CNBC Closing Bell
2026-10-22 · Hosted by Scott Wapner, Melissa Lee, Michael Santoli · CNBC
Executive Summary
Stocks rallied into the close with the Dow up 350 points and the Nasdaq up more than 1%, lifted by strong SpaceX and Super Micro commentary that reinforced AI demand confidence. Meta reported record revenue of $64.6 billion, up 26% year-over-year and ahead of the $63.4 billion estimate, while raising the low end of its CapEx (capital spending on AI infrastructure) range to $135-145 billion. Tesla and IBM both beat on top and bottom lines, with Tesla posting $0.48 EPS versus $0.43 expected and IBM beating with $2.95 adjusted EPS.
Key Stories & Changes
1. Meta's Record Quarter and Muse Monetization Debate
Revenue: $64.6 billion, up 26% year-over-year, ahead of the $63.4 billion estimate; Meta's highest-ever quarterly revenue
Q4 revenue guidance: $73-76 billion, midpoint above consensus
CapEx: raised low end to $135-145 billion (from $130 billion low end), a $5 billion increase
GAAP EPS: $3.83, impacted by a nearly $10 billion settlement charge; adjusted EPS excluding the charge: $7.40
New non-voting Class C common stock announced, effectively a two-for-one split, trading beginning late January
Company will not provide full-year 2027 revenue guidance
On the call, Zuckerberg said Muse has more than 10 million users, with 25% using it daily and average spending over $200 per user; said Muse will be "highly profitable" if compute efficiency improves
Jeffries' Brent Thill: "pretty stable, steady... no fireworks" on CapEx, but wants more color on Muse monetization and the new share class's purpose
2. Tesla Beats on Top and Bottom Lines Ahead of Merger Talk
TSLA: Tesla — Mixed reaction — EPS $0.48 vs. $0.43 expected; revenue $28.05B vs. $27.95B expected; gross auto margin 18.1% vs. 16.8% expected; FSD subscribers up 67% YoY to 1.74 million
META: Meta Platforms — +5% after hours — Record $64.6B revenue beat; raised CapEx low end; Muse engagement highlighted on call
IBM: IBM — +4% — Beat on revenue ($17.1B vs. $16.82B) and EPS ($2.95 vs. $2.86); software/infrastructure both beat
T: AT&T — Worst S&P performer — Strong results overshadowed by Starlink competitive fears and lower phone subscribers
CMG: Chipotle — Hit 3-year lows — Downgraded at TD Cowan, two price-target cuts
LVS: Las Vegas Sands — -4.5% — EPS and revenue missed; Macau and Singapore properties both weak, though dividend raised
3. IBM's Software Rebound and M&A Commentary
CEO Arvind Krishna told CNBC he wants to "give numbers on which we have extremely high confidence" after last quarter's guidance miss, citing Red Hat returning to double-digit growth after a prior quarter at minus 7%
IBM closed two-thirds of previously delayed deals; full-year revenue guidance maintained at 4-5% growth
Krishna said private equity is "not yet willing" to meet valuation expectations for M&A targets, predicting that could change in late 2026/early 2027
Krishna said IBM's quantum computing capability is not priced into the stock
4. Semiconductor Equipment and Industrial Earnings
Lamb Research: beat by 2 cents at $2.20 adjusted; revenue $8.22B vs. $8.15B estimate; stock fell on guidance seen as underwhelming; 35% of revenue derived from China
Texas Instruments: double beat, $2.66 adjusted vs. $2.42 expected; Q4 guidance of $5.8B vs. $5.7B street estimate; revenue up ~28% year-over-year on industrial, automotive, data center strength
5. Macro: Credit Conditions as the Real Constraint on AI CapEx
Evercore ISI's Julian Emanuel argued credit default swap spreads, not equity valuations, are the key signal: rising hyperscaler yields over recent weeks reflect markets pricing in some capital discipline, not unlimited CapEx growth
Emanuel expects resolution higher into 2027 but is sticking with a flattish S&P call through year-end given midterm election uncertainty
Treasury counselor David Zervos (speaking in a personal capacity) attributed the recent pullback in 10-year yields partly to reduced hyperscaler debt issuance and said he sees "no systemic risk" in the AI buildout, calling it a story of companies putting their balance sheets on the line for compute
Trends Identified
1. Hyperscaler Free Cash Flow Compression Becoming the New Normal
Analysts on the show increasingly treat negative free cash flow (cash spent exceeding cash generated) from AI CapEx as an accepted cost of admission rather than a red flag, provided the underlying business — like Meta's advertising — remains strong. Brent Thill explicitly framed this as the trade-off investors must accept while AI investment outpaces near-term ROI (return on investment).
2. Equity and Debt Issuance to Fund the AI Buildout
Meta's new non-voting share class, following Alphabet's prior $85 billion equity raise, signals a broader shift toward diversified capital-raising — combining debt and equity — to fund an AI CapEx cycle that analysts expect to go "materially higher" into 2027. Thill noted this dynamic will likely recur across Microsoft, Amazon, and other hyperscalers.
3. Credit Markets, Not Valuations, as the Arbiter of the AI Trade
Multiple guests pointed to credit spreads and treasury yields, rather than equity multiples, as the key signal for whether AI spending can continue unchecked. Both Emanuel and Zervos linked recent yield moves to reduced hyperscaler debt issuance, suggesting the bond market is quietly setting the pace for the broader AI narrative.
4. Energy Prices as the Hidden Driver of the Inflation and Rate Story
Zervos repeatedly traced recent inflation stickiness back to an oil price shock (WTI up nearly 80%), arguing that once geopolitical conflicts resolve and energy prices retreat toward $60, much of the current rate anxiety could fade. ---
Sentiment Analysis
Overall Market Sentiment: Cautiously Bullish
The session reflected relief rather than euphoria — strong earnings reinforced confidence in the AI trade without resolving lingering questions about monetization and spending sustainability.
Risk Factors Highlighted
Hyperscaler cash flow drag: Heavy AI CapEx is pushing companies like Meta toward negative free cash flow for the first time in years.
Credit market deterioration: Rising credit default swap spreads above March conflict-era peaks would signal markets are losing confidence in the AI spending cycle.
Midterm election uncertainty: Guests flagged political volatility and potential House/Senate shifts as a source of market sensitivity into November.
Energy price shock persistence: Sustained high oil prices (WTI potentially above $100 in 2027) could trigger broader economic slowdown.
AT&T/Starlink competition: Satellite internet competition emerged as a tangible threat to legacy telecom subscriber growth.
IBM legacy revenue risk: Analyst Brent Thill flagged concern that mainframe deal strength may not persist as AI reshapes enterprise computing priorities.
Unclear AI monetization timeline: Meta's refusal to give 2027 revenue guidance leaves investors without clarity on CapEx payback timing.
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.