🛡️ Ainvest Option Flow Digest - 2026-06-29 — The Hedging Day: Smart Money Buys Downside Insurance and Structures Defined-Risk Spreads Into the FOMC + Mega-Cap Gauntlet

Eleven names, ≈$172M in options premium (plus ≈$37M of paired stock on three delta-hedged crosses) — and the story is how little of it is a naked directional bet. We pulled the OPRA tape and the stock tape on every ticket, and it reshaped the read on eight of the eleven: three "bullish/bearish" prints were really QCC packages with hidden stock legs (WMB 126K, CRWV 96K, IGV 205K shares), two "naked-put" headlines were actually defined-risk vertical spreads (QCOM credit, SNOW debit), a "$62M deep-ITM call buy" first read as position management but next-day OI inverted it to a confirmed fresh opening long, rolled past earnings (TSM), and one big "call buy" the scanner tagged a SELL was, on the tape, a buy (CRWV). Strip the labels away and the real tape is hedged and downside-leaning — a ≈$19M Nasdaq-100 put timed to the July FOMC, a 1-year IGV software-downside hedge, a SNOW bear spread, a CRWV protective-put package — set against a couple of genuine bullish recovery bets (AVGO, NBIS).

🔁 Update (2026-06-30) — This Session's ⏳ Flags Are Now Resolved

The next-day OPRA open-interest snapshot for 2026-06-29 is in, and it inverted three of the reads above:

• 🔄 TSM — the ≈$62M deep-ITM $400 call cross we called "position management, likely a close" was a CONFIRMED FRESH OPEN. OI rose 6,724 → 14,722 (Δ +7,998 ≈ the full 8,000 block). Paired with the June 26 close of the Jul-17 $400 call, the desk rolled its $400 long forward from July into the September quarterly — extending bullish deep-ITM exposure past the July 16 earnings, not winding down. TSM
• 🔄 AMZN — the ≈$22.3M 1.70:1 call ratio spread was NOT a fresh open. Despite prints of 90,814 ($290C) and 53,420 ($250C), aggregate OI barely moved ($290C +620, $250C −502) — an existing ratio spread transferred between holders, adding ≈zero new market exposure. AMZN
• 🔄 QCOM — the "mild lean toward closing" the $220 put was ruled out. Both legs opened ≈ full size ($220P +4,917, $170P +4,955 on 5,134-lot prints) — a genuine fresh bull put spread (STO $220P / BTO $170P). QCOM

✅ Confirmations: QQQ $695P (+17,730), NBIS $350C (+14,396, larger than the block), WMB $80C (+3,000, clean), IGV $90P (+5,001), CRWV $75P (+3,943) / $97C (+1,502), ACLS $250C (+1,954) and SNOW's bear spread ($120P +4,657 full / $160P +1,486 partial) all OPENED as written. AVGO is a nuance: the long Oct $410C opened in full (+25,883) but the short Jul $360C OI fell (−7,484) — a close, making the package a bullish roll up-and-out, not a fresh diagonal short. The bullish read holds. Full per-name detail: _OI_RESOLUTION_NOTES.

🔁 OI Review — Last Session's Provisional Flags, Now Resolved (covers 2026-06-26)

The next-day OPRA open-interest snapshot for the 2026-06-26 session is in. We pulled it for all 30 legs across 19 names and resolved every ⏳ provisional open/close flag. The headline: 16 names OPENED as written, but 3 INVERTED — and the inversions matter, because they were three of the session's biggest "premium" headlines. ≈$100M of trade-day premium (AVGO $18M + TSM $74M + TMUS $8.2M) created zero new directional or short interest: AVGO's "leveraged-long" call BUY was a CLOSE, and the TSM/TMUS "sales" were closes/transfers that opened no new shorts. The standing lesson came alive this session: a big headline premium is not conviction — or even a new position — until next-day OI proves it.

