Momentum Crashes: Why Nvidia Is Falling, Google Is Surging, and VLMs Change Everything
Momentum Crashes: Why Nvidia Is Falling, Google Is Surging, and VLMs Change Everything
YouTube28 min 32 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights

The primary investment theme is the massive AI infrastructure build-out driven by the shift to Visual Language Models (VLMs) over the next three years. Consider Google (GOOGL) a key beneficiary, as its Gemini model and custom TPU chips provide a competitive edge in this next phase of AI. To capitalize on the application of this technology, look to the biotech sector (XBI), which is using AI to accelerate drug discovery and is showing significant market strength. With the market rally broadening beyond mega-caps, investors should also consider increasing exposure to small-cap stocks (IWM). Finally, remain bullish on the broader semiconductor sector while being cautious on traditional software stocks, which face disruption from AI-driven automation.

Detailed Analysis

Artificial Intelligence (AI) Sector

  • The speaker is extremely bullish on the AI sector, stating it is "not a bubble" and is actually "under hyped." They compare fighting the AI trend to fighting Quantitative Easing (QE), suggesting it's a losing battle for investors.
  • A major catalyst is the U.S. government's "Genesis Mission," an executive order aimed at cementing U.S. dominance in AI. This initiative is described as a "Manhattan project for AI" and creates a "too big to fail" status for leading AI companies by weaving them into national infrastructure.
  • The key theme is a major technological shift from Large Language Models (LLMs), which are text-based, to Visual Language Models (VLMs) and Vision Language Action (VLA) models.
    • VLMs train on images, video, and spatial data, not just text. They are the bridge to real-world autonomy.
    • VLAs are the "brains inside the machines" that will enable robotics and real-time actions.
  • This shift to VLMs and VLAs is expected to cause compute requirements to "explode" by 5 to 20 times, creating a massive demand for hardware and infrastructure.
  • The speaker believes the next three years (until 2028-2029) will be defined by a build-out of compute and power infrastructure to meet this demand.

Takeaways

  • Investors should view AI not as a short-term trade but as a long-term, foundational theme backed by both private sector innovation and significant government support.
  • The investment landscape within AI is shifting. While the first wave was driven by LLMs (like ChatGPT), the next wave will be driven by VLMs and VLAs. This suggests looking for companies involved in robotics, computer vision, and the hardware that powers these technologies.
  • The "too big to fail" government backing reduces the risk for major "frontier model" AI companies, making them akin to national champions like Lockheed Martin or J.P. Morgan.

Google / Alphabet (GOOGL)

  • Google is described as "surging" and potentially pulling ahead of competitors like OpenAI.
  • The company's Gemini 3 model is highlighted as a major breakthrough, especially in visual understanding (VLM). A key example was its ability to analyze and understand a chaotic whiteboard photo, a task that demonstrates a significant leap in AI capability.
  • Google's custom TPU (Tensor Processing Unit) chip is presented as a key differentiator and a source of real competition for NVIDIA. The TPU's "systolic array" architecture is more efficient for AI workloads by reducing data shuffling bottlenecks.
  • The market is recognizing this shift, with Google's stock performing well while the "OpenAI complex" (companies associated with OpenAI) has seen a downturn.

Takeaways

  • Google appears to be a primary beneficiary of the shift from LLMs to VLMs, where its Gemini model and TPU hardware provide a competitive edge.
  • Investors looking for AI exposure beyond NVIDIA should consider Google, as it is demonstrating leadership in the next phase of AI development and has the custom hardware to support it.

NVIDIA (NVDA)

  • NVIDIA's stock was down for the week, even as the broader market rallied. This is attributed to the narrative that its monopoly on AI chips is facing "real competition" from players like Google with their TPUs.
  • The speaker clarifies that this competition does not mean NVIDIA is a bad investment. The market has already discounted the fact that NVIDIA will not capture the entire future AI infrastructure market.
  • Despite competition, NVIDIA's revenue growth remains incredibly strong (up 62% year-over-year), and the overall demand for compute power from the VLM/VLA shift is set to explode, which will continue to benefit NVIDIA.
  • The company's upcoming Blackwell chip platform is also a positive catalyst.

