10조 달러 내기: OpenAI는 이제 실패하기에는 너무 큰가?

The $10 Trillion Bet: Is OpenAI Now Too Big to Fail? THE DAILY --- H.E.A.T. Torching Wall Street’s Obsolete Playbook. Delivered straight to your inbox 5 days a week. Subscribe Home Posts The $10 Trillion Bet: Is OpenAI Now Too Big to Fail? The $10 Trillion Bet: Is OpenAI Now Too Big to Fail? Oct 23, 2025 I’ve been a trader and investor for 44 years. I left Wall Street long ago—-once I understood that their obsolete advice is designed to profit them, not you. Today, my firm manages around $5 billion in ETFs, and I don’t answer to anybody. I tell the truth because trying to fool investors doesn’t help them, or me. In Daily --- H.E.A.T. , I show you how to Hedge against disaster, find your Edge, exploit Asymmetric opportunities, and ride major Themes before Wall Street catches on. Table of Contents Announcements --- H.E.A.T. What’s Moving Markets Today How Else Can I Help You Announcements MSTK Launches Today!!!! I’m doing a webinar today about why covered calls – one of the most-touted “safe” ways to generate income – are in reality portfolio poison… and I’ll tell you what to do instead. You can sign up here Speaking of covered calls… I had a great conversation with Max Convexity yesterday about all things options…… And, I just published an article on the topic for Seeking Alpha (may be behind a pay wall, sorry if it is) Why Covered Calls Are A Flawed Strategy Discover why covered call ETFs may not be ideal for income and explore superior strategies like writing puts or 0DTE calls. Click here to read my analysis. seekingalpha.com/article/4831933-why-covered-calls-are-a-flawed-strategy --- H.E.A.T. Back to the AI theme today. What Open AI is doing could lead to the next wave up upside in the AI ecosystem, or a 2000 style meltdown……. Sam Altman just executed the most aggressive capacity land-grab in tech history—locking in headline deals across Nvidia, Oracle, AMD, Broadcom (and circling back to Microsoft and SoftBank)—and in doing so tethered huge swaths of market cap to one company’s execution. OpenAI is targeting 250 GW of compute by 2033 (a mid-sized country’s load) with only ~$13B of revenue today, while counterparties trumpet multi-hundred-billion pipelines. That mismatch is the trade. What’s really happening Altman manufactured competitive FOMO. Oracle inks a $300B contract and re-rates. Nvidia follows with a lease of up to ~5M GPUs (≈$350B by today’s math) plus an option to invest $100B and even guarantee portions of OpenAI’s future debt. AMD offers tens of billions of GW capacity and a large equity kicker. Broadcom matches with a 10-GW custom system and accepts margin dilution risk. Microsoft, once exclusive, allows OpenAI to multi-home while racing new capacity of its own. The result: circular financing, vendor pre-buys, and “too-intertwined-to-fail” incentives that push everyone to keep the flywheel spinning. Why it could work If model utility compounds and inference demand keeps outpacing efficiency gains, these deals become self-funding: more compute → better models/products → more revenue → more compute. OpenAI has already demonstrated hit-making beyond ChatGPT (e.g., Sora’s viral lift), and each viral product justifies another tranche of capacity. Why it could break Execution and funding risk are non-trivial: ( • 1) power bottlenecks (siting, interconnects, tariffs, and grid constraints), ( • 2) circular vendor financing that pulls demand forward, ( • 3) custom silicon risk (time-to-yield, software stack maturity, field reliability), ( • 4) regulatory and governance shocks, and ( • 5) the S-curve: if monetization lags the compute curve, capex turns from asset to anchor. Winners (base case: demand grows, power is scarce, and capacity is king) • Semiconductor & memory scale players with proven ramps: NVDA, AMD, MU, TSM. • Custom silicon and systems integrators that can actually deliver: AVGO (with margin risk managed), NVDA HGX/DGX platform. • Cloud and colocation that capture multi-tenant AI load: ORCL (AI re-rate), MSFT (durable hyperscale), EQIX, DLR. • Power stack and grid enablers: ETN, PWR, SIEGY/ABBNY, CEG, VST; thermal/cooling: CARR, JCI. • HBM, advanced packaging, substrates, optics vendors up the chain (own via the leaders above if you want liquidity). Losers (or at least higher-beta risk) • Second-tier AI chip hopefuls without software moats; custom ASIC programs that slip. • Vendors taking big equity/guarantee exposure if product-market