앤스로픽의 배제, 프론티어 AI를 상품화시키다

THE DAILY H.E.A.T. Torching Wall Street’s Obsolete Playbook. Delivered straight to your inbox 5 days a week. Subscribe Home Posts Anthropic's Banishment Just Commoditized Frontier AI Anthropic's Banishment Just Commoditized Frontier AI Six vendors got in. One didn't. The real trade isn't where Wall Street is looking. May 4, 2026 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. Table of Contents H.E.A.T. News vs. Noise: What’s Moving Markets Today ETF News In Case You Missed It H.E.A.T. The Pentagon signed six AI vendors this week. Wall Street is celebrating the wrong winners — the real money is in the layer above the models, not the models themselves. THE SETUP On Friday morning, the Department of Defense announced completed agreements with six AI vendors for classified-environment work: OpenAI, Google, Microsoft, Nvidia, SpaceX (which now houses xAI), and a pre-revenue startup called Reflection AI. Amazon is reportedly in talks to complete a similar deal. The move follows the Pentagon's January decision to exclude Anthropic from the initial vendor list — until recently the only frontier-model provider operating inside Maven's classified stack via Palantir — designating the company a supply-chain risk unsuitable for military work. Defense Secretary Pete Hegseth, testifying before House Armed Services on Thursday, publicly attacked Anthropic's leadership as politically incompatible with DoD work. In Washington terms, that translates to a single word: unreliable. And reliability is the only currency that matters in classified procurement. The market read this as a story about Anthropic getting punished. That reading is too narrow. The actual signal — the one with capital implications — is that the Pentagon just turned frontier AI into a multi-vendor commodity input, at the precise moment Wall Street was pricing the leading labs as if their moats were structural. An important clarification, because sophisticated readers will catch us if we skip it: Friday's news is the vendor gate, not a revenue award. Task orders and real dollars come later. But the gate is where the decade gets decided — every state CIO, Five Eyes intelligence service, and Fortune 100 CISO reads DoD vendor lists the way mortgage desks read the 10-year. The procurement signal propagates. "We are equipping the warfighter with a suite of AI tools to maintain an unfair advantage and achieve absolute decision superiority." — Emil Michael, Undersecretary of Defense for Research & Engineering BY THE NUMBERS 6 Frontier AI vendors now cleared for classified DoD work DoD, May 1, 2026 $25B Reported valuation Reflection AI is raising at — pre-product WSJ reporting $200B+ FY26 DoD IT + AI procurement envelope GAO / DoD budget docs THE CONSENSUS IS WRONG The consensus reading of Friday's announcement is that the model labs won, that Anthropic lost permanently, and that this is a story about which AI you want in the SCIF. Each of these misreads what the Pentagon actually did. Important to be precise here: the Pentagon is not saying all models are equal. It is saying it refuses to be dependent on any one of them. That distinction matters. "All models are equal" is a capability claim — and it would be wrong. "We refuse to be hostage" is a procurement claim — and it collapses the monopoly premium without requiring the underlying capability story to be flat. One sentence shift, completely different investment implication. When a buyer of this size signals it intends to multi-source a category, that category's pricing power compresses — not next quarter, but over the contract cycle. State-level CIOs, allied intelligence services, and large-enterprise CISOs read these contracts the same way mortgage desks read the 10-year. They will multi-source too. The corollary is that economic value migrates to the layers the Pentagon didn't commoditize. There are three of them: the integration layer that makes models usable inside classified workflows (Palantir, Booz Allen, the prime contractors that own the accreditations), the compute layer that all six vendors equally depend on (Nvidia silicon, hyperscaler capacity), and — increasingly — the open-source weight providers whose business model never depended on closed-API margins in the first place. Nvidia is the cleanest expression of this. Jensen Huang has been telling anyone who will listen that open models win in national security contexts because every weight is auditable. Friday's deal validates the pitch directly: Nemotron — Nvidia's open-source frontier model line — is in. Reflection AI, in which Nvidia is an investor, is in, reportedly at a $25 billion valuation before it has shipped a single model. Nvidia just got two seats at a six-seat table while continuing to sell silicon to all six. That is structurally different from being one of six closed-API vendors competing on benchmark scores. Anthropic's exclusion is real, but it is also legally contested. The company is litigating in two separate cases, its models were used in the Iran operation and the operation to capture former Venezuelan president Nicolás Maduro, and Palantir's Maven platform still runs on Claude infrastructure for a meaningful share of inference. The bear case on Anthropic's commercial trajectory is overstated. The bull case on the other five is more overstated. SPOTLIGHT WHY PALANTIR IS THE QUIET WINNER OF ANTHROPIC'S EXCLUSION The Maven platform — Palantir's classified AI workflow layer — is now the connective tissue between six approved frontier vendors and the warfighter. Before Friday, Maven was effectively a single-vendor pass-through for Claude. After Friday, it is a multi-vendor orchestration layer with switching costs that the Pentagon itself has now sanctioned. Multi-vendor orchestration is a higher-margin, stickier business than single-vendor reseller. The DoD did not commoditize the integrator. It commoditized the integrator's suppliers. That is the entire investment case in one sentence. Booz Allen, Leidos, SAIC, and CACI sit downstream of the same dynamic at the systems-integration level — less elegant exposure, but cheaper multiples and more dollars flowing through them per AI deployment than through any single model API. THE TRADE, IN ONE LINE Long the deployment layer plus compute, short the model-narrative premium. That is the entire issue compressed into eleven words. The basket below tiers the long side by conviction. The pressure points section names where the model-narrative premium is most exposed. WINNERS — TIERED BY CONVICTION Tier 1 is structural and survives most outcomes of the Anthropic litigation. Tier 2 requires the multi-sourcing thesis to play out as expected. Tier 3 is asymmetric — small position size, larger potential payoff. TIER TICKER / NAME THESIS TIER 1 PLTR Maven becomes a six-vendor orchestration layer instead of a single-vendor passthrough. Margin profile improves; switching costs accrue to Palantir, not to any model lab. The DoD just expanded Palantir's TAM at no incremental cost to Palantir. TIER 1 NVDA Wins twice on Friday: Nemotron contracted directly, and portfolio company Reflection contracted alongside it at a reported $25B mark. Continues to sell silicon to all six approved vendors. Open-source national-security thesis — which Huang has been seeding for two years — now has a sovereign anchor customer. TIER 1 MSFT Already inside the Pentagon at the cloud and productivity layer. Friday's deal extends the relationship to its proprietary AI tooling. The least sexy name on this list and probably the most reliable cash compounder from this dynamic. TIER 2 BAH / LDOS / SAIC Defense systems integrators capture per-deployment dollars regardless of which model wins. Anthropic litigation outcome is irrelevant to their book. Trading at substantially lower multiples than the model-adjacent names. Underowned relative to the AI-defense narrative. TIER 2 GOOGL Cloud plus Gemini now cleared for classified work alongside DeepMind alumni at Reflection. Less a thesis on Friday's deal specifically, more a rerating catalyst on the perception that Google was excluded from defense AI. TIER 3 Reflection AI (private) Reported $25B valuation with no shipped model is the speculative line item. But Nvidia anchor, DeepMind founders, sovereign Korea deal, and a Pentagon contract before product launch is a profile that priors say compounds. Watch for secondary market access or eventual IPO. Position size accordingly. PRESSURE POINTS — WHERE THE RISK IS TIMING, NOT FAILURE These are not businesses that fail. They are businesses where the multiple gets re-rated as the multi-sourcing dynamic plays out. The risk is margin compression and crowded positioning, not zeroes. TIER TICKER / NAME THESIS PP1 OpenAI (private) Now one of six in the Pentagon stack rather than the assumed default. Implications for the secondary-market valuation are real if the multi-sourcing pattern propagates to enterprise. Still the consumer winner; the defense-tier premium is what's at risk. PP2 Closed-API