CNBC Halftime Report

2026-07-10 · Hosted by Scott Wapner · CNBC

Executive Summary

The Halftime Report focused heavily on extreme volatility in the semiconductor/momentum trade, with Josh Brown noting the SMH has had almost no days with less than a 2% intraday move recently, driven partly by roughly $200 billion in leveraged single-stock ETFs (equivalent to $500 billion in notional exposure) distorting price action. Jim Lebenthal doubled his Micron position, framing it as both a value and momentum play given Micron's outsized contribution to S&P earnings growth. Stephanie Link revealed she bought Nvidia for the first time ever, calling it "best and breed on sale" at the cheapest valuation since 2019. The desk debated whether beaten-down hyperscalers (Meta, Microsoft down 25-32% from highs) represent a buying opportunity, with Josh Brown arguing Apple is best positioned to capture AI value without the CapEx risk. Citigroup was highlighted as a multi-year turnaround story under CEO Jane Fraser, up 65% over the past year. Anthropic announced former Fed Chair Ben Bernanke is joining its Long-Term Benefit Trust governance body.

Key Stories & Changes

1. Momentum/Chip Trade Extreme Volatility

  • Mike Santoli noted only a couple of days in the past month where semis moved less than 2% intraday

  • Josh Brown: ~$200 billion in leveraged single-stock 2x ETFs now mainstream, equivalent to $500 billion in notional exposure, exacerbating daily swings unrelated to fundamentals

  • SMH RSI reset to a "healthy" 49; average 10-day rate of change for SMH stocks hit negative 15% during the recent drawdown

  • Only ~17% of SMH stocks are at 50-day highs despite the index being near 52-week highs, reflecting significant internal churn

2. Jim Lebenthal Doubles Down on Micron

  • Bought more Micron at $60 as his final trade, selling September $70 calls for $7 premium, bringing cost basis to ~$53

  • Argues Micron should represent 15-20% of total S&P earnings growth this year; SK Hynix and Samsung together represent ~15% of EEM earnings growth

  • Frames Micron as simultaneously a value stock (fundamentals, contract visibility, upcoming capital return) and a momentum stock — says these styles can intersect

  • SK Hynix's US listing (next day) noted as ~7x oversubscribed

3. Stephanie Link Buys Nvidia for the First Time

  • Cites valuation: 18x P/E, 15x EBITDA — cheapest since 2019

  • Nvidia has underperformed the semiconductor group by 53% YTD and 72% over the past year

  • Cites 97% server GPU market share, proprietary software platform, new compute platforms, and expansion into CPUs

  • Revenue growth of 130% year-over-year last quarter, 85% gross margins, free cash flow expected to double to $206 billion next year

  • Josh Brown agreed, noting a "shareholder transition" — momentum crowd has left Nvidia for names like Micron, leaving a lower-risk, fundamentals-driven investor base

4. Hyperscaler Valuation Debate

  • Morgan Stanley Wealth Management: Mag 7 valuation premium over the other 493 S&P names is at just 10%, the lowest in a decade

  • Goldman's desk recommending buying the dip in hyperscalers

  • Josh Brown pushed back: markets historically don't award premiums to companies plowing 100%+ of cash flow into CapEx (comparing it to industrial conglomerate discounts); says this is a "building phase, not harvesting phase"

  • Meta down 25.5%, Microsoft down 32%, Amazon/Alphabet down ~13.5% from 52-week highs

  • BlackRock's Rick Reader (cited from a prior interview) said his firm has rotated within mega-caps toward names less directly tied to AI CapEx

5. Apple Framed as the AI Value-Capture Winner

  • Josh Brown's thesis: Apple's 2.5 billion device users make it the "gateway to consumer AI" without bearing CapEx risk — likened to owning "the other end" of the AI buildout rather than the infrastructure itself

  • Apple not spending at hyperscaler levels; stock near all-time highs, contrasted with Meta/Microsoft's steep CapEx-driven drawdowns

  • Upcoming catalysts: John Ternus taking over as CEO-designate role September 1, first foldable iPhone launch (~500K-1M units by Q4), iPhone 18 (20-22 million units), agentic Siri with LLM flexibility (ChatGPT, Claude compatibility)

