① What Happened Today
Semiconductors got sold off. AI compute demand is strong, but delivery and financing risks are rising.
Semis under pressure: Korea's KOSPI fell 3.73%, Samsung dropped 3%, and SK Hynix fell 4%. US semis were down about 2% premarket. Neocloud names NBIS and CRWV fell 5% and 4%. Weak Asian tech dragged on market mood.
NVDA: Chinese companies reportedly ordered over 3 million export-compliant RTX PRO 5500 GPUs, worth over $35 billion at list price. But Nvidia reportedly may only be able to supply 5%–10% of the orders. Strong demand doesn't mean revenue shows up right away — supply is the key.
AI compute expansion: Anthropic is in talks with Nscale for a new compute deal worth about $45 billion. OpenAI is speeding up a $30 billion funding round. AI demand is still hot, but money spent, equipment supply, and project delivery will decide whether companies actually make money.
META: Ads and Muse AI offer growth, but the market is more focused on whether 2027 costs and capex will be much higher than expected. The stock has already run up, so even an earnings beat may not erase spending worries.
Today's earnings: SPCX (SpaceX) and NFLX (Netflix). Also, Cloudflare is holding its Connect conference — watch for NET news.
② What Money Is Betting On / Protecting Against
[Buyer] DRAM | 2027/01/15 60 Call$DRAM 20270115 60.0 CALL$
About 87k contracts opened total | representative trade amount about $8.47M | stock around $48.16–48.45 when traded
Money bought a lot of long-dated calls, betting memory chips still have upside over the next year or so.
Betting on: AI server demand keeps rising, the memory cycle continues, and the stock can challenge $60+.
**Key level:** $60 is the strike, about 24% above the stock price at the time. If the stock doesn't break the strike before expiration, buyers can still lose all premium. Long-dated contracts also need time for the move to play out.
[Buyer] SOXX | 2026/10/30 360 Put$SOXX 20261030 360.0 PUT$
About 70k contracts opened total | representative trade amount about $1.11M | stock at $505.76 when traded
Money bought deep out-of-the-money puts, possibly to guard against an extreme short-term drop in semis.
Guarding against: Semis have been hit by the Korea selloff, weak memory, and neocloud pullbacks. Some money is using cheap puts to hedge tail risk.
Key level: $360 is far below the $505.76 stock price at the time, so this isn't a normal pullback bet. If the stock doesn't drop hard, the option can lose value fast as expiration nears.
[Buyer] EWY | 2026/12/18 185 Call$EWY 20261218 185.0 CALL$
24.7k contracts | about $9.01M traded | stock at $158 when traded
Money is betting Korea's market bounces by year-end, hoping for a bigger recovery after tech weakness.
Betting on: Korea is under short-term pressure, but memory and semiconductor demand may still support a bounce.
Key level: $185 is about 17% above the stock price at the time. If Korean stocks keep falling, this far-OTM call can also lose time value fast.
[Seller] NFLX | 2026/11/13 58 Put$NFLX 20261113 58.0 PUT$
About 20.8k contracts | about $728.7K traded | stock around $69.65–69.68 when traded
Selling puts before earnings shows some money thinks the chance of NFLX falling below $58 is limited, or is willing to buy the stock at that price.
**Betting on:** Even with weaker US user engagement, Netflix stock may not crash.
**Key level:** $58 is the reference price for assignment, but an earnings gap can quickly raise risk. Collecting premium doesn't mean there's no loss risk.
[Seller] ORCL | 2026/11/20 100 Put$ORCL 20261120 100.0 PUT$
About 11.3k contracts | about $748.8K traded | stock at $130.60 when traded
Selling deep OTM puts shows some money thinks Oracle is unlikely to break below $100 short term.
**Betting on:** AI cloud demand still supports Oracle's business, and the stock won't drop hard soon.
**Key level:** $100 is about 23% below the stock price at the time, but the company's huge capex and funding needs are still potential risks. If the stock drops fast, sellers can lose much more than the premium.
[Combo] NVDA | 2027/01/15 185 Put + 235 Call
$NVDA 20270115 185.0 PUT$ $NVDA 20270115 235.0 CALL$
11.1k contracts each | Put about $4.5085M; Call about $10.2414M | stock at $217.90 when traded
Sell 185 Put, buy 235 Call. This is a bullish risk reversal — taking on downside assignment risk in exchange for upside participation.
Betting on: AI compute demand and orders keep growing, and Nvidia can move toward $235 or higher.
**Key levels:** $235 is the call strike; $185 is the sold put strike. If the stock breaks below $185, the strategy can face a bigger loss. This is a bullish setup with downside obligation, not risk-free arbitrage.
