AI의 마지막 전술적 단계

THE DAILY H.E.A.T. Torching Wall Street’s Obsolete Playbook. Delivered straight to your inbox 5 days a week. Subscribe Home Posts AI's Last Tactical Mile AI's Last Tactical Mile Anduril and Nokia just announced something Wall Street misread. The investable trade isn't the names on the announcement. May 7, 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 News vs. Noise: AI's Musical Chairs ETF News In Case You Missed It H.E.A.T. Anduril and Nokia just announced a 5G tower designed to be deployed in three hours, anywhere on earth, with no infrastructure. Wall Street read it as a press release. We read it as a multi-year procurement signal — and the publicly investable winners are not the names on the announcement. THE SETUP Last week, Anduril Industries — the privately held defense technology company last marked above $30 billion — announced the newest member of its Sentry tower family: a 5G-equipped variant called the 5G Comms Sentry Tower, built in partnership with Nokia Federal Solutions. The product description, in plain English: a self-contained tower with onboard power, compute, and Nokia private 5G radios that — Anduril claims — can be airdropped into a remote location and provide cellular coverage for several kilometers within three hours, with production already at scale. No fixed infrastructure required. Lattice software for administration. Service-pricing model. The trade press read this as a product announcement. We read it as something larger. The 5G CST is not the news. The news is what the 5G CST tells you about where the Pentagon's communications budget is migrating — and which publicly traded companies are positioned to capture that migration. "5G CST is purpose-built for environments where commercial coverage is unavailable or impractical to extend due to cost, time, or infrastructure constraints." — Anduril Industries, May 2026 product announcement Here is the structural setup, in three sentences. For thirty years, US military communications have been built on a stack of tactical radios, satellite communications, and proprietary mil-grade waveforms — a legacy architecture that is expensive, low-bandwidth, and increasingly inadequate for the data volumes that AI-enabled platforms generate. The DoD has spent the last six years quietly migrating toward private 5G as the answer; pilot programs at multiple US bases have moved from research-phase to operational, the Open RAN — Open Radio Access Network, a software-defined cellular architecture — has been embraced as a procurement priority for both cost and supply-chain-security reasons, and CHIPS Act money has flowed to domestic Open RAN equipment makers since 2023. Anduril–Nokia is the most recent commercialization of a procurement direction the Pentagon has been telegraphing publicly since at least 2020, and the most concrete sign that the migration is moving from pilot to production. The reason this matters now is simple: battlefield AI is turning radios into a choke point. Autonomous systems do not fail because the model is dumb. They fail because the node cannot move enough data to make the model useful. A single Lattice-equipped autonomous platform produces sensor fusion outputs that overwhelm legacy tactical radios. Multiply that by hundreds of platforms a forward base will field by 2028 and the math demands cellular bandwidth in places that have never had cellular coverage. The DoD is quietly standardizing around a truth the commercial world learned a decade ago: if you want bandwidth, density, and upgradeable software-defined networks, you build cellular. BY THE NUMBERS <3 hrs Anduril 5G CST stated deploy-to-operational time Anduril, May 2026 400+ Sentry towers deployed globally since 2017 Anduril product disclosure Since 2020 DoD has funded 5G-to-NextG efforts; signal is shift from pilots to procurement-scale deployments DoD 5G strategy, public THE CONSENSUS IS WRONG The consensus mistake is reading the Anduril–Nokia announcement as an Anduril story. Anduril is private and most readers cannot buy it directly. Even if they could, Anduril's secondary-market valuation already reflects optimism about its product roadmap. The investable expression of the announcement is not Anduril. It is the publicly traded names that will receive task-order revenue from every DoD private-5G deployment that follows this template — Nokia at the radio layer, ruggedized hardware suppliers at the platform layer, edge-compute and RF-semi names at the silicon layer, and integrators at the systems-of-systems layer. Anduril is the most visible name in the room. It is not the most investable one. The intellectual move here