🟡 MODERATE (revised down, July 10) — the original score of MODERATE-HIGH rested on "2,000 contracts against just 100 prior OI (≈21x)." That ratio was wrong. Against the true prior OI of 35,052, this block is ≈0.06x the existing open interest, and next-day OI confirms it created only 1,408 new contracts. It remains a real $6.3M facilitated block on a far-dated LEAP — worth watching — but it is not an outsized position relative to what already sits at that strike, and it is a negotiated cross rather than aggressive lit flow. Both facts argue for lower conviction, not higher.

---

📈 Technical Setup / Chart Check-Up

YTD Performance Chart

NVIDIA opened 2026 at $189.84 (January 2) and closed July 8 at $204.12 — up ≈7.5% year-to-date — while holding the title of the world's most valuable public company at ≈$4.9 trillion market cap. The stock has spent 2026 grinding higher on record Data Center revenue (Q1 FY2027: $81.6B total revenue, $75.2B from Data Center alone, +92% YoY) even though the May 20, 2026 beat-and-raise quarter still produced a post-earnings slide — a reminder that expectations here are already sky-high and even great numbers don't guarantee an up day.

Key observations:

• 📈 Steady, less-parabolic climb than some AI peers — reflects NVDA's sheer size ($4.9T) making 20%+ single-quarter moves mechanically harder
• 💰 NVIDIA authorized an additional $80.0B buyback and hiked its quarterly dividend 25× (from $0.01 to $0.25) alongside the May print — a signal of management confidence in free cash flow durability
• ⚠️ "Priced for perfection" risk: a beat-and-raise quarter that still sold off tells you the bar is set very high heading into the next print

Gamma-Based Support & Resistance Analysis

Current Price: ≈$203.40

A quick note before the numbers: this gamma map reflects near-term dealer positioning (nearer-dated expiries), not the 2028 LEAP itself — but it's still the best lens for where NVDA is likely to chop around over the next days-to-weeks, and it's useful context for anyone trading shorter-dated NVDA options around this news.

🔵 Support Levels (Put Gamma Below Price):

• $202.50 — immediate floor, "Very Strong," just 0.4% below spot
• $200.00 — the single largest gamma wall on the board (254.7B total gamma, 95.6B net) — this is the line dealers will defend hardest
• $195.00 — secondary support, "Very Strong," ≈4.1% below spot
• $190.00 and $185.00 — deeper support shelves if $195 gives way

🟠 Resistance Levels (Call Gamma Above Price):

• $205.00 — immediate ceiling, "Very Strong," with the highest net call-gamma reading of any strike (139.1B net) — the strongest single directional wall on the chain
• $207.50 and $210.00 — secondary resistance zone, both "Very Strong"
• $220.00 — the next major wall (73.7B total gamma), ≈8.2% above spot
• $250.00 — interesting: there's already a modest resistance wall sitting almost exactly at the LEAP's $250 strike (15.8B total gamma, mostly call-side), ≈22.9% above spot — a small existing pocket of open interest at the same level this cross just added to

What this means for traders: NVDA is currently pinned in a tight band between the $200 mega-wall (support) and the $205 mega-wall (resistance) — both are the largest gamma readings on the entire chain, which typically means dealer hedging keeps price oscillating in that pocket barring a real catalyst. Net gamma is call-dominated from $195 up through $260 (positive net_gex at nearly every strike in that band), while put gamma dominates only below ≈$190 — a broadly bullish/call-heavy near-term positioning structure consistent with a mega-cap trading near its highs.

Implied Move Analysis

Options market pricing for upcoming expirations (from spot ≈$203.42):

• 📅 Weekly (Jul 10 — 1 day): ±$4.15 (±2.04%) → Range: $199.27 – $207.57
• 📅 Monthly OPEX (Jul 17 — 8 days): ±$11.57 (±5.69%) → Range: $191.85 – $214.99
• 📅 Quarterly Triple Witch (Sep 18 — 71 days): ±$38.52 (±18.93%) → Range: $164.90 – $241.94
• 📅 Longest tracked horizon — "Yearly LEAPS" (Jun 17, 2027 — 343 days): ±$86.29 (±42.42%) → Range: $117.13 – $289.71

Translation for regular folks: the market is pricing a modest ≈2% wobble by tomorrow and a ≈19% swing by the September triple-witch (which captures the Aug 26 earnings report). The longest horizon our tool tracks (June 2027) is still ≈7 months short of the actual Jan 2028 option expiry — but even that shorter-dated cone already stretches its upper bound to $289.71, which comfortably clears the $250 strike. Extrapolate a bit further out to the real Jan 2028 expiry and the implied range widens further still, meaning the option chain is already pricing real odds that NVDA trades meaningfully above $250 well before this contract expires — that's the whole thesis behind buying (or selling) a $250 call this far out.

