🤝 MRVL ≈$34M 0DTE Deep-ITM Call Sale — Cashing Out Expiring Marvell Calls, Not a New Bearish Bet

Last updated: 2026-06-08

✅ RESOLVED — Next-Day OI Update (2026-06-08): ✅ These were 0DTE contracts that expired Friday 6/5 — there is no next-day OI to check. The read holds: a deep-ITM call sale at intrinsic on expiration day = a close / monetize of expiring long calls, not a new directional bet.

📅 June 5, 2026 | 🔥 Unusual Activity Detected

---

🎯 The Quick Take

A desk just sold ≈$34M worth of deep-in-the-money MRVL calls expiring TODAY — both the $215 and $222.5 strikes — in a negotiated multi-leg auction block at 11:47 AM ET. This is not a bearish call or a new directional trade. With MRVL trading at $292.42, those calls carry ≈$77-$77.65 of pure intrinsic value with zero time left to run — selling them is the equivalent of converting expiring long-call positions back into cash, the standard mechanical move on expiration day for anyone holding deep-ITM calls they don't want to exercise or take delivery on. Think: cashing out, not betting against.

---

📊 Company Overview

Marvell Technology (MRVL) is a Santa Clara-based fabless semiconductor company designing custom AI accelerators (XPUs), optical interconnect chips, Ethernet switching silicon, and data-center infrastructure components.

• Market Cap: ≈$270B
• Industry: Semiconductors — Custom AI ASICs, optical interconnect
• The AI angle: Marvell co-designs Amazon's Trainium and Microsoft's Maia AI accelerators, and together with Broadcom controls an estimated ≈95% of the custom AI ASIC co-design market, per Tech-Insider
• YTD 2026: The stock has been one of the year's biggest winners, up sharply — one estimate at ≈130% YTD heading into May earnings, per Money Morning
• Recent milestones: Record Q1 FY2027 beat on May 27, a $2B Nvidia equity investment + NVLink Fusion deal on March 31, and Jensen Huang's "next trillion-dollar company" endorsement at Computex on June 2

Real talk: this is a stock that has more than doubled in 2026 on the strength of its AI co-design franchise, a Nvidia partnership that amounts to a strategic seal of approval, and a raised revenue outlook that has consistently outpaced Wall Street models. The company is not in trouble — which is exactly why today's trade should NOT be read as bearish.

---

💰 The Option Flow Breakdown

📊 The Tape (June 5, 2026 @ 11:47:44 ET)

🤝 BLOCK CROSS — Paired Multi-Leg Auction (cond 131, MULTI_LEG_AUCTION)

Flow-type: 🤝 BLOCK CROSS — printed as a paired multi-leg auction (condition 131). A broker matched a buyer and seller off the open book; there is a known counterparty on the other side. This is a facilitated, negotiated block — not an urgent lit sweep.

Order type (provisional ⏳): STC (Sell-to-Close) — closing an existing long-call position. Vol 2,309 vs prior OI 4,300 (Leg 1) and 3,600 (Leg 2) — in both cases, vol < OI, which is consistent with a partial close of existing long positions, but cannot be proven as open vs. close from today's tape alone (see callout below).

The intrinsic math:

• $215 Call intrinsic = $292.42 − $215 = $77.42; sold at $77.65 — essentially zero time premium, selling almost purely intrinsic
• $222.5 Call intrinsic = $292.42 − $222.50 = $69.92; sold at $70.15 — again, ≈$0.23 of time premium above intrinsic
• Total proceeds: ≈$18M ($215 leg) + ≈$16M ($222.5 leg) = ≈$34M gross — but this is NOT fresh premium at risk, it is the intrinsic (stock-equivalent) value of the positions being cashed out

---

⏳ Come Back Tomorrow for the OI Confirmation — 0DTE Edition

There is a meaningful wrinkle with today's 0DTE trade: both contracts expire at market close today (June 5, 2026). On a standard multi-day expiration, you'd check next-morning OPRA OI to confirm open vs. close — but for 0DTE, these contracts cease to exist at the bell. There is no next-day OI snapshot.

