Nasdaq And S&P Hit Records As Chip Stocks Rip. The AI Trade Is Back On.

The chip-stock rally that paused last week resumed Friday and accelerated Tuesday. The Qualcomm-Stellantis tie-up gave the cycle a fresh narrative — AI is no longer just a data-center story, it's a vehicle story, an industrial-equipment story, a manufacturing-floor story. Qualcomm provides the chips that handle real-time inference inside the car; Stellantis handles the metal. The deal vindicates the "AI everywhere" thesis without requiring you to believe any single hyperscaler's capex plan.

For the book, the read-through is clear: every name in the AI-power-and-compute sleeve benefits. Nvidia (NVDA) and Advanced Micro Devices (AMD) keep their leadership in training silicon. Qualcomm (QCOM) becomes the inference play for edge devices and now vehicles. Taiwan Semiconductor (TSM) and Samsung make all of it. Vertiv (VRT), Eaton (ETN), Arista Networks (ANET), Bloom Energy (BE), Constellation Energy (CEG), and Vistra (VST) supply the power and the networking. The fact that Korea's Kospi closed at a record on Samsung + SK Hynix tells you that the chip cycle is global, not just American.

We added Qualcomm (QCOM) to the Tier A Aggressive and Tier A Halal models today and reinforced Vertiv (VRT) and Bloom Energy (BE) across all model tiers. See the watchlist sync at the bottom and the cascaded positions on the Portfolios page.

Pope Leo XIV Calls AI A "Threat To Normalize An Anti-Human Vision." It's The Loudest Cultural Pushback So Far.

On Monday in Vatican City, Pope Leo XIV released his first encyclical letter — titled "Magnifica Humanitas" — and the entire document reads like a sharp warning to Silicon Valley. The pontiff said artificial intelligence "threatens to normalize an anti-human vision" and that the concentration of immense digital power "in the hands of a few private actors must be countered." The framing was biblical: humanity is choosing, he wrote, between "constructing Babel" (top-down, pride-driven, profit-driven) and "rebuilding Jerusalem" (diverse hands working together).

The most notable detail of the unveiling: Anthropic safety co-founder Chris Olah stood next to the Pope at the presentation. Vatican officials clarified that this wasn't an endorsement of Anthropic specifically — just a gesture toward the industry as a whole — but Olah used the moment to underscore his own concern: "We will always be influenced by those incentives," he said of AI companies, citing commercial pressure, competition, pride, and ambition. That is why outside voices — faith leaders, regulators, independent researchers — matter. "Who are willing to be our earnest, thoughtful critics."

For the investment book, the encyclical matters less for what it changes about model demand and more for what it signals about the political and regulatory backdrop. Glen Weyl, the faith-and-technology researcher cited in the article, calls the Pope "perhaps the single most important person in the world on AI at this moment." When the moral authority of the world's largest Christian denomination lines up with the workers worried about job losses, the college graduates booing AI-evoking commencement speakers, and the residents protesting energy-hungry data centers, you get an environment where Washington's appetite to regulate goes up. Last week, President Trump delayed a voluntary AI-model testing executive order; that delay is unlikely to last forever.

We're not changing positioning on this. The AI-power and AI-compute sleeves stay in place. But we are flagging it as a thesis-shift risk to watch and have written the full read-through here.

Huawei Says It Has An Answer To U.S. Chip Curbs. The Industry Has To Take That Seriously Now.

Shenzhen-based Huawei Technologies announced Monday that it has developed a workaround — which it calls "LogicFolding" — that allows it to make chips on par with leading 1.4-nanometer products from Intel (INTC), Taiwan Semiconductor (TSM), and Samsung by 2031. Critically, Huawei claims it can do this without the one-of-a-kind extreme-ultraviolet lithography machines made by ASML (ASML) of the Netherlands — the equipment that the U.S. has blocked it from accessing since 2022. He Tingbo, president of Huawei's chip arm, called the approach "feasible and affordable." Huawei says it has already mass-produced 381 chip models using earlier iterations of this method, and the upcoming Kirin smartphone chip launching this fall will be the first commercial product on the new architecture.

The technical story is real, and it cuts in interesting directions. The conventional path squeezes more transistors onto a single silicon wafer; Huawei's path stacks multiple layers of circuits inside a single chip and reduces the time to move data between them. Analysts at Omdia call it "an alternative path forward" rather than a clear winner over EUV lithography. But even an alternative path is enough to change the strategic picture: it means the U.S. export-control regime can slow China down but cannot stop it from getting to the next process node, especially in AI chips.

For our book, the read-through is mixed and worth thinking about carefully. The ASML thesis we've held since early last year — ASML as the toll-collector on every advanced chip globally — gets a little weaker if Huawei's approach works. The big U.S. and Korean fab story (TSM, Samsung) is unaffected for now — they still dominate the leading-edge nodes through 2031. The Nvidia (NVDA) and AMD design story is also unchanged — Huawei's chips will compete with H20 / H200 class hardware, not Blackwell. But the geopolitical premium on rare-earth-independent and Western-fab silicon goes up: we're keeping our overweight to TSM and adding watch-list exposure to Lam Research (LRCX), KLA (KLAC), and Applied Materials (AMAT) where Western-fab capex still has to be funded.

