🤝 ARM $17.4M ITM Call Block Cross — RESOLVED: It Was a Mutual Exit, Not an Overwrite, Ahead of July 29 Earnings

📅 July 8, 2026 | 🔥 Unusual Activity Detected

✅ Update — July 9, 2026: The next-day OPRA open-interest snapshot resolved this trade. Open interest at the $270 strike FELL 13,195 → 10,061 (−3,134) — a decline exactly equal to the trade size. That mathematically rules out the covered-call/overwrite story and confirms this was a position-closing trade on both sides: the seller closed a long call, the counterparty bought back a short. Title, quick take, and narrative below have been updated from the original "unprovable" framing.

---

🎯 The Quick Take

Someone crossed 3,134 Arm Holdings $270 calls expiring August 21 at $55.60 on July 8 — a ≈$17.4 MILLION SELL, printed right at the bid as a negotiated block with a known counterparty on the other side. ARM was at $295.58, so this $270 strike was already in-the-money. When we published, size (3,134) sat below the existing open interest (13,000), so the tape alone could not prove whether this opened a new short call or closed an existing long. The next-day open-interest snapshot settled it: OI fell by exactly 3,134 contracts. Open interest can only fall when the seller is closing a long and the buyer is closing a short — so this was not an overwrite for yield. It was a mutual unwind: 3,134 contracts of existing exposure erased from the board, three weeks before a binary earnings print. Translation: two institutions agreed to walk away from this strike ahead of July 29 — one banking a gain on a long call, the other retiring a short.

---

📊 Company Overview

Arm Holdings plc (ARM) is the dominant licensor of the CPU instruction-set architecture that powers roughly 99% of the world's smartphones, and it's pushing hard into data-center CPUs, edge/IoT, automotive, and — as of March 2026 — its own production silicon:

• Market Cap: ≈$367.6 Billion (companiesmarketcap)
• Industry / Sector: Technology — Semiconductors (fabless chip IP / architecture licensing)
• Current Price: $295.58 (52-week range $100.02 – $452.70) (Yahoo Finance)
• Primary Business: CPU architecture licensing + royalties, Compute Subsystems (CSS), and the new Arm AGI CPU — its first in-house data-center chip, co-developed with Meta
• Ownership: ≈90% owned by SoftBank Group; CEO Rene Haas was also named CEO of SoftBank Group International effective April 21, 2026 (SoftBank)
• 2026 story so far: Stock roughly doubled YTD at its mid-June peak (≈$420-452), then pulled back ≈35% off that high into today's print (24/7 Wall St)

---

💰 The Option Flow Breakdown

📊 What Just Happened

The Tape (July 8, 2026 @ 14:29:16 ET):

Mechanism, plainly: This printed at the bid ($55.60 bid / $56.90 ask) as a negotiated block cross 🤝 — a broker matched a known seller with a known buyer off the lit order book. That means there's no "aggressor" to read here (no lifting the ask, no hitting the bid in the retail sense) — direction is unprovable from the print itself.

✅ RESOLVED — Next-Day OPRA Open Interest Confirms a CLOSE on Both Sides

The resolving OPRA open-interest snapshot (posted July 9, 2026 pre-market, reflecting end-of-day July 8) is in. When we published, the trade size (3,134 contracts) sat below prior open interest (13,195), so size ≤ OI meant the tape could not prove open vs. close. It can now:

Why this is decisive, not a lean. Open interest counts contracts that exist. A seller either opens a short (OI up) or closes a long (OI down). A buyer either opens a long (OI up) or closes a short (OI down). Open interest therefore falls only when both parties are closing. Here it fell by 3,134 — precisely the trade size. So:

• ❌ The covered-call / overwrite story is ruled out. A fresh short call would have pushed OI up toward ≈16,300. It went the other way.
• ✅ The profit-taking exit story is confirmed — and it's stronger than we could have guessed. The seller sold out of an existing long $270 call (STC), and the counterparty simultaneously bought back an existing short (BTC). Both desks left the strike.

What it means: this $17.4M print did not add a single contract of new exposure. It removed 3,134 contracts of it, 21 days before ARM's July 29 earnings. That is de-risking on both sides of a negotiated block — not a yield trade, and not fresh conviction in either direction.