Takeaways

  • While NVIDIA's stock may face short-term headwinds as the market adjusts to a more competitive landscape, the company is still a central player in the AI hardware boom.
  • The narrative is shifting from NVIDIA being the only winner to one of several key winners. The massive growth in compute demand is a rising tide that should still lift NVIDIA's boat significantly.
  • Investors should not interpret competition as a reason to be bearish on NVIDIA, but rather as a sign that the AI hardware market is maturing and broadening.

Small Caps & Market Broadening (IWM)

  • Small caps, as a group, had their biggest week since November of last year, indicating a potential shift in market leadership away from just the mega-cap tech stocks (MAG7).
  • The speaker believes the market is "broadening out," meaning more sectors and company sizes will participate in the rally. This is supported by improving economic indicators like PMIs (Purchasing Managers' Index) and the expectation of Fed rate cuts.
  • The speaker notes that winners next year will likely be different from this year's winners due to these factor shifts.
  • There is a strong correlation noted between the performance of small caps (IWM) and the biotech sector (XBI).

Takeaways

  • Investors may want to consider increasing their allocation to small-cap stocks (e.g., through an ETF like IWM) to capitalize on the market broadening.
  • The strong performance of small caps is a sign of a healthier, more risk-on market environment, which could have legs, especially if the Fed begins cutting rates.

Biotech & Pharma Sector (XBI)

  • The Pharma sector had its best month in 30 years relative to the S&P 1500.
  • The biotech ETF XBI is trending higher and outperforming the tech-heavy Nasdaq 100 (NDX).
  • This strength is directly linked to the AI revolution. AI is collapsing the cost of medicine and accelerating drug discovery.
    • VLMs are critical for medical imaging and diagnostics.
    • VLAs are enabling robotic lab automation.
  • The speaker highlights the goal of using AI and biotech to reverse aging and cure diseases within the next decade, which has massive economic and social implications.

Takeaways

  • The biotech sector is emerging as a major beneficiary of applied AI. This is not just a story stock theme; it is showing up in strong market performance.
  • Investors looking for ways to play the AI trend outside of pure technology companies should look at the biotech sector, potentially through ETFs like XBI.

Semiconductors (General) & Software (Short)

  • Semiconductors: The speaker is broadly bullish on the semiconductor sector. The massive compute requirements for VLMs and VLAs mean the demand for chips will continue to grow rapidly.
    • They expect the rally in semis to "broaden out" beyond just NVIDIA. Smaller companies involved in hardware for computer vision and other specific applications are poised to benefit.
    • Micron (MU) is mentioned as a company that will continue to do well due to the memory-intensive nature of new AI models.
  • Software: The speaker is bearish on the software sector, referring to it as the "short side" of the AI trade.
    • The reasoning is that AI, particularly models like Claude, will soon automate 90-100% of software development.
    • They do not expect software stocks to outperform semiconductor stocks going forward.

Takeaways

  • The investment opportunity in semiconductors is expanding beyond the main players. Investors should research smaller, specialized chip and hardware companies that support the new VLM/VLA paradigm.
  • Caution is advised for the traditional software sector. While some companies may integrate AI well, the entire industry faces a massive disruption from AI-driven code generation, which could compress margins and make many existing business models obsolete.

Bitcoin (BTC)

  • The speaker highlights a positive correlation between Bitcoin and the PMI (Purchasing Managers' Index), a key economic indicator.
  • They believe PMIs are set to move significantly higher, which, based on historical patterns, would be a strong tailwind for Bitcoin's price.
  • Trading activity in Bitcoin has also "bounced up significantly," indicating renewed market interest.