Here is what the tape can tell us directly:

• Vol 2,309 < OI 4,300 (Leg 1) and 3,600 (Leg 2): size alone cannot prove open vs. close — this could be closing part of an existing long, or theoretically opening a new short into expiry. However, selling deep-ITM calls at intrinsic value on expiration day is overwhelmingly a position-exit behavior — no rational actor opens a new short position on 0DTE deep-ITM calls at $77 when there is essentially no premium to collect and maximum delta risk to the close.

• The ≈$0.23 time premium on both legs (the amount above intrinsic) is so negligible that it essentially confirms this is not a premium-collection play. A new short seller looking to "collect" on expiration day wouldn't target deep-ITM strikes — they'd sell near or slightly out of the money.

Our read: this is provisionally ⏳ STC (a close/monetize). The lean is high — the intrinsic pricing, the < OI volume, and the 0DTE deep-ITM context all converge on a position exit. But the tape cannot confirm whether it was a long position being closed (STC) or a paired institutional structure we don't have full visibility on.

What we CAN say: this is not a new bearish bet against MRVL. The $34M is existing intrinsic value being converted to cash.

---

🤓 What This Actually Means — Plain English

Let's decode exactly what happens when you sell a deep-ITM call on its expiration day.

Step 1 — What is a deep-ITM call?

A call option is "deep in the money" when the stock price is far above the strike. Here, MRVL is at $292.42 and the strikes are $215 and $222.5 — meaning these calls are $77 to $70 in the money. Every dollar MRVL moves, the call moves almost a dollar too (delta ≈ 1.0 on deep-ITM 0DTE calls). They behave almost identically to owning 230,900 shares of stock (2,309 contracts × 100 shares/contract).

Step 2 — Why sell on expiration day instead of holding?

On expiration day, deep-ITM calls have two settlement paths:

• 🔄 Exercise — call the stock at $215/$222.5, take delivery of 230,900 shares (requires massive capital)
• 💸 Sell to close — sell the calls back in the market at intrinsic value, receive cash without taking stock delivery

Selling at intrinsic is the standard, mechanical way to close an expiring call position without exercising. A fund or desk that accumulated these calls months ago — perhaps as a leveraged, capital-efficient way to own MRVL's AI/semiconductor rally — is now at expiration and simply converting that exposure back to cash. No mystery. No bearish signal. Just closing the books on a profitable long.

Step 3 — Why the block auction (cond 131)?

Selling 2,309 contracts of deep-ITM options in the open market on expiration day would move the book. Instead, the desk used a multi-leg auction (a negotiated, facilitated transaction) where a counterparty agrees to take the other side at a specific price. The ≈$0.20+ above intrinsic they received is the small premium the counterparty paid for an orderly, negotiated exit rather than having the seller disrupt the market.

Step 4 — The ≈$34M figure in context

The ≈$34M in "premium" is really just ≈$77-$70 per share of intrinsic value × 230,900 share equivalent — it represents stock-like gains accumulated in the underlying MRVL rally, not $34M of new capital being deployed. If MRVL was at $215 a few months ago when these calls were originally purchased (or near it), the intrinsic alone reflects the entire appreciation of the stock move since then.

Bottom line: a desk cashed out expiring long-call positions in MRVL on expiration day via a negotiated block auction. This is a profit-take / position close, not a new directional bet. Do NOT misread the SELL as bearish.

---

📈 Technical Setup / Chart Check-Up

YTD Performance

MRVL has been one of the most explosive large-cap performers of 2026. The stock was ≈$130% higher YTD heading into the May 27 Q1 earnings print, then extended further on Jensen Huang's June 2 Computex endorsement — hitting intraday highs near $333.50. Today's GEX snapshot has spot at ≈$282.34, which means the stock has pulled back from those Computex-spike highs. The trade printed at $292.42 earlier in today's session. The chart story is: parabolic move up, euphoric spike post-Computex, and now digesting those gains near the $280-$290 zone — which maps directly onto the gamma floor described below.