is the same one we made in the memory issue last week. Don't buy the bottleneck. Buy the tollbooth. In the memory issue, the bottleneck was HBM, and the tollbooths were the equipment makers — Hanmi, Disco, BESI, Advantest. In this issue, the bottleneck is battlefield AI bandwidth, and the tollbooths are the names that build, harden, and integrate the tactical 5G stack. SPOTLIGHT THE FOUR-LAYER TACTICAL 5G STACK The defense private-5G supply chain — like the memory supply chain — is not one industry. It is four. Mapping the layers tells you exactly where the money flows when the migration accelerates. Layer 1 — Radio access and core network. Nokia and Ericsson dominate. Nokia Federal Solutions is the US-incorporated subsidiary that holds the relevant clearances and has been winning DoD pilot work for years. Ericsson is the closest comparable. Both are Open RAN leaders. This is the layer that gets the marquee partnership announcements and is most directly exposed to the architectural shift. Layer 2 — Ruggedized hardware and tower platforms. The companies that build the physical infrastructure that survives in austere environments. Anduril dominates the autonomous-tower category but is private; the publicly traded analogs are the ruggedized-systems specialists — Mercury Systems, Curtiss-Wright, Kratos. These are the names whose products get certified for forward operating bases, missile fields, and deployable command posts. Layer 3 — RF semiconductors and edge silicon. Every 5G radio is a stack of specialized silicon: RF front end (power amplifiers, filters, switches), networking and switch silicon for the data plane, RAN compute and acceleration for the baseband, timing and sync chips, and edge-compute modules for local AI inference. Qorvo and Skyworks are the cleanest pure-play exposures to the RF front end — though investors should note their defense-segment exposure is smaller than the handset narrative implies. Marvell has meaningful exposure to networking and custom silicon that shows up in both core and edge builds. The point is not to pick a single silicon winner. The point is that every tactical 5G deployment pulls demand through the entire silicon stack, and the names that supply it are mispriced on consumer-handset overhangs that mask the infrastructure exposure. Layer 4 — Systems integrators and software platforms. Booz Allen, Leidos, SAIC, CACI capture per-deployment dollars regardless of which radio, which tower, or which chip wins. Palantir is the platform-level beneficiary as the data fabric for everything the new comms stack carries. L3Harris is the pivot story — the legacy tactical-radio incumbent that has either correctly migrated to 5G or will be displaced by it, depending on execution. The quiet point: The Anduril–Nokia announcement touches all four layers. Most investors will only think about Layer 1, because that is where the partner is named. The other three layers are where the underowned exposure sits. THE TRADE, IN ONE LINE Long the four-layer tactical 5G stack — radios, ruggedized platforms, RF semis, and integrators — through the FY26-to-FY28 deployment cycle. Don't buy Anduril at secondary marks. Buy the publicly traded names that supply, integrate, and depend on every Anduril-style deployment that follows. Below we tier the names by conviction. WINNERS — TIERED BY CONVICTION Tier 1 names are structurally levered to the tactical 5G migration with publicly disclosed defense exposure today. Tier 2 names require the migration to scale on the FY26–FY28 window. Tier 3 names are asymmetric — smaller positions, larger potential payoffs. TIER TICKER / NAME THESIS TIER 1 Nokia (NOK) The named partner on the Anduril 5G CST. Nokia Federal Solutions is the cleared US subsidiary winning DoD private-5G work directly. Trades at a meaningful discount to Ericsson on Open RAN positioning despite being the better-positioned name on US defense exposure. We acknowledge the decade of value-trap baggage; the call here is structural, not narrative — Nokia is sitting in the federal channel where the early dollars flow, and the federal channel is what is changing. TIER 1 Palantir (PLTR) Lattice — Anduril's software platform — is the closest competitor to Palantir's Maven. But every battlefield 5G deployment generates the data fabric Palantir specializes in operationalizing. Wins on Anduril deployments and on every alternative deployment that uses Maven instead. Layer 4 anchor. TIER 1 Mercury Systems (MRCY) The publicly traded ruggedized-systems specialist most directly exposed to forward-deployed AI compute. Builds processing modules and RF subsystems that go into exactly the kind of