Takeaways

  • Investors with a bullish view on the economy (specifically, a manufacturing recovery indicated by rising PMIs) may see this as a positive signal for Bitcoin.
  • The analysis suggests that Bitcoin could act as a cyclical asset that performs well during periods of economic acceleration.
Ask about this postAnswers are grounded in this post's content.
Video Description
This week delivered one of the best performance since May, with the S&P 500 finishing up nearly 4%, the Nasdaq 100 up almost 5%, and small caps posting their third-best week in two years. Beneath the surface, the big story for the month was factor rotation: long momentum was crushed, its worst month relative to small caps since the GFC lows driven largely by retail capitulation. Size, growth–value dynamics, and pure momentum all show early signs of a potential regime shift tied to a PMI rebound forming beneath the surface. Earnings revisions remain at a four-year high, breadth is expanding beyond the Mag 7, and a Dec Fed rate cut is now likely pointing toward a broadening market in 2025 as the economic cycle accelerates. The second major theme is the Genesis Mission, a newly announced AI executive order that effectively creates a Manhattan-Project-style national program linking the DOE’s supercomputers with private frontier labs. This formalizes AI as national-security infrastructure and all but guarantees that leading AI companies become too important to fail. Eric Schmidt’s long-running influence, the fusion of government and private AI research, and the shift from text-only LLMs toward vision-language models (VLMs) and vision-language-action systems (VLAs) signal the next phase of AI: physical-world reasoning, robotics, and autonomy. This shift demands 5–20× more compute, triggers new hardware competition (Google TPU vs Nvidia), and forces hyperscalers to build massive “AI highways and gas stations” (datacenters + power + memory) to handle multimodal workloads. Finally, the biotech and healthcare angle is accelerating fast. Pharma just posted its best relative month in 30 years, XBI is turning up, and VLMs/VLAs are critical enablers for medical imaging, robotic lab automation, longevity research, and AI-driven drug discovery. Anthropic, Sinclair, and multiple Moonshots episodes highlight how AI may collapse healthcare costs, improve entropy-driven medical outcomes, and reshape entitlements, deficits, and inflation. Combine this with a broken U.S. poverty benchmark and AI-driven job disruption, and the macro backdrop becomes even more dependent on technological progress. Bitcoin tracks closely with PMI cycles, and if PMIs rise into the 60s—as leading indicators suggest—the crypto complex, small caps, and broadening equities should benefit significantly. 00:00 — A Shockingly Strong Tape The S&P 500 jumped nearly 4% this week, the Nasdaq 100 almost 5%, and small caps posted their third-best week in two years. Beneath the surface, the message was clear: breadth is expanding and market leadership is shifting. 02:18 — Momentum Crashes, Retail Capitulates The long side of momentum had its worst month relative to small caps since the GFC lows. Retail traders flushed out hard, and the factors most tied to retail activity saw major unwinds. 04:40 — Size and Growth/Value Begin a Regime Shift Size factor is rolling over after a multi-year run since the ChatGPT launch. Growth vs value is being driven by shifts in the PMI cycle, signaling new winners for 2025. 06:55 — PMIs Point Toward Acceleration IP diffusion, capital goods—are pointing toward PMIs heading towards the 60s in 2026. Every time we’ve seen this historically, factor rotation follows. 09:14 — Earnings Revisions Hit a 4-Year High Global earnings revision ratios are breaking out, and the “493” are quietly outperforming the Mag 7. This is classic early-cycle behavior and confirms the broadening story. 12:32 — Fed Cuts Coming; Don’t Fight the Tape Rate-cut expectations moved back toward 80% this week. Market internals, earnings, and macro indicators all line up with a “don’t fight the Fed, don’t fight the tape” environment. 14:50 — Genesis Mission: The U.S. Manhattan Project for AI Trump signed a sweeping AI executive order creating a unified computing backbone across the DOE’s 17 national labs. This effectively turns frontier AI companies into national-security infrastructure and pushes them into “too important to fail” territory. 18:25 — The Shift to VLMs and Robotics AI is now moving from text (LLMs) to vision-language models (VLMs) and vision-language-action systems (VLAs)—the models required for robotics and autonomy. These workloads are 5–20× more compute-intensive and will reshape hardware demand for the next decade. 22:40 — Nvidia vs Google TPU: The Architecture Battle Google’s new TPU architecture is a genuine challenge to Nvidia’s dominance, with systolic arrays bypassing key memory bottlenecks. This doesn’t kill Nvidia, but it explains why Google is surging while the NVDA complex is under pressure. 26:15 — AI = QE for the Mind AI is becoming a macro force equivalent to QE. If you fade AI, you’re fading both the government and the dominant driver of corporate alpha—and that has never been a winning strategy.
About Jordi Visser
Jordi Visser

Jordi Visser

By @jordivisserlabs

Empowering seasoned professionals to navigate the future of finance, technology, and AI. What We Offer: - Cutting-edge ...