---

🤓 What This Actually Means — Plain English

When we first published, a sold, in-the-money call, printed as a cross split into two very different stories. The open-interest snapshot has now eliminated one of them.

• 🛡️ Story #1 — Covered-call / overwrite for yield: RULED OUT. The idea was that someone holding long ARM stock sold this $270 call against it to collect $55.60/share (≈19% of the stock price) as income — a classic STO (Sell-To-Open). But an STO creates open interest, and open interest fell by 3,134. This never happened. Nobody wrote a new covered call here.

• 📤 Story #2 — The exit. CONFIRMED. Someone who was already long this exact $270 call — bought weeks or months ago when ARM was cheaper — cashed out ahead of the binary July 29 earnings print (STC — Sell-To-Close). And the open-interest math tells us something the original article couldn't: the buyer on the other side wasn't opening a new long either. They were buying back a short call they already had (BTC — Buy-To-Close). That's the only way open interest drops by the full trade size.

What "both sides closed" actually means for you. This wasn't one desk expressing a view and another taking the other side of it. It was two desks who both already had exposure at this strike agreeing to erase it, together, three weeks before a print that has historically been violent (ARM's last quarter was a record beat and the stock still fell ≈7%). Read it as risk coming off the table, not as a directional signal. The long holder banked $55.60/share rather than carry an ITM call through earnings; the short holder paid up to stop being short calls into the same event. Both were reducing, and neither was betting.

What we still cannot tell you: who they were, what their cost basis was, or whether either leg was hedged with stock elsewhere. A block cross is a negotiated deal — the identities and the surrounding book stay invisible to OPRA.

Unusual Score: flagged for size (≈$17.4M single print, well above typical ARM single-leg block activity) — but per our house rule, a cross gets the 🤝 tag, never the "aggressive whale" framing, because there's a known counterparty and the trade is negotiated, not swept off the book.

---

📈 Technical Setup / Chart Check-Up

YTD Performance Chart

ARM's 2026 has been a rollercoaster: the stock roughly doubled YTD by its mid-June all-time high (≈$420-452), then gave back a sharp ≈35% into today's $295.58 print — a violent round-trip on a name that's still up big for the year but has clearly cooled off heading into earnings.

Gamma-Based Support & Resistance Analysis

Current Price: ≈$295.58-$296.33 (minor intraday drift between data snapshots)

• 🟠 $300 — nearest overhead resistance (moderate strength, total gamma exposure ≈3.45B, just ≈1.2% above spot). This is the single largest gamma level on the board and the first real speed bump for any near-term rally.
• 🔵 $285-$290 zone — put-gamma heavy (total gamma ≈0.83-0.96B, put-dominated), acting as the nearest dealer-hedging support directly below spot.
• 🟠 $270 (the sold strike itself) — notable call-gamma pocket. Even though $270 sits ≈8.9% below current spot, net gamma there is call-dominant (net +0.80B) — a sign there's real open interest and dealer call exposure concentrated exactly where this trade struck. That's consistent with 13,000 contracts of prior OI already sitting at this strike.
• 🔵 $250 — deeper structural level (total gamma ≈0.75B, heavily put-weighted) if ARM keeps sliding.

What this means for traders: Compared to a mega-cap like AMD (tens of billions in gamma per strike), ARM's total gamma footprint is much smaller in absolute terms — meaning price can move more freely between levels without dealer hedging flows pinning it as tightly. The $300 level is the nearest "ceiling" dealers will lean against; below spot, $285-$290 is the first cushion.

Implied Move Analysis

Options market pricing across ARM's key expirations:

• 📅 Weekly (Jul 10 — 2 days): ±8.64% (±$25.57) → Range: $270.44 – $321.58
• 📅 Monthly OPEX (Jul 17 — 9 days): ±16.45% (±$48.69) → Range: $247.32 – $344.70
• 📅 August OPEX (Aug 21 — this trade's expiration, ≈44 days): implied range roughly $200.11 – $391.91 (≈±32% around spot) — this is the window this exact $270 call trade needs to navigate
• 📅 Quarterly Triple Witch (Sep 18 — 72 days): ±45.16% (±$133.67) → Range: $162.34 – $429.68

Translation for regular folks: The options market is pricing in HUGE potential swings for ARM — a ±16.45% move possible just by next week's monthly OPEX, and a market-implied range between roughly $200 and $392 by the time this $270 call expires on August 21. That's a stock that can plausibly finish this trade deep in-the-money (well above $270) or comfortably out-of-the-money (well below $270) — which is exactly the kind of uncertainty that makes selling an ITM call (for income OR to exit) a reasonable move heading into a binary earnings print.