---

Gamma-Based Support & Resistance

Current GEX Snapshot Price: ≈$282.34

The gamma exposure map shows a tight cluster of major levels directly around current price, followed by a larger resistance wall at $290 and $300:

🟠 Call Gamma Resistance (Orange Bars — Overhead Sellers):

• $285 — Moderate resistance, 3.05B total GEX, net GEX +1.15B call-dominant. This is the first cap just above the GEX snapshot price. Market makers selling into the $285 zone slow MRVL's recovery off the lows.
• $290 — Strong resistance, 6.39B total GEX, net GEX +0.27B near-balanced (call/put GEX very close). This is the key battleground level — the heaviest resistance wall below $300. With the trade printing at $292.42, the stock was trading just above this level intraday.
• $300 — Strong resistance, 7.59B total GEX, net GEX +1.51B call-dominant. This is the largest single gamma wall on the board and the most significant overhead cap. The $300 round-number level also aligns with psychological resistance and the prior zone before the Computex spike. Clearing $300 would be a meaningful technical signal.
• $310 / $320 — Additional moderate-to-strong resistance further out (3.12B and 4.45B total GEX respectively), marking the zone the stock must reclaim to approach former all-time highs near $333.50.

🔵 Put Gamma Support (Blue Bars — Downside Floors):

• $280 — Moderate support, 4.94B total GEX, net GEX +0.77B (slight call lean). This is the immediate gamma floor, ≈0.8% below the GEX snapshot price. Market makers will buy MRVL aggressively near $280.
• $270 — Moderate support, 4.51B total GEX, net GEX ≈ -0.03B (near-perfectly balanced). The $270 strike has nearly equal call/put gamma — a true anchor level where the market is pricing in significant two-sided interest. A break to $270 would signal a more serious correction.

What this means for you: MRVL is sandwiched in a very tight gamma zone. The stock is essentially pinned between the $280 support floor and the $290-$300 resistance cluster. The $300 level — the largest gamma wall on the board — is the key line that separates "digesting the Computex move" from "resuming the AI-driven rally toward former highs." For today's 0DTE options specifically: the GEX structure was irrelevant to how the trade resolved (the calls expired deep ITM regardless), but it gives context for forward-looking positioning.

---

Implied Move Analysis

Current IV-derived ranges (spot ≈$282.43):

Real talk: these implied-move ranges are enormous — a $106 range for just the next 7 days. That's how volatile MRVL is priced right now. The ±18.8% weekly move reflects the combination of post-Computex spike volatility, AI-capex sentiment swings, and the upcoming August 20 Q2 earnings. The $280 support floor (from GEX) sits almost exactly at the lower boundary of the June 12 weekly implied range ($229.36), suggesting the market sees the current price zone as a genuine battleground, not a one-directional move.

Key observation: the $335-$338 zone (near the all-time high) aligns with the top of the weekly and early opex implied ranges — a potential near-term bull target if the AI narrative reignites.

---

🎪 Catalysts

✅ Already Happened (In the Books)

• Q1 FY2027 Record Beat (May 27, 2026): Revenue $2.418B (+28% YoY), Data Center $1.833B (+27% YoY), Non-GAAP EPS $0.80 (+6.7% beat). Management raised FY27 guide to ≈$11.5B revenue (+40% YoY) and FY28 to ≈$16.5B, per TIKR. "Exceptional AI bookings" cited by CEO Matt Murphy.

• Nvidia $2B Equity Investment + NVLink Fusion (March 31, 2026): Nvidia invested $2 billion in Marvell and signed a broad partnership connecting Marvell's custom XPUs and optical interconnect to the Nvidia AI Factory via NVLink Fusion. Stock surged 13%+ on the news, per TradingKey.

• Jensen Huang's "Trillion-Dollar" Computex Moment (June 2, 2026): Nvidia CEO Jensen Huang joined Marvell CEO Matt Murphy on the Computex 2026 stage and publicly predicted Marvell would become the next trillion-dollar company, per CEOWORLD. Stock spiked to $333.50.

• Hyperscaler ASIC Design Wins: Marvell co-designs Amazon's Trainium and Microsoft's Maia accelerators and has a multi-generational five-year supply agreement with AWS, per TS2. Together with Broadcom, they control ≈95% of the custom AI ASIC co-design market.

• Acquisitions Closed (Feb 2026): Celestial AI (photonics) and XConn Technologies (interconnect) both closed in early 2026, strengthening the optical-interconnect roadmap.

