Ainvest Option Flow Digest - 2026-05-18: $76M Across 7 Names — Two Earnings Bets Land This Week, One $32M Stock-Replacement, and a $13M Put Someone Sold for Income

A calm read of where large options money went today — and, just as important, where it did not. Position sizing and patience matter more than any single print below.

The Quick Read

We tracked roughly $76M of notable options premium across 7 tickers today. The theme is timing: two of the largest positions sit directly in front of confirmed earnings dates this week, so they are event bets with a clock on them — not slow-burn convictions.

• Biggest single ticket: MSFT — a $32M deep in-the-money call that behaves much like owning the stock with a capped downside (a "stock-replacement" structure), spanning the late-July earnings print.
• Most time-sensitive: NVDA — a $6.4M two-strike call ladder expiring June 1, two trading days after NVIDIA's confirmed May 20 (after close) earnings. This is close to a pure earnings bet.
• The one that resolves first: MDB — a $16.5M paired call structure (≈$1.3M net credit) built around MongoDB's confirmed May 28 earnings, with the long leg expiring May 29.
• Income, not a directional bet: GOOG — a whale collected ≈$13M selling puts at $380 (spot ≈$395). That is a premium-collection / willing-to-own setup, not a bullish "loaded up" trade.
• Longer horizon: CRWV and ARKK both bought January 2027 LEAP calls — multi-quarter theses, not this week's story.
• The lone bearish print: GLXY — a small $1.2M put expiring May 29, essentially a short-dated bet that a 60%+ run cools off.

A reminder before the details: large option prints show you what one participant did with size — not what will happen. Many institutional trades are hedges, rolls, or income strategies, not directional convictions. Read these as starting points for your own work, size positions so a total loss on the premium would not hurt, and never chase a print after the move has already happened.

Today's Flow at a Glance

Premium totals reflect the figures in each detailed analysis. Where a trade is a credit (GOOG, and the net of the MDB diagonal), the table says so explicitly — collecting premium is a very different risk profile from paying it.

The Whale Lineup

1. 🐋 MSFT — $32M Deep-ITM Call, Built Like Stock With a Floor

See why a whale paid $32M for calls that act like Microsoft shares →

• What's happening: $32M into the Aug 21 $390 calls with the stock around $417 — deep in the money (delta ≈0.8+). This is a stock-replacement structure: it captures most of Microsoft's upside while capping the worst case at the premium paid, instead of the much larger capital at risk in owning the shares outright.
• The catalyst: FQ4 FY2026 earnings (estimated ≈July 29, not yet company-confirmed) sits inside the expiry; Microsoft Build is June 2-3. The recent FQ3 print was a clean beat (Azure +40% cc), but the $190B calendar-2026 capex plan is the live bear argument and is covered honestly in the analysis.
• The question: Is this conviction into the AI monetization story, or a financing-efficient way to hold a core position through two catalysts? The structure suggests the latter as much as the former.

2. 🐋 MDB — $16.5M Diagonal Wrapped Around a Confirmed May 28 Earnings Print

Decode the earnings-timed MongoDB diagonal and what each leg is doing →

• What's happening: A paired, same-second print — bought $7.6M of May 29 $295 calls (deep ITM) and sold $8.9M of Sep 18 $330 calls, for ≈$1.3M net credit. This is a diagonal, not a simple bullish roll.
• The catalyst: MongoDB Q1 FY2027 earnings is confirmed for Thursday, May 28 after the close. The long $295 leg expires the very next morning (May 29) — it is built to capture the post-earnings gap. The short Sep 18 $330 leg sells longer-dated premium to partly fund it (and carries its own risk into a second earnings event).
• The question: MongoDB has a history of large post-earnings moves in both directions. This trader is paying for that move while financing part of the cost — a sophisticated structure with real two-sided risk.

3. 🐋 GOOG — Someone Collected ≈$13M Selling Puts, Not Buying Calls

Understand the $13M short-put credit and why it is an income trade →

• What's happening: A whale sold 10,000 Aug 21 $380 puts and collected ≈$13M in premium with the stock near $395. The most they make is that credit; in exchange they take on the obligation to buy Alphabet at $380 if it drops there (≈3.8% below spot). This is willing-to-own / premium collection — bullish-to-neutral, not a "bet."
• The catalyst: Google I/O is May 19-20 (the day after the trade); Q2 earnings land ≈July 22-23, inside the expiry. The DOJ antitrust remedy is a genuine, live overhang and is treated as a real risk in the analysis.
• The question: Is selling downside near all-time highs into a live antitrust process smart income or picking up premium in front of a known risk? Both readings are fair.

4. 🐋 NVDA — $6.4M Call Ladder That Expires Two Days After Earnings

See the short-dated NVIDIA earnings ladder and the implied-move math →

• What's happening: Two same-second buys — $4.3M of June 1 $245 calls and $2.1M of June 1 $260 calls, stock ≈$223. Both out of the money, ≈2 weeks to expiry. Open interest was not shown on the tape (we did not invent it).
• The catalyst: NVIDIA Q1 FY2027 earnings is confirmed for Wednesday, May 20 after the close — these calls expire June 1, just two trading days later. This is about as close to a pure earnings/implied-move bet as the tape produces.
• The question: The options market is already pricing a large move. For this ladder to pay, NVIDIA has to move more than what is priced in — and quickly. This is a high-risk, defined-loss lottery-style structure; the analysis stresses sizing accordingly.