Net GEX Bias: Call-gamma dominant across the near-term chain (net positive at nearly every strike from $195–$260) — consistent with a market still leaning constructively on NVDA even after a run to record valuations.

---

🎪 Catalysts

🔥 Confirmed Near-Term Catalyst

Q2 FY2027 Earnings — Tuesday, August 26, 2026, after market close (Calendarial, MarketBeat)

• Guided revenue: ≈$91.0B ± 2%, which excludes any China Data Center compute revenue — watch for whether Data Center growth sustains its ≈90% YoY pace and whether China gets re-included
• This is ≈17 months before the Jan 21, 2028 LEAP expiry — just the first of roughly six more quarterly reports the contract will live through

📊 Recent Catalysts (Already Happened — Context)

• May 20, 2026 — Q1 FY2027 results: $81.6B revenue (+85% YoY), $75.2B Data Center (+92% YoY), $80B buyback authorization, 25× dividend hike — a beat-and-raise that still saw the stock slide on the print, underscoring how high the bar already sits.
• June 5, 2026 — SpaceX ↔ Google compute deal: SpaceX is leasing 110,000 NVIDIA GPUs, with Google paying $920M/month from October 2026 through June 2029 (Motley Fool) — multi-year, contracted demand that extends past this option's expiry.
• IREN partnership: up to 5 gigawatts of AI compute capacity being built on NVIDIA silicon (FXLeaders).
• January 13, 2026 — BIS H200 China licensing framework: case-by-case approvals allowed with a 25% export fee, capped at under 50% of U.S. sales (BIS) — the single biggest regulatory swing factor of the last two quarters, and current guidance excludes China entirely, meaning any re-inclusion is pure upside optionality.

🚀 The Structural Thesis That Actually Underwrites a 2028 Call

This is what a genuinely long-dated LEAP is really betting on — not any one print, but the product roadmap and demand backdrop through the option's life:

1. Rubin volume production begins Q1 2027 — HBM4, 3nm, ≈3.3× the compute of Blackwell B300; Vera Rubin already entered full production around June 1, 2026, with R100 sampling in Q4 2026 (Tom's Hardware).
2. Rubin Ultra arrives H2 2027 — the NVL576 rack (up to 576 GPUs), ≈15 ExaFLOPS FP4, 384GB HBM4E — a marquee product catalyst that lands before this option expires (Tom's Hardware).
3. Feynman architecture (2028) is already on the public roadmap, giving the market "next-gen visibility" that can support multiple expansion into and past the option's expiry.
4. ≈$650–725B of aggregate 2026 hyperscaler AI-capex (Amazon ≈$200B, Microsoft ≈$190B, Google $175–185B, Meta $115–135B) is the demand backdrop underneath every one of these product cycles (Yahoo Finance).
5. Analyst targets already clear the strike: consensus rating is Strong Buy (≈60 buy / 16 hold / 3 sell), with an average 12-month price target of ≈$309–317 and a Street high of $500 (Baird) — well above the $250 strike, even on a 12-month view that's only about a third of this contract's remaining life (MarketBeat).

---

🎲 Price Targets & Probabilities

Combining gamma structure, implied-move data, analyst targets, and the catalyst roadmap:

📈 Bull Case (30% probability)

Near-term target: break and hold above $210 gamma resistance → structural target: $300+ by mid-2027, potentially $309-500 (Street range) before Jan 2028 expiry

• 🚀 Rubin ramp (Q1 2027) and Rubin Ultra (H2 2027) both land on schedule and hyperscalers keep accelerating capex
• 📊 August 26 earnings beats the $91B guide with Data Center growth holding near 90% YoY
• 🇨🇳 China H200 licenses get approved at scale, adding revenue that's currently excluded from guidance entirely
• 📈 A clean break above the $205/$207.50/$210 gamma wall cluster opens room toward $220-$230 near-term
• If this scenario plays out, the $250 LEAP call would move solidly in-the-money well before its 2028 expiry

🎯 Base Case (45% probability)

Range-bound consolidation, roughly $190-$230 over the next several months

• ✅ Solid but not spectacular execution — Data Center growth decelerates from ≈90% toward the 50-60% range as the comps get harder
• ⚖️ Guidance in-line, China remains excluded, no major surprises either direction
• 🔄 Price continues oscillating around the $200/$205 gamma mega-walls for weeks at a time, consistent with the "priced for perfection, beats still sell off" pattern seen in May
• 📊 This is the toughest scenario for the $250 strike near-term — it would need real time (and a fresh catalyst like the Rubin ramp) to work

