CNBC The Exchange
2026-06-05 · Hosted by Kelly Evans · CNBC
Executive Summary
Kelly Evans framed a session where, unusually, “today’s trade broadly speaking is not at all driven by AI” — healthcare, financials, and the consumer led as United Health and Goldman topped the Dow while Broadcom fell 14% on lackluster guidance. The episode probed whether the market can hold up without AI leadership, the long-awaited broadening into small caps (outperforming large caps by ~600 bps YTD), and a granular look at the $1.8 trillion SpaceX IPO (95x sales) including a bull case for data centers in space. Continuum (QNT), a Honeywell quantum spin-off, opened at $68 after pricing above range at $60, prompting a debate on quantum’s ~$75 billion sector legitimacy. Bitcoin slumped to ~$63,000 (mining cost ~$87,000), with Hut 8 highlighting its pivot to AI data centers (stock +176% YTD). A key emerging theme: rising AI cost discipline, with companies “model routing” cheaper Chinese/open-source models (Claude Opus costs ~19x DeepSeek) — a risk to OpenAI and Anthropic’s premium-pricing IPO narratives.
Key Stories & Changes
1. A Non-AI-Led Market Day
Dow winners: United Health and Goldman; credit-card names and Merck in the top five; even lagging Walmart +1%
AI trade lagged: semis and memory lower; Broadcom -14% on weak guidance
Travis Prentice (Informed Momentum): a “reversal trade” / healthy consolidation; semi and memory medium-term trends remain “very powerful”; time to manage risk on parabolic names
Costco posted an 8% monthly comp; Victoria’s Secret also strong — a “stock-by-stock” market
2. SpaceX IPO & Data Centers in Space
$1.8 trillion valuation = ~95x last year’s sales; largest IPO ever, ~3x the largest US listing; more than double its valuation a year ago
~30% to retail; FT reports Goldman says SpaceX AI revenue can grow 100x by 2030
Duncan Davidson (Bullpen Capital): dismisses “lambs to the slaughter” fears; expects ~15% aftermarket pop; bullish on data centers in space (StarCloud already launched one); Super Heavy to lower launch costs; agents (vs. LLM training) ideal for space data centers
Davidson: “We’re at the beginning of the IPO AI wave, not the end”
3. Continuum (QNT) Quantum IPO
Honeywell spin-off priced above range at $60, opened at $68 (~12%+); US government gave $100M for a quantum-chip supply chain
Wedbush’s Antoine Legault: a “marquee” institutional-grade public listing legitimizing the ~$75 billion quantum sector; expects it to reach hundreds of billions, possibly $1 trillion+, in 5–10 years
Quantum advantage / inflection expected ~2028–2030; modest 2025 revenue (tens of millions); direct comparable to IonQ
Dave Cody (former Honeywell): quantum will be “transformational”; one computer could exceed all of today’s compute power
4. Bitcoin Slump & Mining Pivot to AI
Bitcoin at lowest since February (~$63,000), fourth straight week of losses; Oppenheimer says mining now costs ~$87,000 per coin
Hut 8 CEO Asher Genoot: stock +176% YTD on pivot to AI data centers; $17 billion of deals signed; 8+ gigawatts development pipeline; spun out mining into “American Bitcoin” (50%+ margins, mining near $40,000/coin)
Network difficulty falling as miners exit makes remaining mining more efficient
5. Labor Market & AI Washing
Challenger: job cuts rose 16% in May to the highest May total since 2020, with AI cited as the leading cause
Evan Sohn (Revelio Labs): AI is “reshuffling, not breaking” the job market; nearly half of US job postings now require AI skills; info-sector wages up ~4.4–4.5% MoM; advises workers to acquire AI skills
Surge in professional/business-services job openings tied to AI consulting demand
Street expects ~80,000 May payrolls; Sohn at 123k
6. AI Cost Discipline & Model Routing
Deirdre Bosa: companies increasingly skip Anthropic/OpenAI for cheaper Chinese/open-source models
Claude Opus costs ~19x more than DeepSeek but isn’t 19x more capable
