CNBC The Exchange
2026-09-23 · Hosted by Kelly Evans · CNBC
Executive Summary
The S&P 500 tried for a fourth straight day of gains as the AI trade extended its rally, with the Nasdaq briefly touching a record before holding onto gains. Dan Ives (Wedbush) called Meta's Muse agent "a game changer" that proves the tech giant's massive AI CapEx (capital expenditures, i.e. infrastructure spending) can be monetized, arguing the market is still only in the "third inning" of AI spending and won't reach equilibrium until 2029-2030. A weak two-year Treasury auction (graded a "C-minus" by Rick Santelli) pushed yields to fresh highs, with the curve now at its flattest in a year and a half.
Key Stories & Changes
1. AI Trade Extends Rally, Dan Ives Sees "Third Inning"
Nasdaq held onto a record-setting session gain; S&P was fractionally higher, Dow down 186 points
AMD became the 14th US company to hit a $1 trillion market cap; Apple hit an all-time high for the first time since July
Sandisk up 6% on bullish analyst commentary
Dan Ives called Muse "a game changer," arguing investors previously discounted Meta's AI spending but are now seeing tangible results with a 3.5 billion-user installed base to tap into
Ives: this is a multi-year tech bull market with no equilibrium until roughly 2029-2030, estimating markets are only ~15% through the true AI spending cycle
Flagged Alexander Wang's public persona as effective marketing for Meta's AI push, humanizing the technology versus "AI overlord" narratives around other labs
2. Weak Two-Year Treasury Auction Pressures Yields
$69 billion two-year note auction priced at 4.787%, about a quarter basis point above the when-issued level (a small "tail," graded C-minus by Rick Santelli)
Dealer takedown was larger than the 10-auction average, indicating weaker investor demand
Two-year and 10-year yields are now nearly on top of each other; the yield curve is at its flattest in about a year and a half
Any 2-year yield above 4.75% would mark a fresh high going back to June 2024
3. 52-Week Lows Outpacing Highs Despite Near-Record S&P
30 stocks hit 52-week lows versus only 7 hitting 52-week highs, even as the S&P climbed 1.5% to within 1% of its all-time high
The last time this combination occurred was in 1999
Andres Garcia-Amaya (Zoe Financial): the key difference from 1999 is that earnings are actually delivering, not just "clicks and eyeballs"
He argued the current environment looks more like a mid-cycle than late-cycle market, particularly if the Fed's hiking remains modest (25-75 bps versus the 500 bps post-COVID cycle)
4. Jamie Dimon Quantifies the AI Buildout's Economic Scale
Dimon (via clip): the hyperscaler AI ecosystem grew from $3 billion last year to roughly $7 billion this year, potentially "a trillion" next year
Dimon estimated this could add roughly 1% to GDP growth annually, while also modestly adding to inflation via hiring, factory construction, and equipment/commodity demand
5. AI Shopping Agents Move From Novelty to Real Commerce
Garcia-Amaya described using an invite-only AI agent app called Instinct (similar to Muse) that autonomously found, purchased, and arranged delivery of a printer based on a single natural-language instruction
Raised open questions about consumer comfort sharing payment and home-address data with autonomous agents, and the cybersecurity implications
Noted some large partners (e.g., Amazon) appear hesitant to cede customer relationships to Meta's Muse ecosystem, which could slow broader adoption
6. Peloton Unveils New Treadmill Lineup and AI Coaching
CEO Peter Stern announced the Tread Flex, Peloton's first folding treadmill, targeting price and space as the two biggest barriers to home treadmill adoption
New AI coaching features aim to help runners, walkers, and hikers improve performance
Peloton highlighted record adjusted EBITDA and its first full year of positive net income last year, plus multiple consecutive quarters of revenue growth
Subscription business carries roughly 70% margins; CapEx is kept to about 1% of revenue via an asset-light manufacturing model
Announced its first-ever commercial series bike, going on sale October 1, targeting gym operators; commercial segment posted double-digit growth over the past year
7. Direct Primary Care Startup Uses AI to Cut Employer Health Costs
