My Investing Plan For The Next 5 Years

Key Concepts

• AI Infrastructure Scarcity: The current market phase where demand for AI hardware (GPUs, memory, networking) significantly outstrips supply, granting immense pricing power to manufacturers.
• Normalization: The transition phase where supply catches up to demand, shifting power from hardware sellers to software/platform buyers.
• Hyperscalers: Large-scale cloud providers (Google, Microsoft, Amazon, Meta) that own the customer relationship and distribution, positioning them as long-term AI winners.
• Cyclical vs. Durable Sellers: The distinction between companies selling one-time hardware components (cyclical) versus those with long-term service contracts and ecosystem lock-in (durable).
• Disruption of Traditional Media: The shift of audience attention and advertising revenue from traditional television (e.g., late-night talk shows) to digital platforms like YouTube.

---

1. The Four-Phase AI Roadmap

The speaker outlines a five-year evolution of the AI market:

• Phase 1: Scarcity (Current Phase): Characterized by bottlenecks in hardware. Companies selling "picks and shovels" (Nvidia, TSMC, ASML, Micron, Broadcom, power/cooling firms like Vertiv) hold immense pricing power.
• Phase 2: Normalization: Supply begins to meet demand. Investors will pivot from focusing on raw growth to the "durability of profits." Highly cyclical companies (memory manufacturers) face significant risk, while companies with structural moats (ASML, Nvidia) are expected to remain more stable.
• Phase 3: Power Shift to Buyers: Hyperscalers (Alphabet, Amazon, Microsoft, Meta) gain control by utilizing custom silicon (e.g., Google TPUs) and multi-vendor strategies. They monetize AI through massive distribution and long-term customer relationships.
• Phase 4: Software Resurgence: Software companies will split into two groups: those that integrate AI to become more profitable (winners) and those that are commoditized by AI agents (losers).

---

2. Investment Strategy and Portfolio Positioning

The speaker emphasizes a shift from "scarcity" winners to "long-term" winners:

• Direct Exposure: Maintains a position in ASML as a durable seller due to its monopoly on EUV machines and long-term service contracts.
• Core Bets: The portfolio is heavily concentrated in the "Phase 3" winners: Google, Microsoft, Amazon, and Meta. These are viewed as the most predictable long-term winners due to their proprietary data, massive distribution, and ability to monetize AI across existing ecosystems.
• Risk Management: The speaker avoids "Phase 1" companies that are purely cyclical (e.g., Micron, SanDisk) despite their recent market cap surges, citing the risk of a valuation "re-rate" once supply catches up.

---

3. Case Study: Ferrari’s EV Strategy

• The Issue: Ferrari’s new EV has been poorly received by the market, with the stock dropping 5.7%.
• Key Argument: The design is criticized for being generic and lacking the "Ferrari aesthetic." The speaker argues that Ferrari’s value is rooted in its brand history and the "roar" of its combustion engines. By creating a generic-looking EV, the company risks diluting the brand equity that justifies its premium pricing.

---

4. Fail of the Week: The Cancellation of The Late Show

• The Misconception: Many attribute the cancellation of Steven Colbert’s show to political conflict with Donald Trump.
• The Reality: The speaker argues the true culprit is YouTube.
• Supporting Data:
  • Total late-night network revenue has halved since 2018 (from $440M to $200M).
  • The Late Show faced annual losses of ~$40M with an overhead of ~$100M per year.
  • Structural Shift: Younger audiences prefer short-form clips on YouTube over hour-long traditional broadcasts. The traditional late-night model is deemed "dead" due to high production costs and declining advertiser interest.

---

5. Synthesis and Conclusion

The AI transition is a multi-year process. While current market gains are concentrated in hardware sellers (Phase 1), investors should be wary of cyclicality. The most sustainable long-term value lies with the "Hyperscalers" (Phase 3) who own the customer relationship and can integrate AI into existing, high-margin business models. Simultaneously, traditional business models—like late-night television—are being aggressively disrupted by digital platforms, serving as a reminder that structural shifts in technology and consumer behavior are the primary drivers of long-term market winners and losers.