🔄 Inversions

• TSM — the $74M "earnings-IV harvest" (sell deep-ITM $400 call + $350 put) was framed as a fresh premium-collection open. It inverted: OI fell on the $400 call (22,944 → 21,291, Δ −1,653) and was flat on the $350 put (21,779 → 21,521, Δ −258). Across 30,000 contracts, no new short interest was created — a close / transfer (STC / unwind), not a new IV-harvest open. TSM
• AVGO — the $18M deep-ITM $100 call BUY was framed as a leveraged-long "buy the dip." It inverted: OI fell 720 → 471 (Δ −249). A deep-ITM call BUY where OI drops is a buy-to-close / unwind, not fresh bullish accumulation. AVGO
• TMUS — the $8.2M Aug $195 call SELL was framed as a capped-upside overwrite. It inverted: OI flat (16,282 → 16,309, Δ +27) against a 16,000-contract sale — no new short interest. A close or transfer, not a new income trade. TMUS

✅ Confirmations

The genuine new vol-selling held — SMH ($100M put-sale, OI +51.5K/+49.8K), IGV ($74M, +99.9K/+100K), NOW ($22M, +11.1K/+11.1K) all OPENED (STO). The QQQ delta-hedged complex opened all three legs (stays non-directional). Bullish opens confirmed on UBER (all 3 legs), VSAT (bull call spread), ASTS, DOCN, COHR; bearish/protective opens on RIOT, BURL, AXTI; hedged/covered opens on GS, CART, CAPR, FOXA. Full per-name detail: _OI_RESOLUTION_NOTES.

What changed: strip out the three inversions and the day's real new directional/short flow was meaningfully smaller than the ≈$280M "premium SOLD" headline. The lesson, again: a big BUY headline is not bullish conviction and a big SELL headline is not new premium-collection until next-day OI proves the position opened.

⚡ The Quick Read

Today's tape spanned 11 names for ≈$172M in options premium — but the headline labels are misleading, and the real story only shows up after you pull the stock tape (the single biggest ticket, TSM's $62M, turns out to be ≈zero new direction). It sorts into four themes:

1) Downside insurance into the late-July macro gauntlet — software, semis & index. The loudest real signal is hedging ahead of the events. QQQ drew a ≈$19M Nasdaq-100 $695 put (Jul-31) — a single-leg auction with no stock leg, deliberately expiring right after the July 28–29 FOMC (hawkish under new Chair Warsh) and the July 29–30 mega-cap earnings wave, but before NVDA. IGV placed a ≈$23.9M QCC package — 5,000 one-year $90 ATM puts (≈$5.5M) delta-hedged with a 205,000-share stock block — a 1-year downside hedge on the software complex, itself a flip from the $74M put-sale on the same ETF three sessions ago. SNOW put on a ≈$2.33M bear put spread (long $160 / short $120, Jan-2027) — defined-risk downside, and a notable flip from its bullish $260-call flow just three sessions ago.

2) Crosses and packages where the headline misleads. The day's single biggest ticket, TSM's ≈$62M deep-ITM $400 call block cross, was framed on the trade day as likely position management — but the next-day OI inverted that read: it was a CONFIRMED FRESH OPEN (OI 6,724 → 14,722, Δ +7,998 ≈ the full block). Paired with the June 26 close of the Jul-17 $400 call, the desk rolled its $400 long forward from July into September, extending bullish deep-ITM exposure past the July 16 earnings — a genuine (negotiated, known-counterparty) directional position, not housekeeping. Two more "trades" are really stock positions in disguise: WMB's "$1.2M bullish $80 call" printed with a 126,000-share stock block ($9.55M) 252ms later — a QCC delta-hedge, direction ambiguous; and CRWV's ≈$7.4M $75 put came paired with a 96,000-share block ($8.94M) — a delta-neutral long-vol / protective package, not a bearish bet (and a separate $1.1M call there was actually bought, not sold as the scanner labeled it).

3) Defined-risk spreads, not naked premium (≈$40M structured). The pros structured rather than gambled. QCOM sold a ≈$15.4M net-credit bull put spread (short $220 / long $170) — mildly bullish, max loss capped at $10.3M, not two naked short puts. AMZN worked a ≈$22.3M 1.70:1 call ratio spread targeting ≈$290 by Jan-2027 — but next-day OI showed it was a TRANSFER, not a fresh open: aggregate OI barely moved on either leg ($290C +620, $250C −502 vs prints of 90,814 / 53,420), so an existing ratio spread changed hands rather than new conviction going on. The geometry is still range-bound with a hard ceiling above ≈$341 — but read it as a position transfer, not a new bet.

