CNBC Halftime Report
2026-10-27 · Hosted by Scott Wapner · CNBC
Executive Summary
The Investment Committee kicked off a pivotal week — mega-cap tech earnings, a Fed decision, and Anthropic's S-1 filing — with the market green across the board. Joe Terranova argued that strong price reactions have lagged strong earnings all season, warning that "stocks peak before earnings peak" and that investors must now be highly selective within the AI trade rather than buying it indiscriminately. Steve Weiss remained the most cautious voice, citing Middle East oil risk and tightening financing for neoclouds, while Jim Lebenthal and Stephanie Link argued the market is underappreciating how three major funding risks — Anthropic's IPO, SpaceX's $40 billion debt issuance, and Meta's $95 billion securities offering — were answered positively today, removing a key overhang.
Key Stories & Changes
1. Anthropic S-1: "Better First Step" Than Feared
Anthropic plans to list on NASDAQ under ticker ANT; plans to plow most IPO proceeds into CapEx (compute, infrastructure)
H1 2026 revenue of $16.3 billion, up more than 1,000% year-over-year; margins above 80%
Operating loss narrowed to just below $700 million in H1, down significantly from the prior year; profitable on an adjusted basis
Total net losses of $81 billion for H1, but the vast majority attributable to convertible-debt accounting
Founders retain voting control via a Class F share (seven co-founders, 50.1% of the vote); structured as a public benefit corporation with a dual mandate
Risk factors (nearly 100 pages) cover political backlash, self-preserving/coercive model behavior (resisting shutdown, blackmail-like conduct), compute/debt access, and customer concentration (two customers ~25% of last year's revenue)
Anthropic has committed at least $270 billion in compute spend to Amazon, Google, and Microsoft over coming years
Rohit Kulkarni: three factors drove the positive stock reaction — profitability trend line, over $100 billion in cash, and a disclosed enterprise footprint of 12,000 customers paying $100,000+ annually
Stephanie Link: "This is once in a generation, if not once in a lifetime" opportunity, calling Anthropic "clearly the leader in monetizing intelligence" over the last 24 months
2. Three Financing Risks Answered Positively Today
Anthropic's S-1, SpaceX's $40 billion debt issuance, and Meta's ~$95 billion securities offering all landed the same morning
Jim Lebenthal argued this answers "will the funding be available?" — a real risk alongside Middle East tensions and inflation — and said the "return on investment" debate is "a 2028 discussion, not a 2026 discussion"
Lebenthal: "I feel equally certain that the markets will rally to new highs by year end"
3. Market Internals Diverge From Headline Index Levels
The S&P 500 is only about 2% off its all-time high, but the average stock and equal-weight index show much weaker technicals, trading near 52-week lows in multiple groups
Steve Weiss argued this reflects real underlying weakness, not complacency; Joe Terranova countered that leadership is beginning to rotate from defensives (healthcare, staples, utilities) toward offense (financials, industrials)
Terranova: "Stocks peak a year before [CapEx] spending peaks" — a warning that some AI infrastructure investments may already be ahead of the cycle's turn
4. Hyperscaler Earnings Week: Alphabet, Microsoft, Amazon
Alphabet and Microsoft report Wednesday (same day as the Fed decision); Amazon reports Thursday
Jim Lebenthal likes Alphabet most among megacaps, calling the stock (~$315) "extremely cheap" given growth, AI deployment, and cloud; also likes Meta on weakness
DA Davidson note cited: Microsoft, Amazon, and Google will succeed "whether AI slows down or speeds up" given attractive relative valuations
Stephanie Link: watching for roughly $2.3 trillion in aggregate cloud backlog across hyperscalers as a positive signal, and emphasized strong interest-coverage ratios mitigate debt-sustainability concerns
Panel agreed that CapEx guidance, not just revenue growth, will determine whether the market rewards or punishes results this week
5. Microsoft's 60,000-Employee Buyout
Coincides with the week's earnings and is viewed as reinforcing the CapEx-versus-headcount tradeoff already in focus
6. Apollo Takes ~10% Stake in Mubadala Capital
Apollo (over $1 trillion total AUM) takes a minority stake, reported near 10%, in Mubadala Capital, the asset management arm of Abu Dhabi's sovereign wealth fund ($60 billion in core alternative assets)