austere-environment tactical platforms Anduril is fielding. Underowned in defense baskets that overweight the primes. TIER 1 Booz Allen / Leidos (BAH / LDOS) Defense systems integrators capture per-deployment dollars on every DoD 5G rollout regardless of which radio or tower vendor wins. The same logic we applied to the AI vendor issue applies here: you do not need to pick the winning model, you need exposure to the integration layer that operates underneath all of them. TIER 2 Ericsson (ERIC) The Layer 1 alternative to Nokia. Less direct DoD exposure today but a clean Open RAN play. Wins if the procurement migration broadens beyond Nokia-anchored deployments. Useful as a paired exposure with NOK rather than a substitute. TIER 2 Curtiss-Wright (CW) / Kratos (KTOS) Layer 2 ruggedized-hardware specialists with established defense customer relationships. Curtiss-Wright is the more mature, lower-multiple name; Kratos is the higher-beta, higher-volatility play. Both win on tactical communications buildouts that go through traditional procurement channels rather than direct-to-Anduril. TIER 2 Qorvo / Skyworks (QRVO / SWKS) RF front-end semiconductors that go into every 5G radio shipped — commercial or defense. Defense-qualified part wins are growing as a share of mix. Both names trade at depressed multiples on consumer-handset overhangs that mask the underlying defense and infrastructure exposure. Asymmetric setup at current valuations. TIER 2 Marvell (MRVL) Networking and custom silicon exposure that shows up in both core network and edge builds. Less defense-specific than the RF-front-end names but increasingly relevant as edge-compute workloads in tactical 5G deployments push for custom silicon. Layer 3 diversified exposure rather than pure-play. TIER 3 L3Harris (LHX) The pivot trade. Legacy tactical-radio incumbent whose franchise is either being modernized into the 5G era or quietly displaced by it. If LHX executes the pivot, the stock re-rates as a Layer 1+4 hybrid. If it doesn't, the legacy radio franchise becomes the Pressure Point. Asymmetric in either direction. Position as a research idea, not a high-conviction long. TIER 3 Anduril (private, secondary) Private-market exposure available only to qualified investors via secondary platforms. Marks already reflect significant optimism about product roadmap. We think the public-market substitutes in Tier 1 are better-priced and currently underowned. Mentioned for completeness, not recommended. PRESSURE POINTS — WHERE THE RISK IS TIMING, NOT FAILURE These names are not businesses that fail. They are businesses where the consensus is mispricing the risk of the architectural migration we have described. TIER TICKER / NAME THESIS PP1 SATCOM (case-by-case, not a blanket short) Important nuance: tactical 5G is local distribution, not long-haul backhaul. More 5G nodes producing more sensor data means more backhaul demand — which often means more SATCOM, particularly LEO. The pressure is segmented, not blanket. Specific legacy radio-terminal budgets and proprietary tactical-waveform programs may lose dollars to ground-based 5G distribution gear. Backhaul capacity, resilient links, and LEO constellations may gain dollars from the same migration. Treat SATCOM as a segmented opportunity, not a substitute story. PP2 Legacy tactical radio franchises The L3Harris tactical-radio book is the clearest example. Same company appears in our Tier 3 winners list because the pivot story cuts both ways. Investors who own LHX as a tactical-radio play rather than a 5G migration play are exposed to the wrong end of the same trade. PP3 GPU and accelerator names as defense AI plays We made this point in the memory issue and it applies again here. NVIDIA, AMD, Broadcom are AI customers in defense workflows, not AI beneficiaries. They sell the compute that feeds the bottleneck. Accelerator makers do not own the bandwidth bottleneck. They depend on it being solved. PP4 Pure-play primes without comms exposure Lockheed, Northrop, Raytheon will sell platforms into every battlefield AI deployment. They will not capture the comms-stack rents we have described. If you own primes for AI-defense exposure, you own the wrong layer of the stack for this specific trade. Hold them for the platform-level reasons; do not double-count them as comms exposure. THE HONEST BEAR CASE BEAR CASE This can absolutely be a 2028 story — and markets will punish you for being early even if you are right. The DoD has been running private-5G pilots since 2020. Commercialization timelines in defense procurement are notoriously slow. The Anduril–Nokia announcement is a single product launch, not a budget