📉 Bear Case (25% probability)

Target: test $180-$190 gamma support, or lower on a broader AI-capex re-rating scare

• 😰 China policy tightens further instead of loosening, or a hyperscaler signals capex deceleration ("not enough cash for AI" narrative already circulating)
• 🚨 Rubin execution slips — HBM4 supply or TSMC 3nm capacity constraints delay the ramp
• 💸 Broader AI-valuation compression hits the most expensive, most crowded mega-cap the hardest
• 🔨 A break below the $200 mega-wall (254.7B gamma) could cascade toward $195, then $190/$185 support shelves

If this cross was actually a seller (writing calls, not buying them): the bear/base case is exactly where that position wins — the calls expire worthless or cheap, and the ≈$6.3M collected becomes pure profit. This is precisely why we can't read a directional verdict off a cross alone.

---

💡 How Four Different Traders Might Read This

🎲 YOLO Trader

If you independently believe the multi-year Rubin → Rubin Ultra → Feynman roadmap, you can buy a small number of the same NVDA Jan 2028 $250 calls — ≈561 days of runway, defined risk (you can't lose more than the ≈$3,145/contract premium), and the average analyst target (≈$309–317) already clears the strike. The catch: theta bleeds constantly over 1.5 years, a beat-and-raise already produced a sell-off once (May 2026), and China policy is a binary that can swing the stock 10%+ overnight. Size it as money you can lose entirely.

📈 Swing Trader

Instead of a 1.5-year outright, express the same bullish thesis on the Aug 26 earnings with a defined-risk call spread (e.g. buy a September $210 call / sell a $230), sized small — the Sep-18 triple-witch implied move (±18.93%) already prices real odds of a move into that zone. The $200/$205 gamma mega-walls say NVDA is coiled; let a range break + earnings be your trigger rather than chasing.

💵 Premium Collector

The cross itself is a purchase, so mirroring it isn't your trade. If you're constructive, sell premium below the $200/$205 gamma support — a cash-secured put or a put credit spread collects income while that wall does the work, and you're paid to be willing to own NVDA lower. Never sell naked calls on the most valuable company in the world into an AI-capex supercycle.

🌱 Beginner

Don't copy this blind. We cannot prove the cross was even a buyer — copying an unprovable-direction trade is copying a coin flip, not "smart money." A LEAP this far out has enormous time value at risk and there's no rush. Let the Aug 26 earnings and tomorrow's OI snapshot add real information first, and use this as a lesson: a big LEAP cross confirms positioning, never direction.

Risk level: High (full premium at risk over a long hold) | Skill level: Advanced

---

⚠️ Risk Factors

What could go wrong — and what the tape simply can't tell us:

• 🤝 Direction is unprovable, full stop. This was a negotiated block cross with a known counterparty. There is no NBBO aggressor signal, no lit-market "took the offer" proof. Anyone telling you with confidence "this is a bullish $6.3M bet" is overstating what the tape actually shows.
• 👥 No visibility into who's on either side. OPRA data cannot show broker/MMID, customer identity, order ID, or whether either counterparty already held a related position (stock, other options, or a hedge in another instrument entirely). This could be a hedge fund, a market-making desk, a corporate treasury program, or something else entirely — we simply don't know.
• 🇨🇳 China policy is a binary swing factor. Current guidance ($91B Q2 guide) excludes China Data Center revenue entirely. The Jan 13, 2026 BIS framework allows only case-by-case, fee-laden approvals. Any tightening removes optionality; any loosening adds it — either way, headline risk is real and can move fast.
• 💸 Valuation leaves little room for error. At ≈$4.9T market cap, NVDA is priced for continued near-perfect execution. The May 2026 beat-and-raise-then-selloff is direct proof that good numbers don't guarantee a good stock reaction anymore.
• 🏭 Rubin/Rubin Ultra execution risk. HBM4 supply, TSMC 3nm capacity, and the complexity of the NVL576 rack all have to come together on schedule. Any meaningful delay pushes out the thesis that underwrites a $250 strike this far out.
• 📉 "Not enough cash for AI" narrative. Some analysts are already questioning whether hyperscaler capex can keep compounding at 2026's pace — if that spending decelerates, it hits NVDA's growth story directly.
• ⏰ Time decay is a real cost even if you're eventually right. A 1.5-year hold means substantial premium can bleed away in a range-bound market even if NVDA ultimately gets to $250+ — timing matters, not just direction.
• 🎢 Competition isn't standing still. AMD's MI-series (also facing China restrictions), and custom hyperscaler silicon (Google TPU, Amazon Trainium, Meta MTIA) all chip away at the edges of NVDA's dominant share over a multi-year window.

