📊 Ainvest Option Flow Digest - 2026-06-04 — Desks Buy Protection and Sell Premium: Puts, Spreads & a $22M Straddle

📅 Thursday, June 4, 2026 | 🔥 7 institutional structures, ≈$68M net (≈$84M gross notional)

Last updated: 2026-06-05

🔁 Next-Day OI Resolution (2026-06-05). The 6/5 OPRA OI snapshot is in. ❗ One inversion: CRWV $100C was a CLOSE/profit-take (OI −3,547), not the fresh ITM overwrite first read — a long-call holder taking profit after the Vera Rubin pop. ✅ Confirmed opens (6): APH (+1,918, bullish), CCL (+19,749, bearish/hedge), CDNS (370P +14,977 — heavy bearish/hedge accumulation; 320P flat), DXCM (+1,902), MRVL (+1,006, willing-to-own), VOYG (both legs ≈+4,200, short straddle). Every 6/4 article carries a ✅ RESOLVED box + Last updated: 2026-06-05 stamp.

⚡ The Quick Read

If 6/3 was a "structure day," 6/4 was a defense-and-income day. The tape split cleanly into two camps: desks buying downside protection (CCL puts, a CDNS bear put spread, a DXCM LEAP put) and desks selling rich premium (a CRWV ITM call-write, a deep-ITM MRVL put sale, and a ≈$22M VOYG straddle). Only one outright bullish call bet (APH) crossed. As always — these are negotiated blocks with known counterparties, not panic sweeps; read the structure, not the headline dollar.

• 🐻 CDNS — the standout narrative: a ≈$13.4M-net bear put spread ($370/$320, Aug-21) on Cadence at record highs. And it rhymes with yesterday — 6/3's CDNS $400 call resolved as a CLOSE/profit-take; today the same name gets downside protection. A two-day de-risking arc on a stock at all-time highs.
• 🐻 CCL — ≈$4.8M of Sep $26 puts on Carnival into its late-June Q2 earnings (fuel headwind + consumer-spending worry).
• 🐻 DXCM — ≈$1.8M of Jan-2028 $60 puts on Dexcom — a multi-year bearish/tail-hedge against the GLP-1 + Abbott overhang.
• 📤 CRWV — (updated 6/5) resolved as a CLOSE/profit-take at the $100 strike (OI −3,547), not a fresh overwrite — a long-call holder ringing the register after the Vera Rubin pop.
• 🤝 MRVL — ≈$13M deep-ITM put sale ($350P, 2027): a desk gets paid to be willing to own Marvell at an effective ≈$217 (≈30% below spot). Bullish/willing-to-own.
• 🛰️ VOYG — a ≈$22M short straddle ($60, 2028) on space name Voyager: selling rich long-dated volatility, betting it pins near $60 (undefined risk — pros only).
• 🤝 APH — the lone clean bull: ≈$1.7M of Aug $150 calls on Amphenol into its July 29 earnings (AI-connector demand).

Headline premium is NET (long legs paid − short legs collected). Gross across all legs was ≈$84M; the ≈$68M net is capital at risk after the CDNS spread nets out. Don't blindly follow blocks — half of today's flow is protection and income, not directional conviction.

🔁 OI Review — Last Session's Provisional Flags, Now Resolved (Updated 2026-06-04)

The next-day OPRA OI snapshot for Wednesday 2026-06-03 is in. THREE reads FLIPPED: CDNS's "$24M ambiguous block" was a CLOSE/profit-take (STC); EWZ's "fresh bull call spread" was actually a call-strike ROLL up $40→$45; HOOD's "willing-to-own put-sale" was a CLOSE/position-reduction. The day's big open question — LRCX's bull-vs-bear direction — resolved BULLISH (a confirmed roll up-and-out). Plus 7 clean confirmations (GLD, ON, NVTS both legs, INTC, LYB, WULF, IWM) and 2 tweaks (MRVL, SMH). Every 6/3 article carries a ✅ RESOLVED OI box and a Last updated: 2026-06-04 stamp.

