Tom Lee: Why This AI Rally Isn't a Bubble

F
Fundstrat May 06, 2026

Audio Brief

Show transcript
This episode covers the current state of the stock market rally, evaluating whether semiconductor and artificial intelligence equities are becoming overvalued. There are three key takeaways. First, semiconductor valuations remain historically reasonable. Second, market growth is driven by a tangible scarcity of compute power. Third, AI will act as a massive macroeconomic catalyst over the next five years. Expanding on these points, the forward price to earnings ratio for semiconductor stocks sits around twenty two. This is well below the historical peak of thirty five, suggesting the risk to reward balance remains favorable. Additionally, this rally is grounded in actual earnings and supply chain scarcity, spanning both hardware and energy providers. Finally, AI productivity is expected to add two percentage points to US GDP annually, driving an estimated six percent baseline growth in S and P five hundred earnings. Investors should contextualize these current valuations against historical data and physical supply constraints to navigate this ongoing growth trend.

Episode Overview

  • This episode examines the current state of the stock market rally, specifically evaluating whether semiconductor and AI-related equities are becoming overvalued after hitting record highs.
  • Fundstrat's Tom Lee provides a framework for analyzing market risk and reward, arguing that the underlying fundamentals of the current tech rally remain sound despite recent price surges.
  • The discussion is highly relevant for investors, analysts, and market watchers seeking to understand the macroeconomic impact of AI productivity and how to contextualize current market valuations against historical data.

Key Concepts

  • The Valuation Context for Semiconductors: Despite a strong recent rally, the forward P/E ratio for semiconductor stocks sits around 22x. Because this metric has historically reached as high as 35x over the last two decades, these assets are not yet considered fundamentally "expensive," suggesting the risk/reward balance remains favorable.
  • The Scarcity Premium: Current market growth is not solely driven by speculative hype; it is grounded in fundamental earnings and a tangible scarcity in compute power and related supply chains. This includes both the hardware (semiconductors) and the energy required to power them.
  • Macroeconomic Impact of Artificial Intelligence: AI is projected to be a massive macroeconomic catalyst, expected to add approximately two percentage points to US GDP annually over the next five years. This structural economic boost translates to an estimated 6% baseline growth in S&P 500 earnings driven purely by AI-enabled productivity.

Quotes

  • At 0:34 - "stocks are rising for the right reasons... good earnings, but also just evidence of really the scarcity of compute and the supply chain whether it's semis and energy." - Explains the fundamental drivers behind the current tech rally, moving beyond pure market sentiment to highlight tangible supply and demand dynamics.
  • At 0:46 - "the market hasn't made these stocks expensive yet. The forward PE of the semi index is still only 22 times... they had gotten as high as 35 times over the last 20 years." - Contextualizes current valuations against historical peaks, providing a quantitative basis for continued market optimism.
  • At 1:06 - "AI for the US is probably adding two percentage points to GDP each year for the next five years. You know that's like six percent S&P earnings growth coming from AI productivity." - Quantifies the anticipated macroeconomic and corporate earnings impact of artificial intelligence integration, framing it as a long-term economic driver.

Takeaways

  • Evaluate technology investments by comparing current forward P/E ratios against their historical peaks, rather than simply looking at recent price appreciation, to avoid prematurely exiting secular growth trends.
  • Factor the "scarcity of compute" into portfolio construction by looking beyond consumer-facing AI software and considering foundational supply chain layers, such as semiconductor manufacturers and energy providers.
  • Maintain a balanced risk assessment by monitoring physical macroeconomic constraints—such as potential shortages in petroleum and energy products—which could serve as headwinds against AI-driven productivity gains.