CNBC Halftime Report
2026-06-03 · Hosted by Scott Wapner · CNBC
Executive Summary
The Halftime Report was dominated by a live interview segment with Goldman Sachs CEO David Solomon at the Economic Club of New York, where he acknowledged being in a period of “more greed than fear” while expressing genuine uncertainty about the AI demand curve. The panel also debated whether the market’s AI momentum trade — momentum factor up ~35% quarter-to-date — is sustainable or signals an impending correction. Marvell’s +32% surge after Jensen Wong’s trillion-dollar endorsement anchored the AI infrastructure discussion, while committee members made notable portfolio moves including selling Netflix and Uber on time stops, and adding Twilio, Generac, American Airlines, Hyatt, and Dick’s Sporting Goods to their portfolios.
Key Stories & Changes
1. David Solomon (Goldman Sachs) — “More Greed Than Fear”
Solomon, speaking at the Economic Club of New York, called the current environment one where there is “definitely more greed than fear”
On AI: “extraordinarily excited about the opportunity” for productivity growth; believes AI is a “generational technological shift”
On the demand curve: “enterprises will go slower” than current expectations; low-margin businesses will be very cautious about token spend
Cited Walmart’s token spend surge as evidence that costs are “super expensive” for enterprises
Warned that AI compute demand will not be a straight line — drew parallel to late 1990s undersea cable build-out that ultimately gave way to wireless, leaving initial investors with losses
On labor: dismissed “massive structural unemployment” thesis but acknowledged AI will “interrupt jobs and dislocate”; called for upskilling and government/business policy response
Acknowledged Alan Greenspan’s “irrational exuberance” speech came 3.5 years before the Nasdaq peaked at 5200 — implying current exuberance could persist before an eventual correction
On equity supply (Alphabet, SpaceX, etc.): “there’s enough capital” in $150T global equity markets; Alphabet’s $80B deal trading “very well” as data point; but acknowledged deals are “unprecedented in size”
On timing: “something could happen that changes the lens… quite quickly” — greed can turn to fear fast
2. Marvell — Jensen Wong’s Trillion-Dollar Endorsement
Marvell up ~32% — driven by Jensen Wong’s statement at Computex that it could be the next $1 trillion company
Stephanie Link: optical networking running on inference, growing 50-55%; custom ASICs (80% market share with Broadcom) growing 20%, expected to reach 100% by FY29
Earnings power revised sharply: “a month ago I thought maybe $5 [EPS], now maybe $10”
Joe Teranova: optical thesis is similar to memory earlier in cycle; people witnessing “what went on with memory” and now seeing optical as the next bottleneck
Stock up 232% — could go to 275%+; Nvidia has a $2 billion investment in Marvell
3. Market Momentum Debate
S&P 500 up 19% from March 30 lows; NASDAQ up 31% in same period
Momentum factor (MTMU/SPMO): up >35% quarter-to-date and “more than doubling the market” since Iran war began
High-beta, high-momentum stocks at 9-week gain of over 40% into new highs — only prior precedents: November/December 1999 and January 2021 (both close to final tops of those rallies)
Rick Reader (BlackRock CIO, Fixed Income): worried about “crowding in different markets” but acknowledged multiples “not that scary” — manageable if earnings growth continues
Wells Fargo: “later innings of a sugar high rally”; sell signal from sentiment indicator triggered in May for first time since November 2021; put/call ratio at 2 standard deviations
Brian Belski: fundamentally very bullish; “25-year secular bull market remains in place” but needs a 5-10% normal correction to “clear the froth”; won’t time the top
4. Committee Trades
Joe Teranova sold: Netflix (time-stopped after 30 days down from purchase); Uber (same time-stop discipline)
Joe Teranova bought: Twilio (Bank of America momentum list; agentic AI exposure via messaging/voice); Generac (AI data center backup power; hit 4-year high; signed hyperscaler data center agreement)
Brian Belski bought: Dick’s Sporting Goods and Academy Sports and Outdoors (consumer leisure; smaller cap plays on broadening out trade); American Airlines and Hyatt Hotels (travel; moved down-cap from United/Hilton to smaller names to increase portfolio name count and reduce average market cap)
Stephanie Link: owns Dick’s Sporting Goods; noted 0.6% same-store comp in Footlocker integration; “Fast Break Store” concept driving double-digit growth; 15x forward — undervalued
5. Goldman Sachs / AI Enterprise Adoption
Solomon: $800B hyperscaler CapEx this year; even if flat next year (not growing to $1.6T), “you still don’t have enough data centers”
Only 11,400 data centers exist globally; analysts estimate 30,000 needed; each gigawatt facility costs $45 billion to build out
