It’s Not a Bubble, It’s a Wave | WAYT?

Key Concepts

• Agentic Brokerage: An investing platform (e.g., Public.com) that allows users to create AI agents to automate portfolio tasks like buying dips or rebalancing based on specific conditions.
• Inference vs. Training: The distinction between training AI models (the current massive spend) and running them (inference), which is expected to be the next major growth phase.
• Capex Super-Cycle: The massive, sustained capital expenditure by hyperscalers (Google, Amazon, etc.) on data centers and AI infrastructure.
• "Wave" vs. "Bubble": The argument that current AI spending is a "wave" (a sustained, productive infrastructure build-out) rather than a "bubble" (which implies demand evaporation and bankruptcy).
• SaaS Apocalypse: The recent period where software-as-a-service (SaaS) stocks were aggressively re-rated lower as capital shifted toward semiconductor and AI-infrastructure companies.

---

1. Nvidia Earnings and Market Outlook

• Earnings Expectations: Nvidia is expected to report $79.1 billion in revenue (80% YoY growth) and $1.77 in EPS (119% YoY growth). Data center revenue is projected at $73.2 billion, effectively representing the entire company.
• Anticipation vs. Reaction: Nvidia is characterized as an "anticipation stock" rather than a "reaction stock." It often rallies leading up to earnings because customers (hyperscalers) have already signaled their intent to increase capital expenditure.
• Key Storylines:
  • Blackwell Demand: Whether the new chip remains sold out.
  • Margin Durability: Monitoring for any signs of cyclical deterioration.
  • Inference Opportunity: How Nvidia defends its moat against custom ASICs (Google/Amazon) and competitors like Cerebras.
• Valuation Context: Despite its massive run, Nvidia is currently one of the "poorer" performing stocks in the SMH semiconductor ETF year-to-date, as smaller players like Intel and Micron have seen higher percentage gains from lower bases.

2. The "Wave" Theory of AI Infrastructure

• The Argument: Josh Brown argues that the current AI build-out is a "wave," not a "bubble." Unlike the 1999 internet bubble, which was funded by speculative IPOs, current spending is funded by the cash flows of the world’s most profitable companies.
• Structural Constraints: Citing analyst Gavin Baker, the speakers note that there is no "slack" in the system. Leading-edge wafer capacity and power are structurally constrained, meaning the build-out cannot be turned off or on quickly.
• The "Shortage" Indicator: The fact that only 10 basis points of the population are currently using these models, yet there is already a massive compute shortage, suggests the cycle is in its early stages.

3. Space Sector Opportunities

The speakers highlighted three companies gaining attention ahead of a potential SpaceX S-1 filing:

• Intuitive Machines (LUNR): Positioning itself as a "space prime" contractor for NASA and the Space Force, focusing on lunar logistics and infrastructure.
• AST SpaceMobile (ASTS): Providing direct-to-cell satellite coverage; has secured significant contracts with major carriers like AT&T and Verizon.
• Planet Labs (PL): Known as the "Google of the sky," it provides daily Earth imaging and is increasingly integrating AI inference directly onto satellites.

4. Shifting Investment Heuristics

The discussion referenced Adam Parker’s research on "unlearning" past investment rules:

• Market Cap Concentration: Historically, owning the largest stocks was a poor strategy, but this has flipped in the last decade.
• Quality vs. Junk: The median high-quality stock has seen multiple contraction since 2020, while some lower-quality names have outperformed.
• Short Interest: Contrary to traditional wisdom, growth stocks with low short interest have outperformed those with high short interest since 2020.

5. Software vs. Semiconductors

• The Divergence: A chart comparing the S&P 500 Software Index to the Semiconductor Index shows a historic divergence since the launch of ChatGPT.
• Paradoxical Play: The speakers suggest that if one fears an AI bubble, the "paradoxical" play might be to buy the beaten-down software sector, which has been abandoned in favor of AI darlings.
• Cybersecurity Exception: Cybersecurity stocks (e.g., CrowdStrike, Palo Alto Networks) are viewed as essential infrastructure that will benefit from, rather than be disrupted by, AI.

6. Notable Quotes

• Josh Brown on the AI build-out: "Bubbles pop because demand evaporates. Waves can rise and fall, but they do continue."
• Josh Brown on the "SaaS Apocalypse": "The only thing worse for a software company that's struggling to convince its customers to stay... is 'and here's more services we want you to pay for because we just bought this other company.'"

Synthesis/Conclusion

The market is currently defined by a massive, capital-intensive shift toward AI infrastructure. While comparisons to the 1999 bubble are common, the speakers argue that the current environment is fundamentally different due to the nature of the funding (corporate cash flow vs. IPO speculation) and the structural constraints on supply. Investors are advised to look past the "AI darlings" to potentially undervalued sectors like software and space, while acknowledging that traditional investment heuristics—such as avoiding high-capex companies—may no longer apply in this new cycle.