🔄 Inversions — These FLIPPED Versus Wednesday's Read

• CDNS 6/3 — $24M $400C 8/21: "ambiguous block-cross" → CONFIRMED CLOSE / profit-take (STC): OI 11,871 → 8,783 (Δ −3,088). OI FELL, so the SELL of 5,500 $400 calls was a long-call holder closing / ringing the register near the all-time high — NOT premium collection and NOT a fresh short. The provisional "could be STO or STC" resolved decisively to STC. A big SELL headline at record highs was an exit, not a new position.
• EWZ 6/3 — $5.2M $40/$45C 12/18: "fresh bull call spread" → CONFIRMED call-strike ROLL up $40→$45: $40C OI 127,653 → 82,128 (Δ −45,525, closed) AND $45C OI 72,470 → 120,794 (Δ +48,324, opened). The "BUY 40C" leg bought to CLOSE; the "SELL 45C" leg sold to OPEN. Net: the desk rolled its call exposure UP one strike ($40 → $45) — a bullish-leaning position adjustment, not a newly-opened defined-risk debit spread. The October-Brazil-election thesis stays as backdrop; the structure was a roll.
• HOOD 6/3 — $2.1M $95P 7/17: "willing-to-own put-sale open" → CONFIRMED CLOSE / position reduction: OI 6,248 → 2,780 (Δ −3,468). OI FELL sharply, so the $95-put activity was unwinding existing open interest, not a fresh willing-to-own short-put open. The "getting paid to own Robinhood at an effective ≈$80.55" framing does not apply — this was an exit at the strike, not an entry.

🎯 The Day's Big Question, Resolved

• LRCX 6/3 — $1.5M call-spread ladder: direction unproven → CONFIRMED BULLISH: the 6/05 legs closed ($330C −448, $340C −1,109) while the 6/12 legs opened ($350C +1,463, $360C +1,738). That is a textbook bullish roll up-and-out — close the near-week 330/340 spread, open a higher-strike, later-dated 350/360 spread. The bear/credit alternative we flagged is ruled out; the mildly-bullish lean is confirmed.

✅ Confirmations — The Original Read Held

• GLD 6/3 — $15M $415C 2028 covered-call STO: OI 467 → 3,393 (Δ +2,926). Buy-write opened as predicted.
• ON 6/3 — $7.9M $190C 2028 call-write STO: OI 43 → 2,561 (Δ +2,518). Far-OTM premium-collection open confirmed.
• NVTS 6/3 — $3.1M $32 collar (both legs): $32C OI 814 → 1,825 (Δ +1,011) and $32P OI 41 → 1,041 (Δ +1,000). Collar/synthetic-short opened on both legs.
• INTC 6/3 — $1.1M $155C 10/16 BTO: OI 557 → 1,595 (Δ +1,038). Long-call lotto opened.
• LYB 6/3 — $1.5M $55P 1/27 BTO: OI 307 → 3,204 (Δ +2,897). Bearish/hedge long-put opened.
• WULF 6/3 — $2.6M $35C 9/18 BTO: OI 372 → 10,629 (Δ +10,257). AI/HPC-pivot long-call opened.
• IWM 6/3 — $0.8M $277/$275 bull put spread: long $275P OI 87,068 → 112,194 (Δ +25,126); the short $277P leg's OI was absorbed by closers (−2,448). Bull put credit spread opened.

⚠️ Tweaks — Partial Corrections

• MRVL 6/3 — $15M diagonal: the 2028 $300C opened fresh (Δ +1,010), but the Aug $165C OI fell 1,266 → 826 (Δ −440) — the near leg was partly a CLOSE, so this is a bullish-leaning roll/restructuring, not the clean "free leveraged long" first described.
• SMH 6/3 — $1.55M bullish combo: the short $600P opened (Δ +1,802, willing-to-own confirmed), but the $625C was OI-flat (Δ +17 vs size 249) — the upside call was NOT a fresh long. The engine of the trade is the put-sale.

What changed in the post-resolution read

Pre-resolution Wednesday PM: ≈$63M net across 13 structures, framed as a mostly-bullish/income "structure day" with several ⏳ provisional opens. Post-resolution Thursday AM: three reads flipped — CDNS ($24M) moved from ambiguous to a closing profit-take, EWZ from a fresh bullish spread to a strike roll, and HOOD from a fresh bullish open to a close. The day's marquee question — LRCX's direction — resolved bullish (roll up-and-out). The premium-collection and long-premium opens (GLD, ON, NVTS, INTC, LYB, WULF, IWM) all confirmed.

Standing lesson (loud again): when a SELL prints below open interest (Vol ≤ OI), the classifier's "more likely closing" lean is the warning — CDNS, EWZ, and HOOD all looked like fresh positioning intraday and all resolved as closes/rolls on the next-day OI. A big BUY or SELL headline is not direction; only the next-morning OI is. Wait for it.

👀 Today at a Glance

Net premium = long-leg premiums paid − short-leg premiums collected (capital at risk). Gross double-counts the two sides of a spread/straddle.