---

🎪 Catalysts

✅ The One Hard Catalyst Inside This Option's Window

Q1 FY2027 Earnings — Tuesday, July 29, 2026 (quarter ended June 30, 2026), 23 days before the Aug 21 expiration of this exact trade. Company guidance set on the May 6 Q4 call: Revenue $1.26B ± $50M (≈+20% YoY), Non-GAAP EPS $0.40 ± $0.04 (≈+14% YoY). Watch data-center royalty trajectory (it more than doubled YoY last quarter), Armv9 mix, CSS traction, and any AGI-CPU supply-capacity commentary (marketchameleon, Arm 6-K/SEC). Important: the Aug 21 expiration itself is NOT a company event — it's simply when this contract settles, 23 days after the catalyst that will actually move the stock.

History says this is violent: Arm's Q4 FY2026 print on May 6, 2026 was a record quarter — revenue $1.49B (+20% YoY), licensing $819M (+29% YoY, record), data-center royalty more than doubled YoY, EPS $0.60 beating consensus — and the stock still slid ≈7% on supply-capacity caution around the AGI-CPU ramp (heygotrade, Arm newsroom).

🔥 Recent Catalysts (Last ≈3 Months)

• Arm AGI CPU launch, March 2026 — Meta as lead partner. Arm's first in-house production silicon: a 136-core data-center CPU for agentic AI workloads, delivering >2x performance-per-rack versus x86 (Arm newsroom, Tom's Hardware). Management said booked demand roughly doubled from ≈$1B to >$2B across FY2027-FY2028 on the Q4 call, but supply-chain capacity is the key constraint and first production revenue isn't expected until Q4 FY2027 — early 2027, well past this option's expiration (BigGo Finance).
• Rene Haas expanded role at SoftBank, April 21, 2026. Arm's CEO additionally became CEO of SoftBank Group International, tightening Arm-SoftBank alignment but also amplifying the SoftBank-overhang question (SoftBank, Bloomberg).
• Analyst spread widening: Mizuho at $500 (Street high, June 4, 2026) versus Deutsche Bank at $140 (Street low, March 25, 2026) — one of the widest target ranges in large-cap tech (MarketBeat). At least one analyst has gone to Strong Sell citing valuation plus SoftBank margin-loan risk (Seeking Alpha).

⏳ Expected / Structural (Mostly Land After the Aug 21 Expiration)

• Arm AGI CPU production ramp — first revenue not until Q4 FY2027 (early 2027); a qualitative update is possible on July 29, but no revenue lands before this option expires (mlq.ai).
• Armv9 + CSS royalty-rate uplift — Armv9 carries ≈2x the royalty rate of Armv8 (≈25% of royalty revenue and climbing); CSS royalties run >10% per chip versus 1-3% traditional. A multi-quarter tailwind, not an August event (mlq.ai).
• SoftBank's broader AI-infrastructure buildout (SB Neo neocloud, Portsmouth data center) — a longer-dated demand driver, not an August catalyst (whatfinger).

---

🎲 Price Targets & Probabilities

Using the gamma levels, implied-move data, and the July 29 earnings catalyst, here's how the next ≈6 weeks (through Aug 21 expiration) could play out:

📈 Bull Case (30% probability)

Target: $340-$390 (upper edge of the Aug OPEX implied range)

Q1 FY2027 beats on both revenue and EPS, data-center royalty growth accelerates further, and management raises AGI-CPU supply-capacity visibility. ARM breaks back above the $300 gamma resistance and re-tests the $340+ area, deep into the implied-move upper band. This is the scenario the exiting long holder gave up — they sold their $270 call at $55.60 and would have collected considerably more by holding it through a rally.