Grid constraints: “can’t increase the cost of power… to average Americans — that’s not going to work”; grid upgrade is a binding constraint
Consumer adoption “a much more complicated model than the enterprise model” — still uncertain
6. Crypto Options Signal — Bitcoin Bears Dominant
Ollie Renick (CBOE, Chicago): Bitcoin relative strength vs. stocks peaked last July; spread now >70 percentage points underperformance
In Bitcoin ETF IBIT: more puts than calls; more put buying; more call selling
In MSTR: 2x more puts bought than calls; most popular contract: 130 strike put expiring Friday (~6% downside to break even)
In Coinbase: calls popular but 3x more sold than bought
Stephanie Link: owns Coinbase for long-term; “set it and forget it” — small position; notes exchanges benefit from any buyer-seller pair regardless of direction
Trends Identified
1. Market Structure Is Increasingly Algorithm-Driven
Joe Teranova offered a structural explanation for why momentum keeps working: quantitative funds now dominate market structure, and algorithms systematically chase both fundamental (earnings growth) and technical (price appreciation) signals. This creates a self-reinforcing cycle in AI names where strong earnings → price momentum → algo buying → more momentum → further price appreciation. The momentum factor being up 35%+ quarter-to-date is not purely fundamental; it reflects the growing share of non-discretionary capital systematically amplifying existing trends.
2. Goldman Sachs CEO Is the Voice of Calibrated Uncertainty
David Solomon’s interview was notable for its intellectual honesty: he declined to call a top while explicitly acknowledging the greedy psychology, drew the Greenspan parallel, and flagged that enterprise AI adoption will be slower and more cautious than the infrastructure build-out implies. His point about low-margin businesses being unwilling to pay expensive token costs provides a credible demand-side constraint that may limit how quickly cloud AI revenue translates to justification for hyperscaler CapEx.
3. “Time Stops” Are a Discipline Investors Need to Apply Now
Joe Teranova’s explicit articulation of time-stops — not just price-stops — for lagging positions like Netflix and Uber provided a disciplined framework for navigating a momentum-driven market. In environments where capital relentlessly rotates to the strongest performers (Marvell +32%, Generac hitting 4-year highs), holding sideways-to-down stocks because of fundamental conviction is an opportunity cost trap. The committee’s rotation toward momentum names with fundamental backing (Twilio, Generac) while exiting time-stopped positions reflects rational adaptation to current market structure.
4. Travel and Leisure Broadening Is a Conviction Theme
Brian Belski’s switch from large-cap (Hilton, United) to mid-cap (Hyatt, American Airlines) while keeping the same sector exposure represents a broadening-out thesis: these companies trade at significant discounts to their large-cap peers, have comparable earnings quality, and benefit from the same travel demand trends. Joe Teranova’s Hyatt observation — 52-week high with only 54% analyst buy ratings — captures the contrarian signal in under-owned quality names. —-
Sentiment Analysis
Overall Market Sentiment: Euphoric But Self-Aware
The panel acknowledged momentum extremes and Greenspan-like warning signals, yet maintained bullish stances because fundamentals still support earnings growth — a rational but uncomfortable position that reflects end-of-cycle psychology without a confirmed catalyst for reversal.
Risk Factors Highlighted
9-week 40%+ momentum precedent: Only two prior instances (November/December 1999; January 2021) both occurred close to final tops; current setup is statistically rare and historically dangerous
Greed-fear imbalance: Goldman CEO explicitly calls market “more greed than fear”; Wells Fargo sell signal triggered in May for first time since November 2021
Enterprise AI demand curve is non-linear: Solomon’s warning about slow enterprise adoption, high token costs, and the non-linear nature of tech adoption (wireless disrupted cable) challenges the straight-line CapEx forecast
Grid constraint is binding: Only 11,400 data centers exist vs. estimated 30,000 needed; electricity grid upgrades are slow; cost of power cannot be passed to consumers indefinitely
Momentum factor reversal risk: As market becomes more quant-driven, momentum reversals tend to be faster and more violent; algorithms that buy momentum also create mechanical selling when it turns
AI CapEx ROI still unproven at enterprise scale: The $800B+ CapEx build is premised on future monetization; if enterprise adoption is slower than modeled, hyperscaler revenue growth will disappoint
This episode was covered in today’s The Market Signal — 2026-06-03, a cross-source synthesis of multiple podcast reports.