Halftime Report: Reacting to Nvidia’s Earnings – Brad Gerstner Joins Live

Date: August 28, 2025

Host: Scott Wapner (CNBC)
Featured Guest: Brad Gerstner (Altimeter Capital)
Panel: Josh Brown, Joe Terranova, Surat Seti, Bryn Talkington, Silvana Hanale, Steve Liesman

Episode Overview

This market-moving episode of the Halftime Report centers on the aftermath of Nvidia’s highly anticipated earnings report. With stocks at record highs and Nvidia leading the AI revolution, the panel—joined exclusively by Brad Gerstner, CEO of Altimeter Capital and major Nvidia shareholder—offers real-time analysis on:

• Nvidia’s results, guidance, and stock action
• Broader AI/tech sector trends, including concerns of a bubble
• The implication of customer concentration and market risks
• The government’s role in AI/chip export policy, especially regarding China
• The impact of the AI boom on the software sector Additional discussions cover CrowdStrike’s earnings, Tesla’s current status, vaccine-maker woes, Federal Reserve drama, and select stock picks.

Key Themes & Insights

1. Nvidia Earnings: “A Banger of a Report”

Brad Gerstner’s Take [31:13–34:22]

• Nvidia delivered a "banger" quarter: 56% top-line growth, accelerating at unprecedented scale for the largest company in the world.
• Margins and guidance strong: Profitability at 73%. Q3 guidance was to $54 billion excluding China.
• Nvidia as a national resource: "It's the most important company in the world." – Brad Gerstner [31:59].
• The market is "massively supply constrained" for compute and AI chips.
• The long-term infrastructure spend estimate increased sharply, from $2T to $3–$4T by decade’s end.
• Nvidia’s stock action post-earnings was muted, but this is against the backdrop of an extraordinary run-up from earlier in the year.

"Imagine doing all of this... the largest scaled buildout of compute in history, despite all the wrangling on tariffs in China, despite the transition and ramping up of a new generation of chips in Blackwell that a lot of people thought they couldn’t pull off without some major hiccups." — Brad Gerstner [32:45]

2. Market Read-Through: Confidence Beyond Nvidia

Panel Discussion [02:09–08:27]

• Muted price action: Expected after the stock’s significant run-up into earnings.
• Nvidia’s customer concentration flagged: 3 now represent 56% of A/R; previously 2 were 33% [05:07].
• Strong buyback wave: $1T in S&P buybacks already in 2025, led by mega-caps including Nvidia [06:16].
• AI infrastructure spend: $600B annually is more than doubling year-over-year.
• Latest quarter may "offer some reassurance for investors after signs of stalling" in large-cap tech [06:45].

On Customer Concentration

“There are 7 billion Internet users. All of them are going to be engaging with AI... So the hyperscalers are basically a conduit to get Nvidia technology in front of 7 billion Internet users... So it’s a stupid conversation and I don’t think concentration risk here is the issue." — Josh Brown [08:27]

On Market Risks

Nvidia and its hyperscaler customers cannot defy gravity in a real economic downturn; broader macro risks remain even for AI leaders.

3. AI, China, and U.S. Policy

Brad Gerstner’s View [34:22–40:11]

• On U.S. export policies to China:
  • The U.S. must let Nvidia and other AI firms compete in China—or risk ceding the ecosystem to Huawei.
  • U.S. policy has been slow but is now accelerating under new executive orders.
  • China represents a $50B opportunity for Nvidia, currently excluded from guidance.
  • Gerstner believes there’s a “>50%” probability that Nvidia can sell Blackwell chips (or analogous models) in China.

“If we cede China to Huawei, it makes it vastly easier... for Huawei and the very sophisticated model ecosystem they have... to win around the world.” — Brad Gerstner [36:25]

• On potential government cut of sales:
  • Gerstner doesn’t love the idea (as a shareholder) but sees the merit in the government hedging risks by taking a cut to fund domestic chip initiatives.

4. Is There an AI Bubble?

Panel + Gerstner [42:49–44:04]

• Sam Altman (OpenAI CEO) recently warned of “AI investor overexcitement.”
• Gerstner agrees: Short-term hype is likely; long-term, impact remains underestimated.
• Using Google’s post-IPO run as an analog, investors who bought at “overexcited” short-term highs did exceedingly well if they held long-term.

"My view is... Of course, we have excess exuberance and excitement in the very short run for these companies, but I think when you look at the medium to the long term, AI is still underappreciated for the impact that it’s going to have."
— Brad Gerstner [43:22]

5. AI’s Impact on the Software Sector

Gerstner on Differentiating Winners [52:39–55:34]

• “AI is not killing software. It’s dividing the winners from the laggards.”
• Companies like Snowflake, Databricks, and MongoDB are thriving, because “you can’t run AI without data.”
• The right question: Does AI create tailwinds for data infrastructure? Yes, and it’s driving durable acceleration in these businesses.
• Some application software companies (e.g., Salesforce, Adobe) are struggling, but the “AI winners in software” are being clearly delineated.