5. 🐋 CRWV — $4.3M of January 2027 Calls, Far Out of the Money

See the long-horizon CoreWeave AI-cloud thesis behind the $4.3M LEAP →

• What's happening: $4.3M into Jan 15, 2027 $170 calls with the stock ≈$100 (≈70% OTM). Volume well above open interest — a fresh opening position, not a close. This is a multi-quarter conviction bet on AI/GPU cloud demand.
• The catalyst: Q2 2026 earnings ≈August 18 is the next hard test before this expiry; the bull case (≈$99B backlog) and bear case (thin margins, debt-funded capex, a recent downgrade) are both laid out.
• The question: A deep-OTM LEAP is a small premium for large notional — but it needs a big, sustained move to pay. Most of these expire worthless; position size is the whole game.

6. 🐋 ARKK — $2.9M Betting on an Innovation Rebound Into 2027

See why a whale reached for January 2027 ARKK calls →

• What's happening: $2.9M into Jan 15, 2027 $90 calls on the ARK Innovation ETF, with the fund ≈$74 (≈22% OTM). A long-duration bet on disruptive-growth names (Tesla, Coinbase and the rest of the basket) and an easier-rate backdrop.
• The catalyst: The June 16-17 FOMC and the July-August earnings cluster across ARKK's top holdings are the drivers; the analysis is honest about ARKK's poor 5-year track record versus its strong trailing year.
• The question: This works only if several holdings re-rate together and rates cooperate — a basket bet that needs multiple things to go right in sequence.

7. 🐻 GLXY — A Small $1.2M Put Says This Run Cools Off

See the contrarian short-dated bet against Galaxy Digital →

• What's happening: $1.2M of May 29 $28 puts, stock ≈$27.58 — near the money, ≈11 days out, volume vastly above open interest (clearly a fresh open). The lone outright bearish print today.
• The catalyst: There is no GLXY-specific earnings before May 29, so this largely rides Bitcoin's path and broad crypto sentiment after a 60%+ trailing run; the Helios data-center delivery is an end-of-Q2 item, after this expiry.
• The question: Is this a hedge on crypto gains or a genuine fade call? Either way, it is short-dated and defined-risk (the premium), and the analysis frames it that way.

Read It By Your Style

These are not recommendations — they are how different, disciplined approaches might think about today's flow. Risk control comes first in every one.

• 🎲 YOLO / event trader: The two short-dated earnings names — NVDA (May 20 print, June 1 expiry) and the MDB long leg (May 28 print, May 29 expiry) — are where the action is. Both are binary and can go to zero in a day. If you engage, treat the premium as fully at risk and size to 1% or less of the portfolio. The market already prices a big move; you only win if the move is bigger.
• 📈 Swing trader: MSFT's deep-ITM structure is the most "swing-able" — it behaves like stock with a floor and spans Build + earnings. Consider waiting for a post-Build entry rather than chasing; a defined-risk bull call spread is a cheaper way to express the same direction.
• 💵 Premium collector: GOOG's short $380 put is your trade to study — collecting ≈$13M for the obligation to own Alphabet ≈4% lower. A cash-secured put or a defined-risk bull put spread expresses the same idea without naked tail risk. Respect the live antitrust headline; do not sell premium into an event you have not priced.
• 🌱 Just getting started: Watch, don't trade, this week's binaries. The cleanest lesson today is GOOG: selling a put collects money and is bullish-to-neutral, while buying a put (GLXY) pays money and is bearish. Knowing whether premium was paid or collected tells you more than the dollar headline. Paper-trade the NVDA and MDB earnings outcomes to see how fast short-dated options decay.

Upcoming Catalysts — Event vs. the Option That Plays It

Catalysts and the specific expiration positioned for them are kept separate on purpose, so the timing is unambiguous:

The two confirmed dates this week (NVDA May 20, MDB May 28) are the ones with an immediate clock. Everything else has weeks of runway, which changes how — and whether — you would act on it.

The Bottom Line

Today's tape is a study in timing more than direction: two confirmed earnings prints this week with short-dated bets stacked in front of them, one large structure (MSFT) built for capital efficiency rather than maximum aggression, and one quietly significant income trade (GOOG) that the dollar headline alone would misread.

The most useful discipline this week is patience around the binaries. NVDA and the MDB long leg resolve within days; you do not need to pre-position to learn from them — watching how those options behave through the print is itself worth more than a rushed entry. Large prints are information, not instructions. Size small, prefer defined risk, and let the catalyst — not the headline — do the work.

Full per-ticker analyses are linked in each section above. Options involve substantial risk and are not suitable for every investor. This newsletter is informational and not investment advice. Past performance of any strategy does not predict future results.

Ainvest Option Flow Digest — May 18, 2026