Dr. Donish Nagda (CEO, Resilient Health) described a subscription-based, employer-direct primary care model bypassing insurance middlemen
Currently operates 15 clinics in 5 states, growing roughly 400% year-over-year
Uses AI to automate administrative tasks and offload routine care tasks, aiming to make AI a "deflationary force" in healthcare costs
Comes as employer healthcare costs are projected to rise nearly 10% this year to more than $19,000 per worker (per Aon); companies like Pepsi, Cigna, and PwC are dropping GLP-1 weight-loss drug coverage
8. Municipal Bond Market Funds Disaster Resilience
Dan Close (Nuveen) noted $63 billion in resilience-infrastructure projects currently in the muni market pipeline
Proposed FEMA reforms could cut federal disaster reimbursement from 90% to as low as 50-75%, pushing municipalities to self-fund hardening projects
Examples cited: Miami's $400 million in pumping stations/sea walls (2024) and NYC's Battery Park $650 million sea wall issuance
Long-dated, double-A-rated muni bonds are yielding over 5% at a discount, translating to roughly 8-10% taxable-equivalent yields for high-tax-state investors
Trends Identified
1. AI Spending Framed as Early-Stage, Not Late-Stage
Dan Ives and Andres Garcia-Amaya both argued the AI investment cycle remains in its early innings, with the "application layer" — consumer-facing tools like Muse and Instinct — only just beginning to scale after years of infrastructure (GPU, memory) investment. This framing supports continued bullishness on both AI infrastructure and emerging application-layer companies despite near-term volatility.
2. Agentic Commerce Raises Both Opportunity and Security Tension
The real-world example of an AI agent autonomously completing a purchase illustrates rapid capability gains, but panelists flagged unresolved concerns around data security, payment trust, and reluctance from platforms like Amazon to cede customer relationships — a friction point that could shape how quickly agentic commerce scales.
3. Market Breadth Divergence Echoes (But Differs From) 1999
The elevated ratio of new lows to new highs despite near-record index levels invites dot-com comparisons, but the consensus view is that current earnings strength differentiates today's market from 1999's speculative, earnings-light rally.
4. Fiscal Pressure Pushes Infrastructure Resilience Funding to Muni Markets
Anticipated cuts to federal FEMA reimbursement rates are accelerating municipal bond issuance for climate and disaster-resilience infrastructure, creating a growing asset class for income-seeking investors amid a backdrop of rising natural disaster costs. ---
Sentiment Analysis
Overall Market Sentiment: Bullish, AI-Led
Guests were broadly optimistic about the durability of the AI trade and equity markets generally, while flagging Treasury market softness and breadth divergence as watch items rather than serious threats.
Risk Factors Highlighted
Weak Treasury demand: The two-year auction's soft grade and rising yields signal potential funding cost pressure ahead.
Market breadth divergence: 52-week lows outnumbering highs despite near-record index levels raises concentration and sustainability concerns.
AI agent security and privacy risk: Autonomous purchasing agents raise unresolved questions about payment security and data protection.
Partner hesitancy to adopt AI agent ecosystems: Some large platforms (e.g., Amazon) may resist ceding customer relationships to Meta's Muse, potentially slowing adoption.
Inflation risk from AI buildout: Jamie Dimon noted the AI infrastructure boom could add modestly to inflation via hiring, construction, and commodity demand.
Employer healthcare cost inflation: Nearly 10% projected cost increases are prompting employers to cut GLP-1 and other drug coverage, shifting costs to employees.
Federal disaster reimbursement cuts: Proposed FEMA rate cuts (from 90% to as low as 50%) would shift significant financial burden onto municipalities.
Peloton's mixed investor perception: Stock down roughly 40% over 18 months despite improved profitability metrics, reflecting a disconnect between fundamentals and market sentiment.
This episode was covered in today's [The Market Signal — 2026-09-23](https://marketsignal.beehiiv.com/p/the-market-signal-2026-09-23), a cross-source synthesis of multiple podcast reports.