🎯 Base Case (45% probability)

Target: $270-$310 (chops around the sold strike)

Earnings come in roughly in-line with guidance ($1.26B rev, $0.40 EPS) — solid growth, but management flags AGI-CPU supply constraints again (echoing May's ≈7% post-earnings drop despite a record quarter). ARM oscillates between the $285-$290 put-support zone and the $300 gamma resistance, finishing the option's life close to the $270 strike either side. This is the scenario that vindicates both sides of the confirmed exit — the long holder took $55.60 near the top of a chop range, and the short holder retired an obligation that was never going to pay off big either way.

📉 Bear Case (25% probability)

Target: $200-$250 (test of deeper gamma support)

A soft July 29 print — particularly weak 2H licensing commentary or another AGI-CPU capacity warning — triggers a sharper leg down given the stock is already ≈35% off its highs. ARM breaks the $250 structural support and tests the lower end of the ±32% Aug OPEX implied range (≈$200). In this scenario the $270 call expires worthless — meaning the long holder who exited at $55.60 made exactly the right call, and the short holder who paid to buy back the same call gave up a windfall. This is the outcome the confirmed exit was insuring against, at least on the selling side.

---

💡 How Four Different Traders Might Read This

(Open vs. close is now confirmed — this was a mutual exit, so there is no position to "copy." These are independent ways to express a view into the July 29 earnings, sized for the ±16–32% implied moves in play.)

🎲 YOLO Trader

If you think ARM's earnings move beats the ≈16.45% the market is pricing, a short-dated straddle around the July 29 print bets on a violent reaction (ARM dropped ≈7% on a beat last quarter — realized has repeatedly topped implied). The catch: straddles get brutal IV crush the instant earnings drop — you can be right on the move and still lose. Precise timing, tiny size, exit within a day or two.

📈 Swing Trader

The stock is ≈35% off its highs with $300 as clear gamma resistance and $285–$290 support. Wait for the July 29 print, then trade the reaction off those levels rather than guessing ahead of a binary with expensive options. A defined-risk directional spread post-earnings (once IV crushes) is cheaper than buying rich pre-earnings premium.

💵 Premium Collector

Note the correction: this was not a covered call being written. It looked like one on the surface (a sold ITM call), and the open-interest data proved it was an exit instead. That distinction matters for you — nobody here decided ITM call premium was worth selling into this earnings setup. If you hold ARM shares and still want income, an out-of-the-money covered call is the play — but do it after July 29 (selling before the binary means you're short the earnings gap on top of the cap). Selling into still-elevated post-print IV, with $300 as a natural first target strike, harvests premium without pre-earnings gap risk. Never sell naked calls on a +158%-YTD name.

🌱 Beginner

Stay on the sidelines until July 29 clears. Options are richly priced (±16% move by next OPEX alone), and ARM has a habit of big post-earnings drops even on good numbers. Here's the lesson worth more than the trade: this print looked like a $17.4M bearish call sale, and the open-interest check showed it was two institutions closing out old positions — no new bet at all. A big headline dollar figure tells you how much changed hands, not what anyone thinks. Watch how the $300/$285 levels hold, and don't pay up for a binary.

---

⚠️ Risk Factors

• ✅ Open vs. close is now proven — it was a close, on both sides. Open interest fell 13,195 → 10,061 (−3,134, exactly the trade size). This trade added zero new exposure. Any "bullish" or "bearish" read of it is unsupported: closing trades express an exit, not a view.
• Zero counterparty visibility. A block cross has a known buyer and seller to the broker who matched it, but retail traders (and we) have no way to see who they are, their cost basis, or whether either side is hedged elsewhere (stock, other options, or otherwise). Open interest tells us what happened to the contracts, never who or why.
• Binary earnings risk, July 29. ARM's last quarter was a record and the stock still fell ≈7%. The market is pricing a ±16.45% move by next week's monthly OPEX alone — a sharp gap in either direction is a real possibility.
• Valuation remains rich even after the pullback. Even ≈35% off its June highs, ARM trades at a steep multiple reflecting aggressive AI/data-center growth expectations; a soft 2H-licensing tone on the call could reprice the stock hard.
• SoftBank overhang. ≈90% SoftBank ownership plus reported margin-loan/liquidation-risk chatter around SoftBank's broader AI capex creates a technical supply overhang risk that's largely outside ARM's own fundamentals.
• AGI-CPU is a "show-me" story. >$2B in booked demand sounds great, but first production revenue isn't until Q4 FY2027 (early 2027) — well past this option's expiration — and supply-chain capacity is the stated bottleneck.
• Smartphone unit softness. Guided flat-to-slightly-negative, relying on premium Armv9/CSS mix to offset — any underdelivery on the mix shift caps the legacy royalty base.