🔬 The Seven, In Plain English

🐻 CDNS — Downside Protection at Record Highs (and a Two-Day Story)

A floor-negotiated block bought 10,000 Aug-21 $370 puts and sold 10,000 $320 puts — a ≈$13.4M net-debit bear put spread, max profit if Cadence falls ≈21% to $320 by August. What makes it notable is the rhyme with yesterday: 6/3's CDNS $400 call block resolved (via OI) as a CLOSE/profit-take near the record high, and today the same name gets a downside hedge. Whether it's an outright bearish bet or a hedge on a long position, the two-day footprint reads as a desk de-risking Cadence at all-time highs ahead of July 27 earnings. 👉 Full CDNS breakdown

🛰️ VOYG — Selling the Space Dream's Volatility

A desk collected ≈$22M selling a Dec-2028 $60 straddle (sell 3,800 puts + 3,800 calls) on Voyager Technologies, the space/defense name (Starlab station + defense tech) that IPO'd last June. Translation: they're monetizing the rich, hard-to-pin implied volatility of a blue-sky space stock, betting it drifts toward and pins near $60 by 2028. The bigger put leg tilts it toward willing-to-own accumulation at an effective ≈$27. The catch: a short straddle has undefined risk — a big move either way hurts. This is a sophisticated, well-capitalized desk's trade, not a retail template. 👉 Full VOYG breakdown

🤝 MRVL — Getting Paid to Own Marvell at a 30% Discount

A desk sold 1,000 deep-in-the-money Jun-2027 $350 puts for ≈$13M — the $350 strike is ≈$41 above spot, so collecting $133 sets an effective entry of ≈$217 if assigned, roughly 30% below today's $309. That's a bullish / willing-to-own stance. It follows yesterday's MRVL bullish diagonal (which OI confirmed as a bullish roll) — a second straight day of constructive Marvell flow. (Forensic note: the tape carried a cancelled duplicate, so the real size is one 1,000-lot print, ≈$13M, not $26M.) 👉 Full MRVL breakdown

📤 CRWV — Resolved: A Profit-Take, Not an Overwrite

❗ 6/5 OI update: the $100 call OI fell 40,799 → 37,252 (−3,547) — this was a CLOSE / profit-take at the $100 strike (likely a long-call holder selling after the Vera Rubin pop), not the fresh ITM overwrite described below. ≈$11M collected selling 10,000 Jun-18 $100 calls (in-the-money, spot ≈$107) on the NVIDIA-backed GPU cloud — an overwrite that caps upside at $100 right after the +14% Vera Rubin validation pop. Reads as income / capped-upside, neutral-to-mildly-bearish into the run. Note it follows yesterday's CRWV $120-call write — the overwrite strike rolling down, a more conservative cap. 👉 Full CRWV breakdown

🐻 CCL — Hedging Carnival Into Earnings

≈$4.8M buying 20,000 Sep $26 puts on Carnival (spot ≈$27.46), positioning for the late-June Q2 earnings binary. The bear case: a ≈$500M unhedged fuel headwind, an EPS reset, and decelerating consumer-discretionary spending — against record operational numbers and a Buy-rated Street. Could be an outright bearish bet or a hedge on a long; CCL printed mid-$23–$24 in late May, so $26 is in range. 👉 Full CCL breakdown

🐻 DXCM — A Multi-Year Bet on the Glucose Overhang

≈$1.8M buying 2,000 Jan-2028 $60 puts on Dexcom (spot ≈$74) — a long-dated bearish/tail-hedge against the GLP-1 demand-destruction narrative and Abbott Libre competition. Dexcom has a history of violent drops (≈−40% in a day in July 2024; a $54 low last November), so a $60 strike sits inside its demonstrated range — though the buyer needs a fresh multi-year low just to break even. 👉 Full DXCM breakdown

🤝 APH — The Lone Clean Bull

≈$1.7M buying 1,250 Aug-21 $150 calls on Amphenol (spot ≈$146), the AI-datacenter connector maker, into its July 29 Q2 earnings. Fresh open (Vol/OI ≈6.7×), breakeven ≈$163.80. The one straightforwardly bullish directional bet of the day, riding the AI-interconnect demand wave. 👉 Full APH breakdown

⏱️ Sorted by Timeframe

• ⚡ Weekly (days out): CRWV Jun-18 $100 ITM call-write (into Vera Rubin headlines + Russell inclusion June 27).
• 🗓️ Monthly (weeks out): CDNS Aug-21 $370/$320 bear put spread (Q2 earnings Jul 27); APH Aug-21 $150C (Q2 earnings Jul 29).
• 📆 Quarterly (months out): CCL Sep-18 $26 puts (late-June Q2 earnings binary inside the window).
• 🦅 LEAP (years out): MRVL Jun-2027 $350 put sale (willing-to-own), DXCM Jan-2028 $60 puts (GLP-1 hedge), VOYG Dec-2028 $60 straddle (vol-sell). LEAPs = patient, thesis-driven, low-theta positions.