---

🎯 The Bottom Line

Real talk: Somebody moved 2,000 contracts (≈$6.3M) of a far-dated, out-of-the-money NVDA call through a negotiated block cross today. That's a real, freshly-opened position — the size-versus-OI math proves that much cleanly. What it does not prove is direction. This printed near the ask, which is a soft lean toward "a buyer paid up," but on a cross that lean is a hint, not a verdict — it's just as plausible this was a premium seller writing calls against an existing position and simply accepting a price near the top of a pre-negotiated range.

What this trade tells us, honestly:

• 🎯 A sophisticated player has real conviction — good or bad — about NVIDIA's trajectory through the Rubin and Rubin Ultra product cycles, spanning roughly six more earnings reports
• 💰 The $250 strike sits well below the average Street 12-month target (≈$309-317), meaning if this leans bullish, the buyer is betting on a scenario the analyst community already broadly expects — just wants the leveraged, defined-risk way to express it
• ⏰ The negotiated, off-book mechanism (not a lit sweep) tells us this is patient, planned positioning — not a reactive, urgent bet on any single headline
• 📊 Right now the stock itself is coiled between the $200 and $205 gamma mega-walls — near-term price action looks more like consolidation than trend

This is NOT a "load up on $250 calls" signal — it's a "watch how this resolves, then decide" signal.

If you're watching from the sidelines:

• ✅ The OPRA open-interest snapshot is in: OI moved 35,052 → 36,460 (+1,408), a partial net open — ≈30% of the block was a transfer, not new risk. It does not resolve direction
• 📅 August 26, 2026 is the next real catalyst — the $91B guide (ex-China) is the number to watch
• 🎯 Use the $200/$205 gamma walls as your near-term map: a clean break of either resolves the current range

If you're bullish on the multi-year thesis:

• 📊 Consider a smaller, shorter-dated position first (see the Balanced idea above) to get a read on how NVDA trades through an actual catalyst before committing to anything 1.5 years out
• 🛡️ Size any LEAP-style position knowing theta is a real, constant cost over that horizon — not just a directional bet

If you're skeptical or bearish:

• 📉 The May 2026 "beat-and-raise-then-selloff" is your best evidence that this stock is priced for perfection — any real Rubin delay or capex deceleration headline could hit hard
• ⚠️ $200 gamma support (254.7B, the largest wall on the chain) is the level to watch — a clean break below there would be the first real technical crack

Mark your calendar — key dates:

• 📅 July 10, 2026 — weekly options expiration (±2.04% implied move)
• 📅 July 17, 2026 — monthly OPEX (±5.69% implied move)
• 📅 August 26, 2026 — Q2 FY2027 earnings, ≈$91B guide (ex-China)
• 📅 September 18, 2026 — quarterly triple witch (±18.93% implied move, captures earnings reaction)
• 📅 Q1 2027 — Rubin volume production begins
• 📅 H2 2027 — Rubin Ultra launch
• 📅 January 21, 2028 — this option's expiration

Final verdict: NVIDIA's structural AI story remains genuinely strong — record Data Center revenue, a locked-in annual product cadence, and analyst targets that already clear this strike. But a $6.3M cross with an unprovable direction is not, by itself, proof of anything beyond "someone with real size wanted long-dated $250-strike exposure, one way or the other." Treat it as a data point pointing toward NVIDIA's importance to the AI trade — not as a signal to copy blind.

This is a marathon position on a marathon company. Don't let one negotiated print do your thinking for you. 💪

Disclaimer: Options trading involves substantial risk of loss and is not suitable for all investors. This analysis is for educational purposes only and not financial advice. Past performance doesn't guarantee future results. Because this trade printed as a negotiated block cross, its direction (buyer vs. seller initiated) cannot be proven from the public tape — treat any directional read in this article as an informed lean, not a fact. Always do your own research and consider consulting a licensed financial advisor before trading. A 1.5-year options position carries substantial time-decay and total-loss-of-premium risk regardless of your directional view.

---

About NVIDIA Corporation: NVIDIA designs the GPUs, CUDA software platform, and data-center networking technology that power the majority of the world's AI training and inference infrastructure, with a market cap of ≈$4.9 trillion in the Semiconductors & Related Devices industry — the world's most valuable public company.

---

Last updated: July 10, 2026 — corrected the prior open-interest figure (100 → 35,052; the original was a scanner data error) and resolved the open/close flag with the next-day OPRA snapshot (OI 35,052 → 36,460, +1,408: PARTIAL NET OPEN, ≈30% transfer). Unusual Score revised MODERATE-HIGH → MODERATE.