appropriation. The procurement signal is real but the dollar flow could lag the narrative by two to three years. We are also being directional rather than line-itemed on FY26 budget specifics. The DoD has named private 5G as a procurement priority in published strategy documents, and individual base-level deployments are documented. We are not making a precise claim about FY26 appropriated dollars because line-item disclosures on tactical 5G are not consistently published. The thesis depends on the migration being real and accelerating, not on a specific budget number we can cite. Anduril could also fail to scale. Defense procurement is unforgiving to companies that promise rapid production and miss. If the Sentry family fails to deliver on the volume claims in the May announcement, the entire commercialization narrative we are building around it weakens — and the public-market names that ride the same migration get re-rated downward. Finally, the public-market names we are tiering as winners have meaningful issues we are not pretending away. Nokia has been a value trap for a decade. Mercury Systems has had execution problems. Qorvo and Skyworks carry consumer-handset overhangs. L3Harris is in our Tier 3 winners and PP2 pressure points simultaneously precisely because the pivot is uncertain. None of these names are clean. The thesis is that the architectural migration is large enough to overcome the individual company-level baggage, but readers who do not believe that should size positions accordingly or pass entirely. WHAT TO WATCH NEXT THE SIX SIGNALS THAT SAY TACTICAL 5G IS REAL The procurement migration we are describing either accelerates over the next eighteen months or it doesn't. These are the concrete signals — drawn from earnings decks, budget documents, and procurement announcements — that will tell you which way it is going. We track these forward; readers should too. 1. FY27 DoD budget request released with explicit private-5G or Open RAN line items above pilot-program scale. The transition from research-and-development funding to procurement funding is the single highest-conviction signal. 2. Nokia or Ericsson naming DoD or federal-solutions contracts above the pilot threshold on quarterly earnings calls. Vocabulary shift from "opportunities" to "awards" is the trigger. 3. Mercury Systems, Curtiss-Wright, or Kratos disclosing tactical-comms-related design wins or program awards — particularly anything that names autonomous-platform integration. 4. Anduril announcing a second or third major Sentry 5G CST deployment — ideally outside CONUS and ideally with a named end-user. Production-scale deployment signals follow product launches by twelve to eighteen months in defense procurement when the migration is actually happening. 5. L3Harris commentary on quarterly calls shifting from "protecting tactical radio franchise" to "transitioning to integrated 5G/radio portfolio." The vocabulary shift is the early warning that the legacy book is being rebuilt. 6. Open RAN consortium milestones — particularly any FCC or NTIA procurement actions that designate Open RAN as a federal procurement standard. This is the policy lever that converts a strategy document into binding capex flow. FIVE TAKEAWAYS 1. The Anduril–Nokia 5G CST announcement is a procurement signal, not a product story. The DoD is rebuilding tactical communications around private 5G. The buildout window is FY26–FY28 and the publicly investable winners are not Anduril. 2. The cleanest one-line expression: don't buy the bottleneck — buy the tollbooth. The bottleneck is battlefield AI bandwidth. The tollbooths are the four-layer stack: radios (NOK, ERIC), ruggedized platforms (MRCY, CW, KTOS), RF semis (QRVO, SWKS, MRVL), and integrators (BAH, LDOS, PLTR). 3. Wall Street is trading AI infrastructure as if it ends at the datacenter door. It doesn't. Battlefield AI generates more data than legacy tactical comms can carry. The companies that supply the bandwidth are mispriced relative to the companies that supply the AI. 4. If you want the cleanest single-ticker expression of this trade, it is Nokia — not because Nokia is sexy, and not because the value-trap history is forgotten, but because Nokia Federal Solutions is sitting in the cleared federal channel where the early dollars flow. The named partner on the Anduril announcement, structurally exposed to every DoD private-5G deployment that follows the same template, trading at a discount to Ericsson on Open RAN positioning despite being better positioned on US defense exposure. 