---

🎯 The Bottom Line

Real talk: Someone crossed $17.4 million of in-the-money ARM calls on July 8, and when we published we said honestly that we couldn't tell whether it was bullish, bearish, or portfolio housekeeping. The next-day open-interest snapshot answered it: portfolio housekeeping. Open interest at the $270 strike fell 13,195 → 10,061, a −3,134 drop matching the trade size to the contract. Open interest only falls when both counterparties are closing, so this was a mutual unwind — a long holder selling out (STC) into a short holder buying back (BTC).

What this trade tells us:

• 🎯 Not an overwrite. The covered-call story is arithmetically impossible: writing new calls raises open interest, and open interest fell. The seller was exiting a long position, not renting out upside.
• 📉 Zero new exposure created. $17.4M changed hands and the market ended the day with 3,134 fewer ARM $270 calls outstanding. This was capital leaving the strike, not entering it.
• ⏰ The timing is the story: both desks cleared out 21 days before the July 29 earnings print — the actual catalyst. The Aug 21 expiration was only ever a settlement date.
• 📊 With the market pricing a ≈$200-$392 range by Aug 21, walking away from an ITM call ahead of a print that knocked the stock ≈7% on a record beat is a defensible risk decision on both sides.

This is NOT a "smart money says sell ARM" signal. It's the opposite lesson: a $17.4M sold-call headline turned out to contain no directional bet whatsoever — and only the open-interest check could reveal that.

If you own ARM:

• ✅ Watch the July 29 print closely — that's the actual event, not this trade or the Aug 21 expiration.
• 📊 Keep an eye on the $300 gamma resistance and $285-$290 support zone for near-term technical levels.

If you're watching from the sidelines:

• ⏰ Tuesday, July 29 is the date that matters — avoid new directional bets before then given the ±16-32% implied moves in play.
• ✅ The open-interest check is in (July 9, ≈06:30 ET): open interest fell by the full trade size, confirming a close on both sides. There is no institutional position here to follow.

Mark your calendar — key dates:

• 📅 July 10 — Weekly options expiration (±8.64% implied move window closes)
• 📅 July 17 — Monthly OPEX (±16.45% implied move window closes)
• 📅 July 29 (Tuesday) — Q1 FY2027 earnings — THE catalyst
• 📅 August 21 — Monthly OPEX, expiration of this exact $17.4M trade
• 📅 September 18 — Quarterly triple witch

Final verdict: ARM's long-term story — Armv9/CSS royalty uplift, data-center royalty doubling YoY, and the new AGI-CPU/Meta partnership — remains genuinely compelling. This specific $17.4M cross has now been resolved as a two-sided exit that created no new position, so it carries no directional signal at all. Let the July 29 earnings print — not this trade — drive your decision.

This is a marathon, not a sprint. Protect your capital. 💪

Last updated: July 9, 2026 — next-day OPRA open interest resolved the open/close flag: OI fell 13,195 → 10,061 (−3,134), confirming a CLOSE on both sides. The title, quick take, plain-English section, scenarios, risk factors, and bottom line were revised from the original "unprovable / possible overwrite" framing (see the ✅ RESOLVED box above).

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. This specific trade printed as a negotiated block cross with size below existing open interest; next-day open interest has since confirmed it CLOSED positions on both sides. Counterparty identity, cost basis, and any external hedge remain invisible to the tape. Always do your own research and consider consulting a licensed financial advisor before trading. Earnings create binary event risk with potential for double-digit gaps in either direction.

---

About Arm Holdings plc: Arm Holdings is the dominant licensor of CPU instruction-set architecture powering ≈99% of the world's smartphones, expanding into data-center CPUs, Compute Subsystems, and its own AI-focused silicon (the Arm AGI CPU), with a market cap of ≈$367.6 billion in the Semiconductors & Related Devices industry.