• Stifel Target Raise (June 2-3, 2026): Stifel raised its MRVL target to $321 from $230 with a Buy rating.

🚀 Upcoming Catalysts (Next 6 Months)

• 📅 Q2 FY2027 Earnings — August 20, 2026 (after close): Per Investing.com, this is the next major hard catalyst. Management's Q2 guidance: revenue $2.7B ±5% (≈+35% YoY), Non-GAAP EPS $0.93 ±$0.05. Watch: data-center revenue trajectory toward the ≈$11.5B FY27 run-rate, Trainium3 content share, 1.6T optics adoption, and gross-margin trend toward 59%+.

• NVLink Fusion / Silicon Photonics productization: Ongoing milestones from the March 2026 Nvidia partnership expected through 2H 2026, per StockTitan. Product disclosures could re-catalyze the stock.

• Trainium3 Ramp (2026): Marvell's content share in Trainium3 (slated for volume production in 2026) is the key swing factor, per TS2. Any intel on Alchip's competing Trainium3 bid is a risk-to-watch.

• Q3 FY2027 Earnings (≈late November / early December 2026): The follow-on print testing whether AI bookings continue to compound at the pace management guided.

---

💡 Trading Ideas

🛡️ Conservative — "Wait at the $280 Gamma Floor"

For investors with $5K-$25K, new to options or MRVL

Today's deep-ITM close tells us nothing negative about MRVL's forward outlook — it's mechanics, not conviction. If you believe the AI co-design story, the $280 gamma floor is your near-term defensive entry level. Market makers will defend this zone aggressively (4.94B total GEX). A pullback to $280, hold, and bounce would be a clean technical entry into the stock outright.

• 🎯 Entry: $280 test-and-hold
• 📈 First target: $290-$300 zone (Strong GEX resistance — a breakout here is the signal)
• 🛡️ Stop: below $270 (puts gamma support in doubt, next floor at $270 is the last major one before a deeper reset)

Why this works: You're buying at the biggest near-term support concentration, with defined risk and a clear catalyst (August 20 Q2 earnings) to trade into. No options premium risk.

---

⚖️ Balanced — "The Q2 Earnings Straddle Setup"

For swing traders with $10K-$50K, 2-3 month horizon

MRVL's implied move for the August 21 OPEX window is ±$140 (±≈50%). That's an enormous range. With Q2 earnings on August 20 and the stock at ≈$282, options are pricing in a massive catalyst move in either direction. If you believe the Q2 beat is likely (revenue beat + Trainium3 commentary), a bull call spread targeting the $300-$320 gamma zone above captures upside without paying for the full implied-move range.

Illustrative structure (verify live prices before executing):

• 📈 Buy MRVL $300 Call, 2026-08-21 expiration
• 📉 Sell MRVL $320 Call, 2026-08-21 expiration
• 💰 Net debit: ≈$8-$12 per spread (estimate; the $300-$320 zone covers the strongest call-gamma resistance)
• 🎯 Max profit: ≈$8-$12 per spread if MRVL is above $320 at August 21 expiry
• ⚠️ Max loss: net debit paid

Why this works: The $300 and $320 gamma levels are the two heaviest call-gamma walls above spot. A Q2 beat with strong Trainium3 / NVLink Fusion commentary could punch through both. The spread limits damage if MRVL stays pinned in the $280-$300 digestion zone.

---

🚀 Aggressive — "YOLO the Trillion-Dollar Story"

For experienced options traders, 2-4 month horizon, $3K-$10K budget

If you're convicted on the Jensen Huang "trillion-dollar" thesis and believe Q2 earnings on August 20 catalyzes a re-test of the $333.50 all-time high (and potentially beyond), the August 21 $300-$330 call debit spread gives you aggressive upside at defined risk.

Illustrative structure:

• 📈 Buy MRVL $300 Call, 2026-08-21 expiration
• 📉 Sell MRVL $330 Call, 2026-08-21 expiration
• 💰 Net debit: ≈$10-$15/spread (estimate; verify live)
• 🎯 Max profit: ≈$15-$20/spread if MRVL reclaims $330+ into August 21 expiry (covers the former all-time-high zone)
• 💀 Max loss: net debit — full premium at risk if MRVL stays below $300

YOLO version: 5-10 contracts = ≈$5K-$15K for a bet on the MRVL breakout story through Q2.