model exposure broadly Any name whose investment case rests on closed-frontier-model premium pricing — including private secondaries — faces a slow re-rate as buyers normalize multi-vendor procurement. Not a 2026 problem; a 2027–2028 problem. PP3 Pure-play defense primes without AI integration Old-guard primes that didn't build AI integration capability now look to be eating the integrator's lunch from a position of weakness. Lockheed, Northrop, RTX still sell platforms — but the software value is migrating to Palantir-class layers. PP4 Anthropic (private, secondary) Real near-term overhang from DoD exclusion and CEO-level public friction with the administration. But two active lawsuits, ongoing classified operations using Claude (Iran, Maduro), and a commercial book that doesn't depend on US defense make the secondary discount likely overdone. Risk is timing of resolution, not viability. THE HONEST BEAR CASE BEAR CASE We could be wrong about commoditization. Frontier models are not actually fungible at the capability frontier — GPT-class, Gemini-class, and Claude-class models still differ meaningfully on long-context reasoning, agentic reliability, and tool use. The Pentagon may discover this and consolidate spending on one or two vendors after a 12–18 month evaluation period. If that happens, the multi-sourcing thesis breaks and the consolidation winner captures most of the defense AI premium. Palantir's multiple is also already reflecting a great deal of optimism. At current levels, Palantir trades at a sales multiple that prices in years of clean execution. A miss on either commercial-side growth or government-side contract pacing produces a sharp drawdown even if the long-term Maven thesis is correct. Reflection AI at a reported $25B before shipping a model is the part of this story that, with hindsight, may look like the cycle peak. The Pentagon contract is real; the valuation is a bet on people, not product. Position size matters more than thesis quality at that pre-product valuation. Finally, the Anthropic litigation could resolve in ways that either restore them to the stack — collapsing the urgency premium other vendors are receiving — or escalate into a broader political fight over AI provider eligibility that drags the whole category into headline risk. FIVE TAKEAWAYS 1. The Pentagon didn't pick winners — it commoditized the model layer and shifted economic value upward to the integrator (Palantir) and downward to compute (Nvidia). That's the trade. 2. The cleanest one-line expression: long the deployment layer plus compute, short the model-narrative premium. Everything else in this issue is downstream of that sentence. 3. Nvidia is the only name that wins twice on Friday's announcement: Nemotron contracted directly, and portfolio company Reflection contracted at a reported $25B valuation pre-product. Open-source national-security thesis is now sovereign-validated. 4. Defense systems integrators (BAH, LDOS, SAIC, CACI) are the lower-multiple, cleaner expression of the same trade — they capture per-deployment dollars regardless of which model wins. Underowned relative to the AI-defense narrative. 5. Anthropic's secondary-market discount is likely overdone. Two active lawsuits, ongoing classified operations using Claude, and a commercial book that does not depend on US defense make the bear case noisier than the headlines suggest. Space Exposure Designed For Potential Weekly Payouts The space economy is no longer just a government program — it’s becoming part of a global industry that is gradually advancing. But why view long-term growth as the only potential path forward? SPCI seeks to offer approximately 100% concentrated exposure to the space industry while striving to generate weekly income through a disciplined put credit spread options strategy. Gain exposure to the “final frontier” and a potential weekly distribution while you wait for the theme to play out. Discover SPCI. Distributor: Foreside Fund Services | Investing involves risk including possible loss of principle. News vs. Noise: What’s Moving Markets Today THE NOISE Wall Street is in full "war on, war off" mode. Brent crude is back near $114 after roundtripping from $118 to $90 and back, the Strait of Hormuz has 600 commercial ships trapped behind Iranian mines that the Pentagon says will take six months to clear, and gasoline is sitting above $4 at the pump. The financial media wants you to obsess over every ceasefire headline and every tanker stuck in the Gulf. Meanwhile, the