🧑‍🤝‍🧑 What Each Type of Trader Should Take Away

🚀 The YOLO Trader The only clean directional lotto is APH calls (bullish into July 29 earnings) — and even that is a measured Vol/OI ≈6.7× open, not a moonshot. The bearish names (CCL, DXCM, CDNS) are protection, not crash bets. Do NOT replicate the MRVL/VOYG/CRWV premium-sells as a YOLO — short puts and short straddles carry large/undefined risk.

📈 The Swing Trader CDNS's bear put spread is a clean event-swing into July 27 earnings with defined risk — a textbook way to play a stretched name without unlimited downside. CCL into late-June earnings and APH into July 29 are the other catalyst-dated swings. Watch the next-day OI to confirm whether CDNS/CCL are fresh opens or hedges before reading them as conviction.

💰 The Premium Collector Today is your buffet — but mind the risk profiles. MRVL (deep-ITM put sale, willing-to-own at ≈$217) and CRWV (ITM call-write) are defined-context income trades if you hold or want the stock. VOYG's short straddle is the danger zone: ≈$22M credit but undefined risk — only for desks that can withstand a big move. Cash-secured or covered only; never naked.

🌱 The Beginner Three lessons today: (1) A put isn't always bearish-betting — CCL/CDNS/DXCM buyers may be hedging longs, not predicting a crash. (2) A SELL can be bullish — MRVL's put seller is willing to own the stock cheaper; that's constructive, not negative. (3) "Net" vs "gross" matters — CDNS looks like ≈$30M of puts but the real capital at risk is the ≈$13.4M net debit of the spread. Read the structure.

🗓️ Upcoming Catalysts (catalyst date ≠ option expiration)

Keep these separate: the event that moves the stock vs. the expiration the desk chose.

• CDNS — 📣 Q2 earnings July 27, 2026. 🎯 Expiration Aug-21-2026 (earnings inside the window).
• APH — 📣 Q2 earnings July 29, 2026 (guided to record $8.1–$8.2B sales). 🎯 Expiration Aug-21-2026.
• CCL — 📣 Fiscal Q2 2026 earnings, late June (the only earnings before expiry; fuel + booking data). 🎯 Expiration Sep-18-2026.
• CRWV — 📣 Vera Rubin headlines; Russell 3000 inclusion June 27; no earnings until August. 🎯 Expiration Jun-18-2026.
• MRVL — 📣 Q2 earnings Aug 20, 2026. 🎯 Expiration Jun-17-2027 (LEAP — earnings is a waypoint).
• DXCM — 📣 GLP-1 / Abbott competition overhang; next earnings late July. 🎯 Expiration Jan-21-2028 (LEAP).
• VOYG — 📣 Starlab / NASA Commercial LEO Destinations milestones (2026–2028); ISS retirement ≈2030. 🎯 Expiration Dec-15-2028 (LEAP).

⚠️ Risk Control & Patience — Read This Before You Copy Anyone

• Half of today's flow is protection and income, not conviction. Bear put spreads (CDNS), protective puts (CCL, DXCM), call-writes (CRWV), and a straddle (VOYG) are defensive/income by design. Copying the SELL legs naked turns a hedged or cash-secured institutional trade into an unlimited-risk retail bet. Don't.
• A short straddle (VOYG) has undefined risk. ≈$22M collected sounds great until the stock moves big either way. This is a well-capitalized desk's trade — not a template to replicate with retail-sized accounts.
• ✅ Open/close now resolved (6/5 OI). CCL opened (bearish/hedge, +19,749), CDNS's 370P opened big (+14,977), and CRWV flipped to a CLOSE/profit-take (−3,547) — exactly why we wait for the OI: a SELL that looked like an overwrite was actually an exit. The next-day OPRA snapshot is the only proof.
• Position sizing beats prediction. Every desk here defined or hedged or got paid for its risk. Patience is a position.

This digest is for educational purposes only and is not investment advice. Options carry substantial risk, including the total loss of premium (for buyers) and potentially unlimited loss (for uncovered sellers and short straddles). Open-vs-close reads marked ⏳ are provisional until the next-day open-interest snapshot confirms them. Always do your own research and manage risk. Past unusual activity does not predict future returns.