③ What to Watch for Related Tickers
NVDA | Orders are strong, but supply and delivery are the test
Huge Chinese orders plus the Anthropic/Nscale talks show AI infrastructure demand is still strong. But supply limits and the financing, permits, and delivery timeline for big compute projects may keep orders from turning into revenue right away. The long-dated 235 Call shows some money betting on medium- to long-term upside. Short term, watch whether semis stop falling and whether the market keeps chasing.
DRAM / SOXX | Long-term bullish memory, but short-term crash hedging
DRAM's long-dated 60 Call and SOXX's near-term 360 Put both saw big opening trades, showing risk views on different time frames: one bets on the long memory cycle, the other guards against a short-term semi crash. Korea's selloff and weak Samsung/SK Hynix could keep weighing on sentiment. Watch whether DRAM holds around $48 (the trade-time price) and whether SOXX keeps falling. These big orders alone don't prove where the market must go.
EWY | After Korea's sharp drop, money bets on a year-end bounce
KOSPI fell 3.73%, Samsung and SK Hynix weakened, and short-term mood clearly worsened. But EWY's 185 Call shows some money still bets on a year-end bounce. $158 was the stock price at the time; $185 is the strike. If Korean stocks keep falling, the bounce bet faces more pressure. If semis stop falling and recover, the trade has a better chance.
NFLX | Earnings key: user stickiness and ad monetization
Goldman keeps a Buy rating but cut its target from $94 to $90. US Q3 monthly active users fell 10% year over year, and watch time fell 20%, showing Netflix faces competition from YouTube and others in the US; other regions look steadier. Ad revenue is expected to double year over year in 2026, and live/event content could help monetization. Beyond revenue and profit, watch how management explains US engagement, content strategy, and ad growth. Options: this week's expiry IV is about 120%, so the market expects a big move. $65 is an important support reference; $70 is key resistance. The money selling the 58 Put bets a deep drop is unlikely, but that doesn't mean earnings risk is gone.
SPCX | Earnings aren't the only focus — Starship is the key catalyst
Morgan Stanley has an Overweight rating and $300 target. The story isn't just Q3 numbers; it's whether SpaceX can turn Starlink, AI compute, and Starship tech breakthroughs into bigger long-term business value. The report expects Q3 revenue around $13B and EBIT around $2.36B. But whether Starship Flight 15 in early November attempts a ship catch could be a more important catalyst. Compute expansion, Neocloud deals, and funding needs also matter. Options: IV is about 62.3%, and this week's expiry IV is as high as 124.5%, showing the market has priced in big swings. $140 is an important support reference this week; $150 is resistance. Big orders show sold puts and bought calls leaning bullish, but earnings and Starship news can both cause sharp moves.
ORCL | AI orders have promise, but capex is key for the stock
Besides Morgan Stanley, JPMorgan expects Oracle may raise its FY30 revenue and operating cash flow targets. But the company may still need about $20B more capital in FY27, with total capex possibly reaching $90–95B. The market really cares about funding sources and when capex peaks. Sellers of the 100 Put bet a short-term crash is unlikely, but leverage and funding risk can't be ignored. Watch whether the company gives a clearer capex and funding path.
④ What to Watch Pre-Market Today
Semis / Memory: After Korea tech's big drop, can US semis stop falling? DRAM's long-dated Call and SOXX's short-term deep-OTM Put can coexist — bullish on the industry long term, hedged against a short-term crash.
NVDA: Chinese orders and new compute deals are positive, but supply limits, construction progress, and whether financing gets done will decide if the good news turns into real revenue.
NFLX: Earnings aren't just about whether revenue beats. Watch US user activity, watch time, ad revenue, and management's comments on content strategy. High IV means big price moves are possible around earnings.
SPCX: Watch earnings and comments on compute deployment, but don't ignore the early-November Starship test flight as a potential catalyst.
META / ORCL: For META, watch whether AI spending turns into ad revenue. For ORCL, watch capex and funding pressure. Strong AI demand doesn't mean every related stock rises together.
NET: Cloudflare's Connect conference may bring product and business updates. Watch whether the news becomes a new stock catalyst.
One-line read: AI demand is still strong, but the market is starting to separate "how big are the orders" from "when will they make money." Short term, first watch whether semis stop falling. On earnings, watch NFLX's user stickiness and SPCX's compute and Starship progress.
This is an educational summary of market info and options trading behavior, not investment advice. Big-order direction and intent are guesses; trade data doesn't necessarily represent final positions or any single institution's full strategy.
Comments