Nasdaq and S&P 500 just printed fresh all-time highs. The talking heads can't square it, so they keep recycling the same "geopolitical risk" segment between commercial breaks. THE SIGNAL The market has already voted: $100+ oil is not the story. The story is that financial conditions eased right back to the lows the moment Trump pulled back from the brink, the Atlanta Fed's GDPNow is tracking 3.5%, Q2 earnings growth is running at 27%, and the hyperscalers are on pace to spend over $700 billion on AI capex this year. That's the fuel under this tape, not Iran. The real risk isn't an oil shock derailing growth — it's that growth is too strong on top of a headline inflation shock, which is exactly why three Fed governors just dissented against the easing bias. The Fed is quietly walking the path from easing → neutral → hiking bias, and almost nobody is positioned for it. Here's what actually matters from here. One, stop trading the tanker headlines — the conflict is contained and POTUS has zero appetite for ground troops or attacks on energy infrastructure with gas at $4. Two, the AI trade is bifurcating fast: cloud growth at GOOGL and AMZN is the only hard proof of monetization right now, and the market is rewarding discipline (AAPL) and punishing cash-burn (ORCL is FCF-negative $25B and a respected fund manager just called it a bankruptcy candidate). Three, watch ISM prices paid at 84.6 — highest since May 2022 — because that's where the second-round inflation problem shows up first, and that's what eventually breaks this rally. Four, the smokestack trade is real. The bottleneck for AI isn't chips anymore, it's gigawatts. Natural gas turbines, power infrastructure, and the boring industrial names supplying the buildout are where the asymmetry sits while everyone else fights over the same six mega-cap names. The crowd is watching the wrong screen. Oil isn't the headwind. An overheating economy that forces the Fed hawkish — into a market trading at 40x CAPE — is. What Iran Tells Us About UFO Disclosure When governments confront unknown threats in their airspace, defense budgets surge and the same aerospace and surveillance companies move hardest. On March 2nd, Northrop jumped 6% and Lockheed 3.3% on the Iran news — and President Trump has since ordered the formal release of government UAP files, with the Pentagon confirming compliance. So if a conventional conflict can move these stocks this fast, what happens when the bigger story breaks? See the UFOD holdings: [thetruthisoutthereufod.com ETF News A Stock I’m Watching $ITRI ( ▲ 0.24% ) Below the radar screen AI power name that had an undercut and rally on Friday. In Case You Missed It Great talk on with the Acquirers Podcast on markets, value investing, inverse Cramer, and Michael Gayed joins me to talk about taking income from your portfolio and how to get more than 4%…… The H.E.A.T. (Hedge, Edge, Asymmetry and Theme) Formula is designed to empower investors to spot opportunities, think independently, make smarter (often contrarian) moves, and build real wealth. The views and opinions expressed herein are those of the Chief Executive Officer and Portfolio Manager for Tuttle Capital Management (TCM) and are subject to change without notice. The data and information provided is derived from sources deemed to be reliable but we cannot guarantee its accuracy. Investing in securities is subject to risk including the possible loss of principal. Trade notifications are for informational purposes only. TCM offers fully transparent ETFs and provides trade information for all actively managed ETFs. TCM's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. Trade notification files are not provided until full trade execution at the end of a trading day. The time stamp of the email is the time of file upload and not necessarily the exact time of the trades. TCM is not a commodity trading advisor and content provided regarding commodity interests is for informational purposes only and should not be construed as a recommendation. Investment recommendations for any securities or product may be made only after a comprehensive suitability review of the investor’s financial situation.© 2026 Tuttle Capital Management, LLC (TCM). TCM is a SEC-Registered Investment Adviser. All rights reserved. Keep Reading Apr 23, 2026 • 10 min read Beijing Is Not Blinking Apr 21, 2026 • 9 min read Glass: The Last AI Bottleneck? Apr 27, 2026 • 12 min read A Trillion Dollars of GPUs Is Worthless If They Can't Talk View more 2