  • $100 billion buyback program cited as ongoing support

6. Citigroup's Turnaround Under Jane Fraser

  • Citigroup up 65% over the past year (vs. PNC +33%, BofA +25%, JPMorgan +18%)

  • Josh Brown: 20-year annualized return still negative 4% (worst among large-cap banks), but international services described as the "crown jewel"

  • Best quarterly revenue in a decade last quarter; 56% jump in EPS year-over-year

  • Banamex spinout expected in 6-9 months; potential exit from Fed consent orders seen as a further catalyst

  • Jim Lebenthal: long-time holder, sees more room to run

7. Software/Cybersecurity Sector Divergence

  • Citi initiated coverage on software with top picks MongoDB, Snowflake, Palantir

  • Snowflake: EPS up 68%, revenue up 33% last quarter, new AWS deal; Jason Snipe holds it

  • CrowdStrike up 68% YTD; ARR grew to $5.51 billion (+24% YoY), free cash flow hit a record $468 million (34% of revenue); "Rule of 40" score of 59

  • Desk noted market is "stratifying" software names rather than universally selling off the sector, following the earlier "SaaS apocalypse" scare tied to AI disruption fears

8. Anthropic Adds Bernanke to Governance Board

  • Former Fed Chair Ben Bernanke appointed to Anthropic's Long-Term Benefit Trust, an independent body advising on AI's economic/societal risks and with authority to appoint board members

  • Bernanke quoted: "The potential of AI is enormous and so is the range of outcomes"

1. Leveraged ETF Flows Distorting Price Discovery

Josh Brown's data on $200 billion in leveraged single-stock ETFs (worth $500 billion notional) illustrates how derivatives-driven momentum trading — untethered to fundamental views — is amplifying daily volatility in popular AI names, a genuine market-structure shift investors must now factor into any entry/exit timing decision.

2. Value and Momentum Styles Converging in AI Names

Both Lebenthal (Micron) and Link (Nvidia) frame their purchases as simultaneously value-driven and momentum-consistent, reflecting how depressed valuations in previously high-flying AI infrastructure names are now attracting traditionally distinct investor bases into the same trades.

3. The "Who Captures AI Value" Debate Sharpening

The panel's extended Apple-vs-hyperscaler discussion (with explicit historical analogies to telecom infrastructure builders not capturing internet-era profits) reflects a maturing market debate: infrastructure spenders (Meta, Microsoft) face CapEx skepticism while device/platform owners (Apple) are rewarded for capturing AI value without bearing its cost.

4. Financials' Quiet Multi-Year Re-Rating

Citigroup's Fraser-led turnaround and broader bank sector strength ahead of earnings season reflect a longer-running rotation into previously unloved value sectors, with catalysts (Banamex spinout, regulatory relief) still ahead rather than behind. ---

Sentiment Analysis

Overall Market Sentiment: Constructively Bullish Amid Volatility

Panelists broadly added exposure to AI-adjacent names despite acknowledging elevated volatility, framing recent weakness as opportunity rather than a structural top.

Risk Factors Highlighted

Leveraged ETF-driven volatility: $500 billion in notional exposure can exaggerate normal selling/buying into extreme swings.

Hyperscaler CapEx sustainability: Markets continue to discount Meta/Microsoft pending clearer ROI signals this earnings season.

"Penalty box" dynamic for Nvidia/Broadcom: Panelists debate whether a ~$5 trillion market cap represents a ceiling the market won't extend past.

Cannot call a bottom: Both Lebenthal and Brown explicitly warn viewers against trying to time entries precisely in this environment.

Style drift risk: Lebenthal cautions investors against abandoning their investing discipline (value vs. momentum vs. technical) mid-rotation.

AI CapEx ROI uncertainty broadly: Central unresolved question of where profits ultimately accrue in the AI buildout (semis vs. hyperscalers vs. device makers).

Software disruption risk (partial): Even resilient names like Snowflake/CrowdStrike face lingering "AI replaces SaaS" overhang, though panel sees this as overstated for now.

This episode was covered in today's [The Market Signal — 2026-07-10](https://marketsignal.beehiiv.com/p/the-market-signal-2026-07-10), a cross-source synthesis of multiple podcast reports.

Keep Reading