주권 AI: 한국, 유럽, 걸프만은 어떻게 실리콘밸리가 필요 없는 AI 블록을 조용히 구축하고 있는가?

Sovereign AI: How Korea, Europe, and the Gulf Are Quietly Building an AI Bloc That Doesn’t Need Silicon Valley THE DAILY --- H.E.A.T. Torching Wall Street’s Obsolete Playbook. Delivered straight to your inbox 5 days a week. Subscribe Home Posts Sovereign AI: How Korea, Europe, and the Gulf Are Quietly Building an AI Bloc That Doesn’t Need Silicon Valley Sovereign AI: How Korea, Europe, and the Gulf Are Quietly Building an AI Bloc That Doesn’t Need Silicon Valley Nov 28, 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 $4 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. ❝ Today is Ex Dividend Date: BITK: .12/Share MSTK: .28/Share ❝ I’m hosting a webinar entitled “Why Covered Call ETFs Suck and What to Do Instead” (More Info Below) December 9 2-3pm. Sign Up Here Table of Contents --- H.E.A.T.--Sovereign AI --- News vs. Noise: What’s Moving Markets Today --- A Stock I’m Watching Why Covered Call ETFs Suck-And What To Do Instead --- H.E.A.T. AI is obviously the top theme we focus on, but it’s not just a US and China thing anymore, and I suspect a lot of the alpha going forward will come from under the radar names in other areas of the world….. AI is no longer a U.S.-and-China-only game – and the rest of the world is done renting their future from Silicon Valley. 1️⃣ Sovereign AI – Korea (and friends) don’t want to be tenants South Korea is taking “sovereign AI” literally: Tripling its government AI budget to ~$6.8B next year Launching a $100B+ National Growth Fund for strategic tech (AI, chips, data centers, etc.) Ordering ~260,000 Nvidia GPUs as the raw compute bedrock Backing a national LLM + local AI stack through Naver, Kakao, Samsung, SK, and AI chip startups like Rebellions and FuriosaAI This is the core sovereign‑AI playbook: Own the compute (data centers + accelerators) Own the models (local LLMs in your language and cultural context) Own the data and rules (domestic regulation, privacy, security) You’re seeing versions of the same script elsewhere: France/Germany : Mistral + SAP pushing a European “sovereign AI” platform U.K. : A sovereign AI unit, trying to attract top AI firms and keep data onshore India : Building domestic models + compute capacity Saudi/UAE : Treating sovereign AI spending like defense – billions of dollars of chips, data centers, and national models Important nuance: most of these sovereign stacks still run on Nvidia and U.S. software. Sovereign AI ≠ “no U.S. tech.” It means no single foreign gatekeeper . Countries want: Local control over data & rules Alternative suppliers if U.S./China geopolitics go sideways A say in where value accrues (not all to U.S. hyperscalers) From a trade lens: this is more global AI capex, not less – but the revenue gets spread across local infra, local models, and local chip startups instead of only the Magnificent Seven. 2️⃣ Europe’s “slow and constrained” AI could actually be the safer bet Europe reads like the opposite vibe… and that’s exactly why it’s interesting. Europe looks “behind” on headline AI: Fragmented markets, heavy regulation, fewer foundational model champions Real chokepoints in power , permitting , and grid congestion Orders of magnitude fewer data centers than the U.S. But that scarcity is starting to look like a feature, not a bug : Power & grid constraints are forcing smarter site selection (Nordics, Spain, Italy) and more modern designs Regulators are pushing for energy/water reporting, local economic impact, sustainability Developers increasingly secure tenants before they build , with 10–15 year contracts – less YOLO, more stable cash flows And crucially: Europe doesn’t need to win the training arms race to win on inference : Estimates put ~70% of long‑run AI demand in inference , not training Inference often must happen inside local borders (latency, sovereignty, data rules) That lines up perfectly with Europe’s strength: high‑quality, connectivity‑rich cloud/data centers rather than speculative mega‑training sites So Europe’s playbook looks like: Fewer, better‑sited, more energy‑efficient data centers Designed from day one for denser racks + better cooling to handle AI inference alongside cloud Tight regulation that, yes, slows things down—but also raises the moat and reduces the odds of a boom‑and‑bust oversupply cycle For long‑horizon capital, that’s often exactly what you want: constrained supply, increasing strategic demand, and assets that are hard to replicate. --- Winners and Losers: Global AI Edition Likely Winners 1. Nvidia (yes, still) Korea’s sovereign‑AI push includes hundreds of thousands of Nvidia GPUs . Gulf states just got clearance to buy tens of thousands of high‑end AI chips. Sovereign AI is, ironically, a massive incremental customer for Nvidia, even if the marketing is “independence from U.S. tech.” 2. Korean Tech Complex (Samsung, SK Hynix + local AI chip startups) Korea is one of the few places with world‑class memory, logic, packaging, and system‑integration under one flag. SK Hynix is already central in HBM for AI; Samsung is a full‑stack semiconductor + device giant. Startups like Rebellions and FuriosaAI are levered calls on national AI policy – if any country can turn “we will build our own chips” into something real, it’s Korea. 3. European data‑center & infra plays Energy‑advantaged regions (Nordics, Spain, parts of Italy) and grid‑friendly locations are structural winners. Operators who can deliver high‑density, fiber‑rich, AI‑ready facilities and navigate European permitting will sit on deep moats. Because builds are slower and more customer‑anchored, these assets could offer cleaner, less bubbled exposure to AI demand than some U.S. hyperscaler capex proxies. 4. Sovereign AI platforms and stack providers (Europe & Asia) Names like Mistral (FR), SAP (DE), Naver/Kakao (KR) and similar regional champions benefit from: Local language + regulatory alignment Government procurement preferences Enterprises who must keep data in‑region They don’t need to beat OpenAI/Google everywhere; they just need to be the default in their jurisdiction . 