Framed as a scale-driven deal reflecting consolidation in asset management and growing Gulf-region investment in AI data center infrastructure and energy
CNBC's Robert Frank: "Capital as a moat is a big thing" in an AI-driven investment landscape requiring ever-larger checks
7. Stock-Specific Moves: EQT/Range Resources, Amgen, Lockheed Martin
EQT: EQT Corp (buying Range Resources) — Deal announced — $13 billion all-stock deal; integration targets Appalachia gas supply for Northern Virginia data center demand
AMGN: Amgen — Featured on positive data — Drug combo cuts death risk by 58% in aggressive lung cancer patients; reports November 3
LMT: Lockheed Martin — Price target raised — Morgan Stanley upgrades to overweight, target to $835 from $690; strong order book (850 F-35s over 10 years)
Trends Identified
1. AI Trade Requires Increasing Selectivity, Not Broad Exposure
Joe Terranova's framing — "before, it was buy the AI trade or don't; now it's buy the AI trade, but you better be selective" — captured a consensus shift. Panelists differentiated sharply between megacap hyperscalers (viewed as attractively valued) and more speculative names like AMD and Broadcom (viewed as harder to justify on earnings).
2. Financing Risk Resolution Is Driving Today's Rally, Not AI Fundamentals
Jim Lebenthal's thesis — that the market rallied because a key funding-availability risk was resolved via Anthropic, SpaceX, and Meta's capital raises on the same day — reframes the AI trade's near-term catalyst as a financing story rather than a pure fundamentals story.
3. Market Breadth Weakness Beneath Resilient Headline Indices
Steve Weiss's focus on equal-weight and average-stock technicals underscores a recurring 2026 theme: narrow index-level strength masking broader technical deterioration, even as bulls like Terranova and Lebenthal see early signs of leadership rotation toward cyclicals.
4. Gulf Capital Is Increasingly Investing Directly, Not Just Funding Others
The Apollo-Mubadala deal reflects a broader shift of Gulf sovereign wealth from passive capital providers to direct co-investors in AI infrastructure, data centers, and energy — with "capital as a moat" becoming a defining theme of the AI buildout's financing phase.
5. Compute Efficiency Metrics Are Validating Continued AI CapEx
As in other episodes this week, Anthropic's falling compute-cost-to-revenue ratio (10x revenue in 2023 to 55% of revenue in H1 2026) is being used by bulls to justify continued hyperscaler spending commitments ahead of this week's earnings. ---
Sentiment Analysis
Overall Market Sentiment: Divided, Leaning Constructive
The committee was split along familiar lines — Terranova and Lebenthal constructive on a year-end rally, Weiss cautious on Middle East and financing risk — but broadly agreed Anthropic's filing was a net positive surprise.
Risk Factors Highlighted
Middle East escalation: Steve Weiss's central risk — an oil price spike toward $150 would be "a disaster" for markets if diplomatic efforts fail.
Neocloud financing stress: Tightening debt and equity markets could force weaker neocloud operators out of business, informing valuations across the AI infrastructure group.
Anthropic revenue deceleration: Growth is expected to continue slowing from historical triple-digit rates; October run-rate data is the key swing factor investors await.
Market breadth deterioration: Equal-weight and average-stock technicals remain weak even as headline indices sit near highs.
CapEx reward/punishment uncertainty: Hyperscaler earnings this week could see the market penalize, not reward, continued spending increases.
Existential/model-behavior risk disclosures: Anthropic's S-1 explicitly flags catastrophic risk scenarios tied to advanced AI model behavior.
Customer and compute concentration: Anthropic's reliance on Amazon, Google, and Microsoft for $270 billion in compute commitments creates potential conflicts of interest and dependency risk.
Open-source/Chinese model pricing pressure: Could compress margins across frontier AI labs, including Anthropic.
Governance complexity: Seven founders sharing 50.1% voting control introduces investor uncertainty distinct from single-founder structures.
This episode was covered in today's [The Market Signal — 2026-10-27](https://marketsignal.beehiiv.com/p/the-market-signal-2026-10-27), a cross-source synthesis of multiple podcast reports.