Critical warning: MRVL is a ±57% implied-move stock on a quarterly basis — it can rip or collapse in either direction. Alchip's competing Trainium3 bid is a binary content-share risk. At ≈99x P/E, any execution slip triggers outsized downside. Only do this with money you can afford to lose.

---

🎲 Price Targets & Scenarios

Using gamma levels, implied-move ranges, and the August 20 earnings catalyst:

📈 Bull Case (MRVL ≈35-40% probability through Aug 21)

Target: $320-$338

How we get there: Q2 FY27 beat (≈$2.8-3.0B revenue vs $2.7B guide), CEO commentary confirming strong Trainium3 content share and accelerating NVLink Fusion product momentum, gross margin trending toward 60%+. The ±$140 implied move suggests the market already assigns meaningful probability to a retest of the $333.50 all-time high. Bull case targets: $300 gamma resistance break (first signal) → $310/$320 zone → $333-$338 former ATH.

🎯 Base Case (≈40-45% probability)

Target: $280-$300 range

MRVL digests the Computex euphoria. Q2 reports in-line with the ≈$2.7B guide but doesn't blow it out. Trainium3 content-share clarity improves but Alchip competition keeps the market cautious. Stock oscillates in the $280-$300 gamma corridor through summer. Gamma walls at $290 and $300 act as natural caps; $280 and $270 act as floors. Base case: grinding sideways into August, then Q2 earnings is the directional trigger.

📉 Bear Case (≈20-25% probability)

Target: $230-$270

AWS signals a pause in AI capex, or Alchip wins a larger-than-expected Trainium3 socket share. Customer-concentration risk (≈76% of data-center revenue from one customer, per TS2) bites hard. At ≈99x P/E, a revenue-growth deceleration triggers a brutal de-rating. The $270 gamma support (4.51B total GEX, near-balanced call/put) is the critical level; a close below $270 opens the door to the ±18.8% weekly implied move lower, which targets as low as ≈$229.

---

⚠️ Risks & Honest Limits

What the tape and analysis CANNOT tell us:

• Open vs. close is not PROVEN for 0DTE: Vol 2,309 < OI 4,300 / 3,600 — size alone cannot confirm this is a close rather than a new short. The 0DTE intrinsic pricing and multi-leg auction structure make a closing exit the overwhelmingly likely explanation, but we cannot rule out a paired institutional structure we don't have full visibility on. Do NOT trade on this analysis assuming the STC interpretation is confirmed fact.

• Who held the long — and why they closed today: The tape shows a SELL but cannot tell us who originally owned these calls, at what cost basis, or whether this was a standalone close vs. one leg of a larger portfolio restructuring. The $34M in proceeds is mostly intrinsic value — how much profit it represents depends entirely on the original entry price, which we cannot see.

• The 🤝 BLOCK CROSS framing: This was a negotiated, facilitated multi-leg auction — not aggressive lit-market selling. A known counterparty was already matched on the other side. The seller did not "slam" the market; they executed an orderly exit through a broker. The language should be: "crossed," "monetized," "closed," not "dumped" or "fled."

• No continuation signal: This trade does NOT signal that the desk is now short MRVL or is bearish on the stock. It signals they are no longer long these specific 0DTE contracts — which would have expired worthless or triggered automatic exercise at the close anyway.

Structural risks for MRVL forward:

• 🎯 Customer concentration: ≈76% of data-center revenue from a single customer (widely understood to be AWS), per TS2. A spending pause or supplier shift is a binary event.
• 🤖 Alchip content-share risk: Morgan Stanley and Melius Research have flagged Alchip as a potential Trainium3/4 winner, which could erode Marvell's AWS revenue, per TS2.
• 💸 Valuation: At ≈99x P/E, execution must be flawless. Any AI-capex air pocket hits growth and multiple simultaneously.
• 📉 Post-Computex overshoot: Average analyst price targets at ≈$233 (per MarketBeat) sit far below the ≈$282-$309 current trading zone — sentiment has outrun published models, creating a potential mean-reversion setup if Q2 guidance disappoints.