4) The genuine directional longs (≈$36M) — bullish AI/semis bets. AVGO layered a ≈$33.3M bullish diagonal (long Oct $410 / short Jul $360) into Broadcom's fall recovery. NBIS bought a ≈$1.7M deep-OTM $350 call expiring in 11 days — an aggressive, low-odds lottery on Nebius. And ACLS worked a ≈$1.1M deep-OTM bullish diagonal call spread (Jul $200 / Aug $250) on Axcelis into its August earnings + Veeco-merger window.

Translation: this is a hedge-and-structure tape, not a conviction-buying one. When the clean directional flow is small, the spreads are defined-risk, and the biggest unstructured ticket is index downside insurance timed to the Fed — that's a desk community managing risk into a known event window, not chasing. 🧭

👀 Today at a Glance

Premiums are per-structure (🤝/🛡️ packages note the paired stock; 🤝 QCOM is a net credit; all others are debits). Day total ≈$172M in options premium across 11 names, plus ≈$37M of paired stock on the three QCC packages (WMB + CRWV + IGV) — but note ≈$62M of that headline (TSM) is a position-management cross with ≈zero new directional exposure, so the real new conviction is far smaller. 🛡️ = tape-verified hedged/insurance. ⏳ = come back next session pre-market (≈06:30 ET) for the OPRA OI that confirms open vs. close.

🔬 The Standouts, In Plain English

🛡️ QQQ — ≈$19M of Nasdaq-100 Insurance, Timed to the Fed

The cleanest unstructured ticket of the day is a hedge. QQQ saw 15,678 July-31 $695 puts bought for ≈$19M — and the tape tells you two important things. First, it printed as a single-leg auction with no stock leg, so unlike the June-26 QQQ crosses (which were delta-hedged with stock), this is a clean, standalone put. Second, the expiry is no accident: July-31 lands right after the July 28–29 FOMC (hawkish, with a live September-hike risk) and the July 29–30 mega-cap earnings wave (MSFT/META/AMZN/AAPL), but before NVDA on August 26. With QQQ near $723 at a ≈35x P/E and ≈30% concentrated in five names, a desk paid ≈$19M for one month of protection across the exact window the calendar is most dangerous. Read it as insurance, not a doomsday call — but the choice to insure is the signal.

🤝 WMB & CRWV — "Trades" That Are Really Stock Positions

Two prints look directional until you pull the equity tape. WMB's $80 call "buy" ($1.2M) printed with a 126,000-share stock block ($9.55M, a Qualified Contingent Trade) 252 milliseconds later — a delta-hedged QCC package where the stock leg is 7.9× the option premium. The direction is genuinely ambiguous: we can't tell from the tape which side bought the stock, so this is not a clean bullish Williams bet. CRWV is subtler still: its ≈$7.4M $75 put came paired with a 96,000-share block ($8.94M) that delta-neutralizes the put — a long-volatility / protective package, not a bearish bet on CoreWeave. And the separate $1.1M $97 call there was bought at the offer (the scanner labeled it a "sell," but the tape disagrees) — a small, genuinely bullish near-term call. Net CRWV read: neutral-to-mildly-bullish, hedged. Neither is the directional trade its headline implies.

🤝 QCOM & SNOW & 🎯 AMZN — Defined-Risk Structures, Not Naked Bets

Three more were mislabeled by size alone and are really spreads. QCOM is a ≈$15.4M net-credit bull put spread (sell the $220 put, buy the $170 put) — mildly bullish on the post-Investor-Day "Dragonfly" recovery, with a capped $10.3M max loss, not two open-ended naked puts. SNOW is a ≈$2.33M bear put spread (long $160 / short $120) — a defined-risk downside bet that's a sharp flip from the bullish $260-call flow we flagged on June 25. AMZN is a ≈$22.3M 1.70:1 call ratio spread targeting ≈$290 by January 2027 — with a hard ceiling: it turns into a loss above ≈$341 and runs net-short delta. A precise, range-bound view, not "loading calls" — and next-day OI confirmed it was an existing spread transferred (OI ≈flat both legs), not fresh new positioning.