5. Bear case is timing. This can be a 2028 story, and markets will punish early positioning. The architectural migration is real. The dollar flow could lag the narrative by two to three years. Use the six signals in the section above to track whether the buildout is accelerating or slipping — and size positions accordingly. 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 News vs. Noise: AI's Musical Chairs Noise: Another peace deal headline, another semiconductor melt-up, another day where Arm trades $6 billion of turnover whipsawing 33% in a few hours on numbers that were mediocre to begin with. The market wants to believe two things simultaneously: that Iran is permanently solved and that every AI company will get every chip it ordered. Neither is true. The tape doesn't care. Signal: The Anthropic-SpaceX deal isn't a one-off — it's the sound of musical chairs starting in the LLM business. Anthropic just doubled Claude Code's rate limits overnight because Elon is leasing out Grok's compute. That only happens when Grok doesn't need it. Meanwhile OpenAI quietly slashed its build commitment from $1.4 trillion to $600 billion and walked away from Norway and the UK. These are not the moves of a company hitting its targets — they're the moves of a company whose CFO is publicly worried about compute obligations outrunning revenue. Anthropic is taking the enterprise and coding workflow, Gemini owns distribution through Search, Workspace, and Android, and the rest of the field is starting to look like inventory. A year from now there will be fewer material LLM players, and the compute they thought they needed will get repurposed — exactly what we're watching in real time. The thing the market is mispricing: It is treating semiconductor, memory, and storage supply constraints as a permanent condition. Samsung adds $190 billion overnight on a deal that might happen with Apple. Intel tacks on $60 billion. Taiwan Semi — the company that would lose the business — adds $90 billion anyway. A quarter of a trillion dollars in market cap on a maybe, while Samsung's union prepares an 18-day strike demanding $27 billion. Fifteen names account for the entire $4.8 trillion the S&P has added this year. Stretch the definition of "chip exposure" to include Google and Amazon's silicon and you get to 87.5% of the index's gain. That is not a market — that is a single trade wearing 500 different tickers. What to do about it: Own the layer that gets paid no matter which LLM wins. The compute shortage is real, but the question of who burns the compute is suddenly very much up for grabs. That argues for capital equipment (LRCX, ICHR), memory (MU, SNDK, SK Hynix), and the photonics/optical layer (CRDO, ALAB) — the picks-and-shovels names that don't care whose logo is on the model. Avoid getting cute with second-tier AI software names that need a specific frontier lab to keep writing checks; the circular-financing web is one missed revenue target away from getting much smaller. Fade the speculation premium. Samsung up 19% on a maybe-deal with Apple, Intel up $60 billion on the same headline, Taiwan Semi up $90 billion on news that should hurt it — these are tells. When the loser of a deal trades like the winner, the trade is positioning, not fundamentals. Trim into the melt-up, don't chase it. On Iran: the day a real deal prints is the day this rally loses its wall of worry. Refining capacity is still the durable energy trade — VLO and MPC don't need a war to work, they need the structural shortage we've had since the 1970s. Use ceasefire-driven crude weakness to add, not subtract. And keep an eye on the next earnings tell. Nvidia on May 20th matters more than any peace memo. If hyperscaler capex commentary starts sounding hedged the way OpenAI's build plans suddenly do, the 87.5% of the S&P that is the chip trade gets a very fast haircut. Position accordingly before the print, not after. 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 $MEMY Holdings Update: We replaced $TEM ( ▼ 7.53% ) $KTOS ( ▼ 7.35% ) and added $PAAS ( ▲ 2.16% ) $FNV ( ▼ 2.75% ) All 5% positions. For a full list of MEMY holdings, visit: https://incomeblastetfs.com/etf/memy Distributor: Foreside Fund Services, LLC A Stock I’m Watching $PENG ( ▼ 0.03% ) I typically don’t like charts that look like this but this name spans the memory AND photonics space and is somewhat under the radar screen at the moment. 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 20, 2026 • 10 min read The Last Undiscovered Bottleneck: Why the Next $NVDA Moment Belongs to Light, Not Silicon Apr 17, 2026 • 9 min read The Weight Loss War Just Got a New Front, and One Stock Has the High Ground. May 6, 2026 • 4 min read What They Do > What They Say View more 3