---

🎯 The Bottom Line

Here's the deal: A desk sold two blocks of deep-in-the-money MRVL calls that expired today. Both trades printed at essentially pure intrinsic value — $77.65 on the $215 strike and $70.15 on the $222.5 strike — which tells you exactly what was happening: someone was converting expiring long-call profits back into cash before the bell. This is the options-market equivalent of selling stock to realize a gain on expiration day. The $34M number is eye-catching, but most of it is the MRVL rally since those calls were originally purchased, not $34M of fresh capital changing hands.

What this trade tells us about MRVL:

• ✅ A desk held MRVL call exposure at the $215-$222.5 level — that means they entered MRVL when the stock was trading near or below those strikes (potentially months ago), and the position was so profitable by today that the calls had ≈$70-$77 of pure intrinsic value. That's a record of MRVL bulls being RIGHT, not a sign they're now bearish.
• 🤝 The block was crossed via a negotiated auction, not urgently dumped — no sign of panic or concern.
• 📅 The dominant forward catalyst is August 20, 2026 Q2 earnings — that is where the next major directional move is likely to originate.

If you own MRVL: Today's action is noise from an expiring options cycle. The gamma structure ($280 support / $290-$300 resistance) and the looming August 20 earnings are what matter now. Watch whether MRVL can reclaim $290 and hold — that level's strong gamma wall (6.39B total GEX) is the line between "digesting Computex" and "breaking out again."

If you're watching from the sidelines:

• 📅 The $280 gamma floor is the near-term entry signal — a test and hold is your risk-managed entry into MRVL's AI story
• 📅 August 20, 2026 — Q2 FY27 earnings after the close (the first major confirming or refuting catalyst since the Computex surge)
• 🔑 The $300 gamma wall is the ceiling to beat — a sustained close above $300 is the technical green light for a move back toward former highs

Mark your calendar:

• 📅 June 12, 2026 — Weekly OPEX; implied ±$53 range ($229-$335)
• 📅 June 19, 2026 — Triple Witch OPEX; range ±$68 ($213-$351)
• 📅 July 17, 2026 — Monthly OPEX; implied ±$103 range ($179-$385)
• 📅 August 20, 2026 — Q2 FY2027 earnings after close (the dominant catalyst)
• 📅 August 21, 2026 — Monthly OPEX day immediately following earnings

Final verdict: Today's trade is a textbook 0DTE deep-ITM position exit — a financially rational, mechanically routine monetize of expiring long-call exposure. It is NOT a bearish statement about MRVL. The real story here is the context: whoever held those calls at $215/$222.5 was sitting on ≈$70-$77 per share of profit — a tribute to just how far MRVL has run in 2026. Forward-looking, everything hinges on August 20 Q2 earnings and whether the AI co-design execution story justifies the ≈99x P/E the market is paying.

---

Disclaimer: Options trading involves substantial risk of loss and is not suitable for all investors. Deep-in-the-money calls on expiration day carry near-100% delta and can result in automatic exercise if not actively managed; selling them at intrinsic converts the position to cash but is only available to holders who have the options to sell. The STC (sell-to-close) classification in this analysis is provisional (⏳) — volume < open interest means we cannot prove open vs. close from the tape alone on a 0DTE instrument. This analysis is for educational purposes only and is not financial advice. Past unusual options activity does not guarantee profitable trading outcomes. MRVL is a high-volatility stock with a ±57% quarterly implied move — strategies involving MRVL options carry the risk of total premium loss. Always do your own research and consult a licensed financial advisor before trading.

---

Last updated: 2026-06-05

About Marvell Technology: Marvell Technology designs custom AI accelerators (XPUs), optical interconnect, and data-center networking silicon. Market cap ≈$270B. Sector: Semiconductors — Custom AI ASICs. The company co-designs Amazon's Trainium and Microsoft's Maia AI chips and received a $2B strategic equity investment from Nvidia in March 2026, cementing its role as a critical node in the AI data-center supply chain.