Full read-through with technical detail and the supply-chain ripple is on the specialty page here.

84% Of Small Businesses Now Plan To Use AI This Quarter. That's How The Productivity Story Actually Pays Off.

The Journal's Tuesday feature on Helene Godin's By the Way Bakery — a 14-person gluten-and-dairy-free shop in Pleasantville, N.Y. — quietly contained the most important AI statistic of the week. According to the survey cited in the piece, 84% of businesses with 20-99 employees and 91% of mid-sized firms plan to use AI this quarter. By the Way had previously tried Enterprise Resource Planning software and abandoned it: too expensive, too complicated. Instead, they're now spending "a few hundred dollars a month" on an AI agent built by a consulting firm called Streamliners that pulls daily order data, forecasts supply, and updates the bakery's POS.

This is the part of the AI story that hyperscalers can't actually capture in their quarterly capex disclosures. Microsoft (MSFT), Alphabet (GOOGL), Amazon (AMZN), and Meta (META) all talk about training spend; they don't talk much about the long tail of small-business adoption, because each individual customer is small. But aggregate adoption at that pace pulls compute consumption up structurally and locks in cloud-revenue growth for years. The "AI bust" thesis — that capex outruns actual demand — gets harder to defend when 91% of mid-sized firms have already committed to using the technology.

Our positioning stays overweight Microsoft (MSFT), Alphabet (GOOGL), Amazon (AMZN), Salesforce (CRM), and ServiceNow (NOW) on this read-through. The full case is written up here.

Gavin Baker's Decade-Long SpaceX Bet Is About To Print. Here's What That Means For The Goldman / Morgan Stanley Trade.

Tuesday's WSJ B1 profile of stock picker Gavin Baker is also a window into how the SpaceX IPO — now 15 days out and targeting a $1.5 trillion market cap — ripples through the public-market book. Baker first bet on SpaceX in 2015 at a $10 billion valuation when he was at Fidelity. He bought again in 2019 at a $30 billion valuation through his hedge fund Atreides Management. Both stakes are about to print billion-dollar gains.

For our book, SpaceX itself isn't buyable yet — only insiders and pre-IPO holders are positioned. The trade is in the underwriters and the infrastructure suppliers. Goldman Sachs (GS) and Morgan Stanley (MS) are co-running the book. With OpenAI and Anthropic confidential filings also expected within weeks, the IPO advisory pipeline at GS and MS is the fullest it's been since 2021. The picks-and-shovels suppliers — Nvidia (NVDA), Advanced Micro Devices (AMD), Vertiv (VRT), Bloom Energy (BE), Arista Networks (ANET) — benefit from the cycle as well.

We're reinforcing Goldman Sachs (GS) and Morgan Stanley (MS) across all tiers this week. Full SpaceX countdown thread here.

"AI Addiction Nearly Ruined Him" — And Why That's Going To Be A Regulatory Story By Q3.

The Journal's Tuesday Personal Journal feature on Joe Alary — a 57-year-old Toronto video editor who built an AI "companion" he named AImee, became obsessed, maxed out two credit cards buying server equipment, spent 20-hour days "coding" with ChatGPT, and nearly lost his job — is the kind of story that compounds into political pressure. He's not alone: there's now a Toronto-based "Human Line Project" support group for people coming out of AI delusions. Cases have ended in hospitalization, suicide, even murder-suicide.

OpenAI's response in the article is data-driven: distress-related and self-harm-adjacent prompts are down 65–80% since the GPT-5 default rolled out in October. The company has also expanded crisis-line referrals and built in a "trusted contact" feature. That's the right direction, but the underlying point stands — these stories layered with the Pope's encyclical layered with the campus-protest energy add up to a regulatory tailwind that will eventually hit hyperscaler valuations.

We're not changing model allocations on this. We are flagging it: a 5-7% multiple compression on the big AI names is a plausible 12-month scenario if the regulatory story accelerates. The hedge in the book is the gold position (IAU at 12%) and the defense / dividend sleeve. Full deep-read here.

BP Fires Its Chairman. A Reminder That Energy Major Governance Risk Is Real.

Today's WSJ A1 reports that BP (BP) removed Chairman Albert Manifold — effective immediately — after the board was told he was "verbally abusive and bullying toward employees and had mishandled company information." Manifold had only been in the role since July. The U.S.-listed shares dropped 3.8% in the immediate aftermath. We don't own BP in the models, but the story is a clean reminder for the energy sleeve — ExxonMobil (XOM), Chevron (CVX), Cheniere Energy (LNG), Williams (WMB) — that governance, succession planning, and culture are real risk factors, especially as the European supermajors continue to wrestle with their post-energy-transition identity.

No changes today. Full BP write-up here.

