CNBC Halftime Report
2026-10-22 · Hosted by Scott Wapner · CNBC
Executive Summary
The Investment Committee debated whether the AI trade is set to resume after Citadel Securities' Scott Rubner argued the midterm elections represent a "clearing event for risk" that will let AI stocks outperform again. Joe Terranova agreed, saying hedge funds and portfolio managers have "run for cover" and positioning is light, setting up a rebuild into year-end. Steve Weiss pushed back hard, arguing it's not that simple — he's rationalized his AI exposure toward higher-quality names like Meta, Alphabet, and Nvidia while avoiding CoreWeave, Anthropic, and OpenAI, which lack EBITDA multiples (earnings before interest, taxes, depreciation and amortization — a cash-flow proxy) to anchor valuation.
Key Stories & Changes
1. The AI Trade "Clearing Event" Debate
Citadel Securities' Scott Rubner (cited by Joe Terranova): midterm elections are a "clearing event for risk," after which clients are "happy to buy higher but not buy now" ahead of a major political event
Terranova: hedge fund and mutual fund positioning has gotten "incredibly light," setting up a rebuild into early November and a strong close to 2026
Steve Weiss disagreed with the "everyone's run from the trade" framing, saying there's been a rationalization toward quality rather than a wholesale exit
Weiss said he's largely exited Vertiv and GE Vernova but kept Meta as his largest position, alongside Taiwan Semiconductor
2. Steve Weiss vs. Joe Terranova on Valuation Discipline
Weiss favors companies with an EBITDA multiple he can anchor to, citing Nvidia at 10 times next year's earnings; he explicitly avoids CoreWeave, Anthropic, and OpenAI because they lack comparable multiples
Weiss noted Anthropic has said it will need $750 billion over the next several years to fund its buildout
Terranova countered he wants exposure to semi-equipment names (KLA Corp, Lam Research, Marvell, Corning, Micron) that have underperformed 10-12% over the last quarter but aren't "high-beta" to the AI trade
Micron is up 220% year-over-year yet trades under 10 times earnings, which Jenny Harrington said reflects skepticism about earnings sustainability rather than cheapness
3. Jenny Harrington's Case for Meta Ahead of Earnings
Harrington: Meta trades at 21 times earnings with 24% expected earnings growth this quarter, calling its earnings more reliable than other AI trade names
Said she's unsure Muse (Meta's AI assistant) translates into near-term earnings growth, but expects the company to describe it positively regardless
On AMD: she noted only about 5% of AMD's revenue is attributable to Meta, raising questions about who makes up the other 95% of AI-driven demand
4. Netflix's Sixth Straight Post-Earnings Decline
Shares at their lowest level since August 2024, down 47% over the past year
Weak Q4 revenue guidance; ad revenue outlook cut to $2.7 billion from a prior $3 billion
CEO Ted Sarandos said the company remains focused on original film/series, licensed content, and live events, calling YouTube-style experiments "relatively small"
Wells Fargo (underweight rating) warned that if engagement doesn't improve, members will move from "less time spent" to "fewer logins" to "cancellations," calling the content-chief change "the canary in the coal mine"
Committee consensus: the stock's cheap valuation (18-19x) does not offset the loss of its growth narrative
5. Calls of the Day: Energy, Hospitality, Defense, Software
XOM: Exxon — Neutral at BofA, $181 target — Limited upside flagged, tied to crude oil direction
VLO: Valero — Neutral at BofA, $471 target — Joe Terranova bullish on refining margins; sees $500 in play
MAR: Marriott — Outperform at Morgan Stanley, $403 target — Jenny Harrington flagged elevated 31x earnings and thin 2.5% FCF yield despite liking the business
NOC: Northrop Grumman — Outperform at Baird, $675 target — Harrington more constructive given $110 billion pipeline and 18x earnings
ZM: Zoom — Upgrade to outperform at Oppenheimer, $130 target — Terranova questioned the rally; highlighted Zoom's early Anthropic stake, now worth multiple billions ahead of its IPO
Trends Identified
1. From Blanket AI Enthusiasm to Selective Quality Rotation
The committee's debate illustrates a market-wide shift from treating "the AI trade" as one basket to differentiating sharply between companies with visible earnings and cash flow (Meta, Nvidia, Alphabet) versus those without (CoreWeave, Anthropic, OpenAI). Terranova explicitly framed this as "narrowing toward quality," a dynamic he expects to expand to other large-cap AI names as the year closes.
2. Valuation Compression Despite Strong Fundamentals
Panelists noted that AI-exposed stocks like Micron (up 220%) and Nvidia now trade at historically low multiples relative to their growth, reflecting persistent skepticism about earnings durability rather than a lack of demand. Carrie Firestone noted the S&P is trading below its 10-year average multiple for similar reasons.
3. Credit-Market Skepticism Bleeding into Equity Positioning
Weiss's refusal to own CoreWeave, Anthropic, or OpenAI because they lack an EBITDA multiple reflects a broader investor shift toward demanding visible cash generation before assigning AI-related valuations — a theme echoing concerns raised elsewhere about neocloud financing risk.
4. Streaming's Growth Narrative Unwinding
Netflix's sixth consecutive post-earnings drop signals the market no longer views the company as a premier growth story, with panelists arguing cheap valuation multiples are irrelevant without a return to subscriber and engagement growth. ---
Sentiment Analysis
Overall Market Sentiment: Divided
The committee was split nearly evenly between those expecting a near-term AI trade resumption tied to the midterms and those urging continued selectivity and skepticism about near-term catalysts.
Risk Factors Highlighted
AI trade concentration risk: Panelists noted the top 10 S&P names represent roughly $30 trillion in market cap, meaning earnings disappointments could have outsized index impact.
Valuation anchor gap: Companies like CoreWeave, Anthropic, and OpenAI lack EBITDA multiples, making traditional valuation discipline difficult to apply.
Anthropic's funding needs: Reported requirement of roughly $750 billion over coming years raises questions about capital market absorption.
Netflix engagement erosion: Wells Fargo warns declining engagement could cascade into subscriber cancellations.
Midterm election uncertainty: Positioning remains light ahead of the election, seen as a near-term source of volatility regardless of outcome.
CapEx cycle maturity: Weiss argues the capital expenditure cycle will eventually peak, compressing valuations for spending-dependent names.
Streaming sector-wide content cost pressure: Panelists flagged rising content costs and squeezed margins across Netflix, Disney, and peers.
This episode was covered in today's [The Market Signal — 2026-10-22](https://marketsignal.beehiiv.com/p/the-market-signal-2026-10-22), a cross-source synthesis of multiple podcast reports.