6. Other Notable Market Segments and Stocks

CrowdStrike and Cybersecurity [11:01–14:41]

• Despite strong performance (“beat on top/bottom line”), CrowdStrike and the broader cybersecurity space have lagged—reflecting sector rotation and overhangs from earlier negative earnings (notably Fortinet).
• Current muted reaction offers a long-term opportunity, per panelists.

Tesla’s Quiet Recovery [15:41–17:45]

• Tesla up 11% in the month, possibly because Elon Musk has been less publicly political.
• Fundamentals still challenged ("lower revenues, lower earnings"), leading to range-bound trading. Suggestion: Covered call strategy makes more sense than outright ownership.

Vaccine Makers: “Untouchable” or Opportunity? [17:45–19:57]

• With political headwinds and negative sentiment, vaccine units at companies like Pfizer and Moderna are “not investable” short-term, though broader pharma businesses may remain so.

Federal Reserve Drama: Lisa Cook’s Dismissal [19:12–26:58]

• Breaking news segment featuring Steve Liesman—Federal Reserve controversy, legal/ethical/political issues around Lisa Cook’s dismissal.

7. Apple and the “Build vs. Buy” AI Debate

Josh Brown & Brad Gerstner [49:04–52:39]

• Apple rumored to have explored acquiring Perplexity for its LLM/AI efforts—panel discusses whether Apple should buy or partner.
• Gerstner’s advice: Focus on partnerships for lightweight edge models (OpenAI, Gemini), not major acquisitions.
• Improved Siri (via small AI models) could swiftly make up for Apple’s current lag in AI voice assistants.

8. Palantir’s Resilience and Missed Opportunity

Joe Terranova & Brad Gerstner [55:34–57:00]

• Gerstner admits skipping Palantir was a “big mistake”—Consensus was to view them as more services than software, but forward-deployed AI engineering now viewed as essential.
• Will consider owning at the right valuation.

Notable Quotes & Timestamps

• “It’s the most important company in the world. Margins at 73%. They just guided to $54 billion ex-China… accelerating at scale, the largest scaled build out of compute in history.”
  – Brad Gerstner on Nvidia [31:59–32:30]

• "If you’re focused on [the $3–$4T AI infrastructure spend] you’re not worried so much about how the stock traded five minutes after the report.”
  – Josh Brown [02:09]

• “So the hyperscalers are basically a conduit to get Nvidia technology in front of 7 billion Internet users... The true risk here is the same risk you face in any other stocks, which is economic risk.”
  – Josh Brown [08:27]

• “We want American AI companies to win around the world... you can’t win around the world... if you cede China to Huawei.”
  – Brad Gerstner [36:12]

• “Of course, we have excess exuberance and excitement in the very short-run... but when you look at the medium to long-term, AI is still underappreciated for the impact that it’s going to have.”
  – Brad Gerstner [43:20]

• “You can’t run AI without data. Data is the oil that runs the AI engine... these companies’ acceleration is durable.”
  – Brad Gerstner on Snowflake, Databricks, MongoDB [53:47]

Timestamps for Key Segments

• [01:16] Opening/Market context & Nvidia setup
• [02:09] Josh Brown first reaction to Nvidia earnings
• [03:13] Panel on price action, Street analyst reactions
• [05:07] Customer concentration at Nvidia
• [06:13] Buyback boom discussion
• [08:27] AI sector and risk perspective
• [11:01] CrowdStrike earnings analysis
• [15:41] Tesla update
• [17:45] Vaccine makers’ dilemma
• [19:12] Federal Reserve / Lisa Cook breaking news
• [31:13] Brad Gerstner exclusive: deep dive on Nvidia
• [34:22] US-China AI policy, chip exports
• [42:49] Discussion of AI bubble concerns
• [46:19] Meta/Facebook’s place in AI and AI hiring headlines
• [49:04] Apple’s AI strategy and LLMs: Build vs. Buy
• [52:39] AI’s impact on software sector winners/losers
• [55:34] Palantir: VC misses and current positioning
• [57:44] Final trades: Stock picks from the panel

Conclusion

This episode delivers a high-density, insider’s perspective on Nvidia’s transformative role in AI, the sustainability of the AI rally, policy risks/opportunities in China, and the deepening divergence among tech/software names in the new AI-centric era. As always, the market narrative is framed with a view toward actionable strategy for investors.

For a full breakdown of rapid-fire final trades, breaking news (including the Federal Reserve saga), and more, listen to the full episode.