Why Bubble Talk is Totally Wrong | TCAF 238

Key Concepts

• Exponential Growth vs. Linear Thinking: The core investment philosophy that markets often fail to grasp the non-linear, exponential nature of technological advancement.
• Compute Shortage: The global deficit in processing power (GPUs/chips) required to fuel the current AI revolution.
• Tokens: The fundamental unit of intelligence and compute; not all tokens are created equal (e.g., simple queries vs. complex drug discovery).
• Life Cycle Change: A framework for identifying companies that have saturated their original markets and are successfully reinventing themselves for new growth.
• Agentic AI: AI systems capable of performing tasks autonomously (e.g., filling out college applications, coding apps) rather than just providing information.
• Capex (Capital Expenditure): The massive spending by hyperscalers on data centers and chips, which the guest argues is justified by an 18-month payback period.

---

1. Investment Philosophy and Methodology

Dr. Encore Crawford, a portfolio manager at Alger, emphasizes a growth-oriented strategy focused on two primary buckets:

• High Unit Volume Growth: Companies disrupting markets with rapid revenue expansion.
• Life Cycle Change: Established companies (e.g., Apple, Microsoft, Nvidia) that have saturated their initial markets and are successfully pivoting to new growth engines.
• Active Management: Crawford argues that because many innovative companies are staying private longer (reaching near-trillion-dollar valuations), active managers must utilize the 15% allocation allowed in 40-act funds to invest in private markets to capture value before IPOs.

2. The AI and Compute Thesis

• The "Short Compute" Reality: Crawford asserts that the world is fundamentally short on compute. She dismisses "bubble" talk, noting that unlike the 2000 dot-com era—where infrastructure was built without the necessary software or demand—the technology today is already functional and in high demand.
• Capex Justification: While hyperscalers are spending heavily, Crawford notes that the payback period for this compute investment is currently around 18 months. She views this as a rational deployment of capital rather than wasteful spending.
• Tokens as Currency: She explains that tokens are the "units of thinking." As AI agents become more capable, the demand for tokens will grow exponentially. She notes that the market will eventually differentiate pricing based on the value of the task (e.g., educational vs. high-stakes scientific research).

3. Notable Company Analysis

• Amazon (AMZN): Viewed as a top-tier holding. Crawford highlights its low valuation relative to its growth and the strategic importance of its "Tranium" chips, which help lower the cost of compute within the AWS stack.
• Nebius (NBIS): Described as a potential "next-generation AI-native hyperscaler" born from the former Yandex engineering team.
• QXO: A building products company led by serial entrepreneur Brad Jacobs. Crawford highlights this as a "life cycle change" play where a fragmented, sleepy industry is being consolidated to drive massive EBITDA growth.
• AppLovin: Originally an ad-tech company for mobile games, it is successfully pivoting to e-commerce and broader advertising, maintaining high efficiency with a flat headcount.
• Intel (INTC): Crawford expresses skepticism regarding Intel’s foundry ambitions. While she praises CEO Pat Gelsinger, she questions the execution risk of building a foundry business from scratch in the U.S.

4. Perspectives on Software and "SaaS Apocalypse"

• The Software Threat: Crawford argues that traditional SaaS companies are at risk because they are often sold on a "per-seat" basis, yet users only utilize 15–20% of the software.
• Agentic Disruption: As AI allows companies to build custom internal tools, the need for expensive, bloated third-party software packages will decline. She suggests that many SaaS companies may see their operating margins compress from 40% to 20% as they lose their pricing power.

5. Significant Statements

• "We’ve been trained to think linearly... but we are in a time of exponential growth, and it’s really hard to get your arms around what exponential actually means." — Dr. Encore Crawford
• On the "bubble" narrative: "What I have found is that sometimes it’s almost easier if you don’t understand something to be like, 'Oh, it must be a bubble.'"
• On the difference between the 2000 bubble and today: "In 2000, we had dreamed the dream, but we actually didn’t have the technology... This time, you’ve actually built the car and the only thing you’re waiting for is gas."

Synthesis

The main takeaway is that the current market environment is defined by a massive, non-linear shift toward AI-driven productivity. Dr. Crawford argues that the "compute shortage" is a real, physical constraint, not a financial bubble. Investors should focus on companies with strong management teams that are either building the essential infrastructure (compute/memory) or successfully pivoting their business models to leverage AI agents. The "SaaS" model is under threat, and the future belongs to companies that can provide tangible, high-value intelligence rather than just software seats.