🐋🎰 The Genuine Bullish Bets — AVGO, NBIS & ACLS

The clean directional longs are the minority. AVGO layered a ≈$33.3M diagonal — long the October $410 calls, short the July $360 calls — a bullish bet that Broadcom recovers into its September quarter after the post-earnings dip (note: these printed as late-reported trades, so we framed the aggression carefully, not as a sweep). NBIS is the day's lottery ticket: ≈$1.7M of deep-out-of-the-money $350 calls expiring in 11 days on Nebius (spot ≈$260) — it needs a fast ≈35% move and will likely expire worthless, but it's a cheap, convex, genuinely bullish swing. ACLS is a similar long shot in chip-equipment maker Axcelis: a ≈$1.1M deep-OTM diagonal call spread (the screenshot showed only the August $250 call, but the tape reveals a paired July $200 call) — a bullish bet on a stock near 52-week highs into its August earnings and the pending Veeco merger, though on an electronic combo we can't prove which leg is long vs short.

📅 Upcoming Catalysts — Mind the Gap Between the Event and the Expiration

🧑‍🤝‍🧑 For Four Kinds of Traders

🎰 YOLO / high-risk: Your convex play is NBIS's 11-day $350 call — cheap, lottery-like, and almost certainly a total loss unless Nebius gaps ≈35% in two weeks; size it as money you can lose entirely. The CRWV $97 call (bought, July-24) is a shorter-dated bullish nibble. Do not copy the big structured trades (QCOM/SNOW spreads, WMB/CRWV QCC packages) at YOLO size — they're hedged or defined-risk for a reason, and the naked version of them has very different math.

📈 Swing trader: The actionable read is the hedging: a ≈$19M QQQ put timed to the FOMC plus a SNOW bear spread and a CRWV protective-put package say desks are paying up for downside into late July. If you're long AI/tech, that's worth respecting. For clean direction, AVGO's diagonal is the cleanest bullish recovery expression; SNOW's bear spread is a defined-risk way to fade an extended name.

💰 Premium collector: Study QCOM's bull put spread — it's the textbook of how to sell premium with a net before the bell: collect the rich $220-put credit, but buy the $170 put so your max loss is a defined $10.3M instead of open-ended. That single long leg is the difference between an income trade and a blow-up. (Contrast it with the naked two-put sales that have bitten sellers in past sessions.)

🌱 Beginner / new to flow: Today's lesson is that the headline label is often wrong. A "call buy" (WMB) can be a delta-hedged stock package; a "put buy" (CRWV) can be protection, not a bearish call; and a scanner can even mislabel a buy as a sell (CRWV's $97 call). Two habits beat the headline every time: (1) check whether the option print has a stock trade on the other side, and (2) check whether it's a spread (two legs) rather than one naked option. Both completely change what the trade means.

⚠️ Risk & Patience — Read This Before You Trade Any of It

• Most of today is hedged or defined-risk — and downside-leaning. The biggest unstructured ticket (QQQ) is insurance, SNOW is a bear spread, and CRWV's big put is protection. Desks are buying downside into a hawkish-FOMC + mega-cap-earnings window. The prudent takeaway isn't to panic-sell — it's to make sure your own risk is sized and hedged before the events.
• A QCC cross is NOT a directional bet. WMB's and CRWV's option "trades" each printed with a large stock block on the other side; the stock leg dominates and the package is delta-managed. Don't read them as bullish/bearish conviction.
• Spreads cap risk; ratios and naked legs don't. QCOM (credit) and SNOW (debit) are defined-risk. AMZN's ratio spread, by contrast, has uncapped loss above ≈$341 from its extra short calls — more bullish is not better there.
• Open vs. close is still provisional on the size-≤-OI legs (QCOM, SNOW, CRWV, WMB). A big print under existing open interest can be opening or closing — next-day OI (≈06:30 ET) is the only proof. Come back tomorrow for the resolution.

Not investment advice. Options carry risk of total loss. Selling options can lose far more than the premium collected, and ratio spreads carry open-ended risk above the upper strike. Unusual flow is a starting point for research, not a recommendation. Always do your own work and size positions you can afford to lose.