5. Power, grid, and renewable developers in the “right” places AI is a power story as much as a chips story. Utilities and grid‑owners in regions with cheap, reliable electricity and friendlier connection timelines stand to win from sovereign AI + inference build‑out. Europe’s repurposed industrial sites (old plants → new data hubs) can become high‑barrier, long‑duration cash machines . Likely Losers (or “Less Obvious Winners”) 1. Countries with hard power constraints and policy whiplash Germany, U.K., Ireland, Netherlands etc. face grid bottlenecks, de facto moratoria, and high energy costs . They won’t be shut out forever, but they risk losing the first wave of sovereign‑AI data‑center siting to better‑positioned neighbors. 2. “Neo‑cloud” and speculative infra with weak tenants Data centers built on 5–7 year contracts with unproven AI/cloud startups face obsolescence and tenant risk . If AI hype cools or the customer fails, you’re left with stranded assets in the wrong place . 3. Local “full‑stack” vanity projects without scale Not every sovereign‑AI ambition will work. Countries that try to do everything in‑house (chips, models, cloud, infra) without the talent/scale to support it risk burning capital for second‑tier tech that still relies on U.S. components. 4. U.S. hyperscalers’ monopoly narrative They still win a ton of business (especially in training), but the idea that “all global AI flows through a handful of U.S. clouds” is fading. Sovereign AI + European inference + Middle East build‑outs mean more regional platforms, regional models, and regional infra owners . 5. Investors treating “AI” as only a U.S. mega‑cap trade The alpha may increasingly be in non‑U.S. infra, regional semis, and local AI platforms , not just buying more NVDA/MSFT/GOOGL at any price. --- News vs. Noise: What’s Moving Markets Today In today’s News vs. Noise , the big story isn’t just that markets are bouncing — it’s why . A couple of weeks ago, positioning was wildly stretched, everyone was max long. You then got exactly what you’d expect: a “healthy” flush, with systematic sellers and discretionary de‑risking knocking the S&P lower and blowing out intraday volatility. As of last week, that positioning had been taken down to basically flat. Once the weak hands and forced sellers were done, the tape was ready to go higher again. The move off the lows is less about some new macro revelation and more about cleaner positioning and a market that can finally go up without tripping every stop‑loss in the system. Layered on top of that, the Fed narrative quietly flipped back in the bulls’ favor. The market is now pricing close to 20 bps of cuts in December – meaning investors see a decent chance of a 25 bp move – versus very low odds just a couple of weeks ago. The exact timing might be noise: if they don’t cut in December, they probably cut in January. The important part is the direction of travel : policy is in an easing regime, global liquidity is still abundant, and whoever Trump picks as the next Fed chair is likely to be more dovish than Powell at the margin. Put differently, you’ve gone from stretched longs + hawkish fear to cleaner books + an easing Fed + rising global M2 , all inside the strongest seasonal window of the year. That’s how you get a 4%+ S&P snapback that sticks . Takeaways Positioning, not “new fundamentals,” drove the washout and the rebound. Crowded longs got taken to flat; once that supply was cleared, the tape had room to heal. The market has re‑embraced the December cut story. Roughly 20 bps of easing are now priced for that meeting — but the real story is that cuts are coming over the next few months, even if the exact month is a coin toss. For risk assets, direction matters more than date‑stamping. December cut + January pause or December pause + January cut both lead to the same place: lower policy rates vs. today. Global liquidity is a quiet tailwind. With most major economies running fiscally expansionary policies, GMI’s composite global M2 is set to rise into 2026 — a supportive backdrop for equities and credit. Positioning lens: Treat this rally as the start of a post‑flush regime, not the end of it — but size with the understanding that the path will still be noisy as the market digests each new datapoint and Fed soundbite. --- A Stock I’m Watching Today stock is Nebius Group (NBIS)…. Locked‑in demand from tier‑1 hyperscalers MSFT and META contracts alone already overwhelm current energizable capacity ; Q3 pipeline reportedly ~$4B, +70% QoQ. These are capacity leases , not pure “we hope they come” usage; that’s real visibility out into 2027–28. Macro tailwind that matches your core thesis All the big capex math says AI racks use 10–20× the power of legacy racks, and cloud capacity is still structurally short – you’re already expressing that via VRT, VST, CEG, PWR, ETN, etc. Nebius is a direct way to own the “outsourced build‑out” piece and sits on the same secular curve as CRWV. Valuation is rich, but not insane relative to growth Market cap ≈ $23–24B , TTM revenue ~$363M; Street is modeling >300%+ growth into 2026 and 60–70%+ upside to consensus PT (~$157). If they execute anywhere near the MSFT + META ramps, you’re probably looking at low‑teens forward P/S on 2026 revenue , which is not crazy for hyperscale AI infra growing triple‑digits. From an asymmetry standpoint: if they stumble badly, you can easily lose 50–70%; if the AI build‑out stays on track and they keep signing and energizing campuses, a 2–3× in 2–3 years is absolutely on the table. How Else I Can Help You Beat Wall Street at Its Own Game Inside --- H.E.A.T. is our monthly webinar series, sign up for this month’s webinar below…. Why Covered Call ETFs Suck-And What To Do Instead Tuesday December 9, 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 Sign Up Here All rights reserved. Keep Reading No posts found View more caret-right 0