Stocks Rise as Nvidia Surges on Earnings Beat | The Close 2/25/2026
Speakers
Summary
- Nvidia's Dominance vs. Market Breadth: Nvidia (NVDA) delivered a significant earnings beat (73% revenue growth, 75% gross margins), confirming the AI hardware boom is intact. However, the market narrative is shifting from "AI as a bubble" to "AI as a disruptor," causing capital to rotate out of pure-play tech concentration and into beneficiaries like Industrials, Materials, and Utilities.
- The Software Crisis: A sharp divergence is emerging between Hardware (NVDA) and Software (CRM, SNOW). Multiple speakers (Khosla, Kass) warn that AI is deflationary for legacy SaaS seat-based models. While hardware booms, software faces existential credit risks and guidance misses (CRM, SNOW both down on earnings), prompting a flight to "Real Assets" and "Physical Economy" sectors.
- The "Real Economy" Pivot: Strategic Value Partners highlights a massive disconnect between tech valuations and real assets. The actionable trade is moving down the AI value chain into the physical infrastructure required to support it—specifically Power, Steel, and Construction—rather than the software layer that might be automated away.
Trade Ideas
NVDA reported $68B in revenue (beating estimates), 75% gross margins, and confirmed demand for H200 chips is outstripping supply. Despite fears of a "bubble" or digestion period, the hyperscalers (Microsoft, Meta, Google) are explicitly increasing Capex. NVDA remains the "only game in town" for the necessary hardware, and the licensing grant for China removes a lingering regulatory overhang. Long NVDA as the primary beneficiary of the confirmed Capex super-cycle. Supply chain constraints (CoWoS packaging, HBM memory) limiting ability to fulfill orders.
Investors are punishing "Mag 7" companies for high AI Capex if revenue doesn't immediately follow, while simultaneously realizing that *someone* has to build the physical infrastructure for AI. This sentiment shift drives a rotation away from concentrated tech exposure into the "beneficiaries" of the spend. If AI requires massive power and physical build-outs, the companies providing the raw materials, energy, and industrial machinery will capture the Capex dollars regardless of which AI model wins. Long the infrastructure layer (Industrials, Materials, Utilities) as a hedge against Tech concentration. A broader economic slowdown reducing demand for commodities and energy.
August argues default rates won't spike; Khosla argues "tail risk is fact" and software credit (a large chunk of private credit) is vulnerable. A significant portion of private credit and direct lending is exposed to software companies bought at 20x+ multiples. If AI disrupts those business models (as seen with CRM/SNOW), the credit backing them becomes toxic. Watch Private Credit closely; avoid exposure to funds heavily allocated to legacy software LBOs. If the "soft landing" persists, credit spreads may remain tight, punishing those who sit out.
CAVA posted 3-5% sales growth guidance (beating estimates) and stock rallied 25%. Despite macro headwinds, the "Mediterranean category" is proving resilient. The company is successfully executing a unit expansion strategy while maintaining traffic, proving it is capturing share from other fast-casual dining options in a K-shaped consumer environment. Long CAVA as a momentum leader in the consumer discretionary space. Wage inflation or a sharp pullback in consumer spending.
Kass argues that if value is extracted from the S&P 500 (specifically software), it must go somewhere else. AI reduces the cost of creating software, which is deflationary for the Tech sector but inflationary for sectors that rely on physical/biological reality. Capital will rotate into "Novel Sciences" (Biotech), "Building the Real World" (Construction), and "Human Connection" (Hospitality). Long Biotech / Construction / Hospitality as the counter-cyclical trade to AI software deflation. High interest rates hurting capital-intensive sectors like construction and biotech.
Salesforce (CRM) and Snowflake (SNOW) both issued disappointing guidance/results, causing stock drops. Victor Khosla notes that software companies leveraged at 40-50% LTV are facing "existential" risks from AI disruption. Second-order thinking suggests AI agents will reduce the need for "seats" in traditional SaaS models. If AI increases productivity, companies need fewer software licenses. This creates a "value transfer" away from legacy SaaS toward AI models and hardware, leaving legacy software with high debt and shrinking moats. Short Legacy SaaS / System of Record Companies (represented by CRM/SNOW) due to deflationary AI pressures. Valuation mean reversion if the market over-penalizes these stocks (as argued by Gail).
There is a massive divergence between "tech valuations" and "real economy business valuations." Khosla specifically cites Steel companies benefiting from AI scrap metal needs and power plants. AI is energy and material intensive. As data centers expand, the demand for high-quality scrap metal (for steel production) and independent power generation increases. These "Real Assets" are currently undervalued relative to the AI tech stack they support. Long Real Assets (specifically Steel and Power) as the physical derivative of the AI trade. Regulatory caps on power prices or a collapse in industrial manufacturing.
Up Next
This Bloomberg Markets video, published February 26, 2026, features Matt Bryson, Gabriela Santos, Victor Khosla vs. Glenn August, Brett Schulman, Zack Kass, Romaine Bostick, Victor Khosla discussing NVDA, XLI, XLB, BKLN, CAVA, XBI, PEJ, ITB, CRM, SNOW, STEEL. 7 trade ideas extracted by AI with direction and confidence scoring.
Speakers: Matt Bryson, Gabriela Santos, Victor Khosla vs. Glenn August, Brett Schulman, Zack Kass, Romaine Bostick, Victor Khosla · Tickers: NVDA, XLI, XLB, BKLN, CAVA, XBI, PEJ, ITB, CRM, SNOW, STEEL