“Model routing” — easy tasks to cheap models, hard tasks to frontier models — complicates the frontier labs’ premium-pricing IPO narratives
Victoria Greene (G Squared): a buyer of Broadcom and CrowdStrike on weakness — Broadcom’s print “was fine,” still growing AI semi revenue ~200% next quarter, $10B free cash flow; favors staying balanced with health/financials
AVGO: Broadcom — -14% (-11% later) — Lackluster guidance; some buying the dip
UNH: United Health — Higher — Dow winner on healthcare leadership
GS: Goldman Sachs — Higher — Dow winner on financials/IPO excitement
QNT: Continuum — Opened +12% ($68) — Quantum IPO legitimizing the sector
HUT: Hut 8 — +176% YTD — Bitcoin miner pivot to AI data centers
BTC: Bitcoin — ~-2% to $63K — Below mining cost (~$87K), 4th weekly loss
Trends Identified
1. The Long-Awaited Broadening Beyond Mega-Cap Tech
For the first time in a while, the market held up on a non-AI-led day, with healthcare, financials, and small caps (outperforming large caps by ~600 bps YTD) leading. The thesis: making AI happen requires broad participation across industrials, energy, and utilities, while small caps offer the “next new trend” (NPUs, sensors for physical AI).
2. Stock-by-Stock Dispersion Within Themes
Travis Prentice and others stressed that even within monolithic trades like AI and software, dispersion is emerging — strong results (Costco’s 8% comp, Victoria’s Secret) coexist with weak ones, and software/cyber require selectivity at resistance levels. The market rewards observing price signals over predicting.
3. AI Cost Discipline Threatens Premium-Pricing Narratives
The shift from “how fast can we deploy” to “where’s the ROI,” combined with model routing to cheaper Chinese/open-source models (Claude Opus 19x DeepSeek), pressures the frontier labs’ growth-and-premium-pricing IPO cases — with potential knock-on risk to the broader AI supply chain levered to their growth.
4. Infrastructure Convergence: Bitcoin Mining, AI, and Quantum
Hut 8’s pivot illustrates how energy infrastructure built for Bitcoin is being redeployed for AI data centers (and how quantum will also require data centers per Dave Cody). Low-cost power and capex efficiency are becoming the unifying competitive moat across compute-intensive technologies.
5. AI Washing in Layoffs vs. Real Skill Demand
Challenger blamed AI for May’s job cuts, but Revelio’s data shows nearly half of postings now require AI skills and rising tech wages — suggesting firms cite AI to cut while selectively hiring, and that the real opportunity is acquiring AI skills, especially in consulting. —-
Sentiment Analysis
Overall Market Sentiment: Broadening / Constructive
The mood was upbeat about the market’s ability to hold up without AI leadership, tempered by AI-cost concerns and stretched IPO valuations.
Risk Factors Highlighted
AI cost gap / model routing: Cheaper Chinese/open-source models (Claude Opus 19x DeepSeek) threaten frontier-lab premium pricing and the broader AI supply chain.
SpaceX valuation: 95x sales for an unprofitable company with debt; data-center-in-space economics still “marginal.”
Quantum’s unproven economics: Modest revenue today; commercial inflection not expected until ~2028–2030; “riskier moment.”
Bitcoin below mining cost: Mining costs ~$87K vs. ~$63K price; fourth straight weekly loss.
Semiconductor consolidation risk: Parabolic names need risk management even within strong trends.
Labor-market softening / AI washing: May job cuts up 16%; AI cited but possibly masking firm-specific weakness.
Screwworm demand impact: Unwarranted fears could soften beef demand despite food-supply safety.
Quantum encryption threat: Future quantum machines could break encryption, forcing defensive spending.
This episode was covered in today’s The Market Signal — 2026-06-05, a cross-source synthesis of multiple podcast reports.