What 1925 Already Told Us About 2026: The 1920s Playbook For Surviving (And Investing Through) The AI Revolution

Every few years a piece of journalism shows up that we end up sending around the office and re-reading two or three times. George Anders's essay in Tuesday's WSJ Journal Report on Artificial Intelligence is one of them. The headline is unassuming — "What the 1920s Can Teach Us" — but the argument inside it is one of the most useful frames we have seen for thinking about how to invest, plan, and live through what the next ten years of AI build-out are going to do to America.

Anders's case is simple: we have been here before. The sociologists Helen and Robert Lynd spent the 1920s in Muncie, Indiana, cataloguing the ways fast cars and open roads were rewriting courtship, church attendance, and the family parlor. Henry Ford asked himself "Are we moving too fast?" and answered "No." Electricians went from a non-profession to 80,000 jobs by 1930. 168,000 Americans worked in rubber factories. 450,000 built roads. Local entrepreneurs opened 20,000 movie theaters across America. In 1920, electricity reached 35% of homes; by 1930, 68%. The auto / electricity / radio revolution didn't just sell cars — it rewrote the entire American labor market and minted multi-generational fortunes in the spillover trades.

We think the AI parallel is the cleanest historical analogue we have. We are deliberately positioning the book in the spillover layer — the chip and power and grid and security names that the AI cycle pulls forward — rather than the consumer-AI application layer (which is the "movie theaters" of this cycle: high-beta, fashion-driven, easy to over-pay for). The piece also lands on the John Doerr "biggest tsunami ever" interview from the same WSJ section: Doerr puts AI bigger than the PC, browser, iPhone, and cloud combined. We agree directionally — and we think the adoption curve is compressing the 1920s decade-long electrification ramp into roughly four or five years.

The long read — with the full 1920s → 2026 spillover-trade comparison table, the safety / regulatory parallels, and the Capital Wealth positioning — is on the specialty page here. This is the piece we'd recommend sending to clients who keep asking us "what makes you confident about this AI thing?" It is also the cleanest historical defense we can offer for the way the portfolios are tilted today.

Ferrari Just Built A $640,000 Electric Glass Spaceship. Jony Ive Designed It. Corning Made The Glass.

On Sunday in Rome, at the Vela di Calatrava stadium, Ferrari Chairman John Elkann unveiled the company's first car without an internal combustion engine. The car is called the Luce — Italian for "light." Starting price is €550,000 / ~$640,000. It does 0–60 in under 2.5 seconds with a top speed above 190 mph, has four motors (one per wheel), and is the first Ferrari ever to seat five people. The entire upper half is glass, much of it supplied by Corning. The physical design was done in partnership with Jony Ive and Marc Newson through their post-Apple design house LoveFrom.

What makes this strategically interesting: Ferrari is launching an EV at exactly the moment Porsche, Lamborghini, and McLaren stepped back from EV roadmaps, and Ford and Stellantis took billions in write-downs on their mass-market EV bets. Ferrari's bet is that the luxury EV market is structurally different from the mass-market — the buyer of a $640,000 second car doesn't care about charging infrastructure, range anxiety, or total cost of ownership. Elkann's most pointed quote, in a clear shot at Tesla: "As a car becomes electric, it doesn't mean that it needs to be a consumer electronics object, which is probably one of the mistakes that the industry has been making in the last 10 years."

For the book: we do not own Ferrari (RACE) — it trades over 40x forward earnings and the Luce launch is in the price. But three names we do own get clean read-through. Corning (GLW) in the Tier B Fundamentals Core sleeves — the Luce glass contract is small in dollars but huge in brand validation. Stellantis (STLA) in the Dividend tier — same Elkann family, same EV-vs-combustion conviction, much cheaper multiple. Constellation (CEG), NextEra (NEE), Vertiv (VRT), and GE Vernova (GEV) across the AI-power sleeve — every electrified mile pulls electron demand forward. Full Ferrari Luce write-up with specifications and the secondary-market depreciation analysis here.

Russia Threatens The Baltics. The Pentagon Pulls Back From NATO. Defense Stocks Keep Catching A Bid.

Today's front page carries two related stories: Yaroslav Trofimov reports that Russia is "increasingly bellicose" against the Baltic states and that European national-security officials warn Moscow could target Latvia, Lithuania, or Baltic Sea islands to test NATO cohesion. Michael Gordon and Robbie Gramer report that the Pentagon is scaling back the forces it plans to send to Europe in a crisis — another step by Washington to reduce its NATO commitment. The combination — greater perceived threat, less U.S. backstop — pushes every European NATO member toward higher defense spend.

Our defense sleeve — Lockheed Martin (LMT), RTX Corp (RTX), General Dynamics (GD), Northrop Grumman (NOC), L3Harris (LHX) — benefits directly. We also hold European exposure through Rheinmetall (RHM.DE via the BAE / Leonardo basket) and BAE Systems (BAESY) where the order book is now extending past 2030. The Pentagon's posture shift is, ironically, the most bullish factor for non-U.S. defense contractors we've seen this year.

Reinforcing Lockheed Martin (LMT) and RTX (RTX) across all tiers today. Full geopolitical brief here.