fit or funding cycles stall (AVGO, AMD have acknowledged margin trade-offs; monitor contract terms). • “Commodity cloud” capacity with slower capex cycles; late movers in grid interconnects. • Power-intensive users with no PPA hedges—miners and thin-margin compute renters—if electricity reprices higher. Three plausible scenarios to underwrite Exponential holds, monetization catches up: Suppliers, power names, and AI infra REITs compound; custom silicon becomes accretive by 2027–28; software margins improve as model utility rises. S-curve stalls, funding tightens: Capex deferrals, write-downs on bespoke systems, equity kickers get painful; infra winners with diversified books (ETN/PWR/CEG/VST/EQIX) outperform vendor-financed chip stories. Power/regulatory choke: Delays extend timelines; utilities and transmission OEMs win while chip names become “deliver-later.” Pairs (long power stack vs. vendor-financed silicon) work. How to invest now (actionable, not theoretical) • Core picks-and-shovels: overweight the power stack (ETN, PWR, CEG, VST) and neutral-to-positive-carry colo (EQIX, DLR). These win in scenarios 1–3. • Selective silicon barbell: own NVDA core; add AMD for share-gain optionality; trade AVGO with downside protection (acknowledge margin dilution/custom risk). Keep MU for HBM leverage. • Exposure control: prefer diversified infra over single-customer custom chip stories; if long AVGO/AMD on OpenAI headlines, pair with put spreads into major milestones. • Power asymmetry: keep our AI-power utilities and transmission names overweight; PPAs and interconnect rights are the new moats. Key risks to monitor weekly Power and interconnect lead times; HBM/packaging supply; contract disclosures around equity/guarantees; regulator posture on vendor financing; product monetization metrics (ARPU, attach, usage retention) vs. compute growth; sovereign and utility PPAs. Bottom line Altman turned capacity into the narrative and made OpenAI systemically important to the AI supply chain. If the revenue curve meets the compute curve, today’s “too big to fail” web becomes self-funding; if it doesn’t, the pain shows up first in custom silicon margins and vendor balance sheets—while the grid and colocation still get paid. Position where cash flows clear in all paths: power, transmission, and scaled silicon—with protection on the most levered headlines. What’s Moving Markets Today Two key things I’m watching are the debasement trade and the momentum trade. Both had some problems yesterday. Gold did stabilize, but crypto sold off. Momentum stocks got crushed. I’d said before, this is the kind of action you will see at the top, but it’s also normal for any uptrend to sell off from time to time. So where are we? Impossible to predict in real time, I would be cautiously buying the dip in areas you like, make sure you have hedges in place, and keep your head on a swivel. We have had a few momentum unwinds this year, and they all have resolved higher. Doesn’t mean this one will, but for now that’s how we are playing it until proven otherwise. --- A stock I’m watching today is Ge Vernova (GEV)….. It’s integral to the AI infrastructure trade and it had an undercut and rally at the September lows. CCCX is another stock of interest that I bought the dip on yesterday. It’s a SPAC still, so you have a $10ish floor, and it’s going to become a quantum stock. Rumors around that Trump’s next investment is going to be quantum…… How Else I Can Help You Beat Wall Street at it’s Own Game Why Covered Call ETFs Suck-And What To Do Instead Thursday October 23rd 2-3PM EST Covered call ETFs are everywhere — and everyone thinks they’ve found a “safe” way to collect yield in a sideways market. The truth? Most of them suck. They cap your upside, mislead investors with “yield” that’s really your own money coming back , and often trail just owning the stock by a mile. Join me for a brutally honest breakdown of how these funds actually work — and what you should be doing instead. What You’ll Learn: 🔥 Why “high yield” covered call ETFs are often just returning your own capital 📉 How most call-writing strategies quietly destroy compounding 🚫 Why owning covered calls in bull markets is like running a marathon in a weighted vest 💡 The simple structure that can fix these problems — and where the real daily income opportunities are hiding Click Here To Register All rights reserved. Keep Reading No posts found View more caret-right 0