AI high from 2025 is not the same for 2026, Jeffrey Small says

Key Concepts

• Artificial Intelligence (AI) Investment: Focus on companies benefiting from the AI boom, both directly and indirectly.
• Hyperscalers: Large-scale cloud and technology companies (Microsoft, Amazon, Alphabet, Meta, NVIDIA) with established business models.
• AI Infrastructure: The underlying systems and resources needed to support AI development and deployment (cloud computing, cooling systems, energy storage).
• Valuation & Risk: Assessing the risk associated with investing in high-valuation AI startups versus established hyperscalers.
• Diversification: Expanding investment beyond core tech stocks to include supporting industries.
• Price-to-Earnings (P/E) Ratio: A valuation metric used to compare a company’s stock price to its earnings per share.

Microsoft’s AI Investment & Market Positioning

The discussion centers on Microsoft’s significant investment in Anthropic, an AI startup, reportedly reaching $500 million annually. This investment positions Microsoft as Anthropic’s top customer, utilizing the AI’s capabilities to power its products. It’s noteworthy that Microsoft was also the original funder of OpenAI, committing $10 billion previously, demonstrating a strategy of diversifying its AI investments. Jeffrey Small, a Retirement Income Source Partner, argues Microsoft is uniquely positioned to profit from the AI explosion due to its existing ecosystem. He explains that while Amazon is focused on owning the “railroad” (infrastructure), Microsoft aims to support “all the engines” (various AI models) running on that infrastructure, giving it a competitive advantage. Microsoft’s Azure cloud business is central to this strategy, allowing it to host and monetize AI infrastructure from multiple providers.

Tech Stock Valuations & Diversification Strategies

The conversation addresses the high valuations of some tech stocks. NVIDIA, for example, has a P/E ratio of 45, which is considered high. Microsoft’s P/E ratio is comparatively lower. Meta’s P/E ratio is currently at 27. Small notes a shift in investor sentiment, moving away from solely chasing AI “hype” and towards investing in companies supporting the AI space. This includes sectors like cooling systems (Train Air-Conditioning Systems) and energy storage, driven by the increased demand from data centers powering AI. He highlights the substantial demand for cooling and power solutions as critical components of the AI infrastructure.

Hyperscalers vs. AI Startups: A Risk Assessment

Small emphasizes the importance of investing in “hyperscalers” – established tech giants – over many AI startups. He points out that many startups have “tremendous valuations” and were significantly impacted in November/December, experiencing stock declines. Hyperscalers, however, already possess established business models that will be enhanced by AI, leading to increased profitability. He specifically names NVIDIA, Microsoft, and Meta as key players, acknowledging that others also have a role. He states, “A.I.’s going to enhance their profitability. So you definitely want to own the hyperscalers in this environment.”

Retirement Investing & Long-Term Perspective

Addressing the question of whether these high-valuation tech stocks are suitable for retirement accounts, Small clarifies that it depends on an investor’s portfolio size and income needs. If an investor has sufficient income from other sources, allocating capital to growth stocks like these, even at “all-time peaks,” can be a viable long-term strategy. He acknowledges the lack of substantial dividends from these companies but argues that long-term growth potential justifies the investment for those with adequate existing income streams.

The Current Stage of AI Development

The discussion touches on the idea that the AI revolution is still in its early stages, described as being “only in the second inning of the A.I. ball game.” While acknowledging potential hyperbole from those seeking to enter the market, the speaker suggests that the fundamental growth potential remains significant.

Logical Connections

The conversation flows logically from Microsoft’s specific investment in Anthropic to a broader discussion of the AI investment landscape. It then transitions into a risk assessment comparing hyperscalers and startups, and finally addresses the practical implications for retirement investing. The shift from focusing on AI hype to supporting infrastructure highlights a maturing understanding of the AI ecosystem.

Data & Statistics

• Microsoft’s investment in Anthropic: $500 million per year.
• Microsoft’s previous investment in OpenAI: $10 billion.
• NVIDIA’s P/E ratio: 45.
• Meta’s P/E ratio: 27.
• Timing of AI startup stock declines: November and December (year unspecified).

Notable Quotes

• Jeffrey Small: “Amazon is trying to own the railroad, and the A.I. building the engines…Microsoft if doesn't need to own the engine, they need to just deal with all the engines.”
• Jeffrey Small: “A.I.’s going to enhance their profitability. So you definitely want to own the hyperscalers in this environment.”

Synthesis/Conclusion

The key takeaway is that Microsoft is strategically positioned to capitalize on the AI boom by focusing on providing the infrastructure for various AI models, rather than solely developing its own. Investing in established hyperscalers, while potentially expensive based on current valuations, is considered a safer bet than many high-valuation AI startups. Diversification into supporting industries like cooling and energy storage is also recommended. The discussion emphasizes a long-term perspective and the importance of aligning investment strategies with individual financial circumstances, particularly for retirement planning. The AI revolution is still in its early stages, presenting both opportunities and risks for investors.