Is AI Actually Saving The Stock Market? — With Tom Lee

Episode Summary

Executive Summary: Tom Lee argues AI is a major market narrative but not the sole driver of stocks: broader earnings growth, dovish Fed expectations, and resilient credit conditions have also powered the rally. He frames AI as a potential long-cycle winner with eventual shakeouts, but warns of two black swan risks—AI becoming too powerful or failing to meet expectations—while seeing Bitcoin and stablecoins as more useful long-term financial infrastructure.

Main Topics: AI as narrative vs. market driver (Priority: 5/5): Lee says AI has become the dominant story supporting bullish sentiment, especially for U.S. equities, but insists it is only one of several factors behind market gains. He distinguishes between narrative-led multiple expansion and actual earnings contribution across sectors. Broader earnings strength and market resilience (Priority: 5/5): He emphasizes that financials, industrials, and parts of tech have all contributed to earnings growth, and that the market has held up through multiple shocks because corporate earnings and liquidity conditions remained better than feared. Black swan events and supply-chain sensitivity (Priority: 4/5): Lee reviews five major shocks since 2020—COVID, supply-chain disruptions, inflation, aggressive rate hikes, and tariff escalation—arguing that none produced the sustained bear market many expected, because the S&P is highly sensitive to manufacturing and supply chains. AI bubble risk and second-order winners (Priority: 4/5): He compares AI to wireless and the internet: not every company will win, valuations may compress, and the sector may experience a shakeout. The eventual winners may be different from today’s leaders, with some companies capturing value later in the cycle. Tariffs, geopolitics, and market positioning (Priority: 4/5): Lee says the market’s V-shaped rebound after tariff fears reflected low recession risk, not just sentiment. He argues tariffs are often temporary, can be extended or rolled back, and are a poor tool for reshoring or countering China’s industrial advantages. Bitcoin, stablecoins, and financial infrastructure (Priority: 4/5): He remains constructive on Bitcoin, arguing that network effects, regulatory adoption, corporate treasury use, and stablecoins as a settlement layer make crypto more useful over time, not less. He sees stablecoins as a genuine Web3 application. Thematic investing and the Granny Shots ETF (Priority: 3/5): Lee explains his ETF approach: focus on major themes rather than macro or single-stock picking. He says thematic investing has historically outperformed and that AI is just one of several themes in the portfolio.

Key Arguments: AI is important mainly as a narrative that supports valuation expansion, but it is not the only factor driving the stock market. Broad earnings growth in banks, industrials, and other sectors means the rally is not solely a tech/AI phenomenon. A market can rally even with hawkish Fed rhetoric if recession risk stays low and credit conditions remain stable. The S&P 500 is especially sensitive to supply chains and manufacturing, which explains why supply-chain shocks mattered so much since 2020. If AI becomes extremely successful, it could create a societal black swan by making human labor less economically relevant. If AI falls short, the likely outcome is not economic collapse but a valuation reset and redistribution of winners and losers. Tariff fears were over-discounted; history suggests waterfall declines usually V-bounce absent recession. The right U.S. policy response is reducing regulatory friction and making it easier to build and manufacture domestically, not relying on tariffs. Bitcoin’s long-term case is network growth, institutional adoption, and its role as collateral/treasury reserve. Stablecoins may be the real killer app of crypto because they improve payments and settlement efficiency while increasing demand for dollars and Treasuries.

Data Points: Magnificent Seven capital expenditures: about $350 billion this year - Used to show that not all big tech spending is pure AI spend; much is maintenance and expansion CapEx. S&P 500 performance since ChatGPT era: up about 50% - Referenced to illustrate the market rally since the AI era began. Waterfall decline threshold: more than 10% within about 2 weeks - Lee’s definition of a rare market selloff pattern that typically V-bounces absent recession. High-yield recession signal: spread would need to widen to 800 basis points - Lee said that would imply recession-like stress; instead spreads widened only modestly. High-yield spread widening during tariff turmoil: about 150-200 basis points - Used to argue the market only priced a growth scare, not a recession. Recession odds vs economists: high yield implied about 10%; economists said 60% - Illustrates Lee’s point that bond markets were more accurate than headline forecasts. S&P rebound from April low: recovered after a waterfall decline - He cited historical precedent for a V-shaped recovery after sharp selloffs without recession. AI market concentration example: NVIDIA +7%, Meta +13%, Microsoft +15% year to date - Shown as winners in the AI trade and compared with laggards. AI laggards among megacaps: Amazon -5%, Google -15%, Tesla -15%, Apple -17% year to date - Used to show dispersion within the Magnificent Seven. Bitcoin price: around $101,000 - Referenced as evidence of Bitcoin’s strength and network growth. Bitcoin early estimate: could reach $25,000 by 2022 - Lee said he modeled this back in 2017 based on wallet activity and network value. Bitcoin explanation power: about 87% - He said wallets and activity per wallet still explain most of Bitcoin’s move. Stablecoin market size: $250 billion - Used to argue the market is still small relative to its potential and banking use cases. Stablecoin Treasury holdings: 12th largest holder of U.S. Treasuries - He said stablecoins collectively hold more Treasuries than many countries. Stablecoin demand comparison: twice Germany’s Treasury holdings - Used to underscore scale and government incentive to support stablecoins. Dollar dominance in traditional finance: 88% - Contrasted with crypto to explain why the dollar remains powerful in global markets. Dollar dominance in crypto quoted pairs: 100% - Used to show crypto’s dependence on the dollar. Stablecoin trading geography: almost 60% in Hong Kong, China, and Japan - Used to show offshore demand for dollar-linked stablecoins.

Pivotal Quotes: "it takes a whole lot of E to offset PE" — Tom Lee: Explaining why narrative and valuation expansion can matter more than earnings growth in the intermediate term. "AI is part of the narrative, but there's a lot of the narrative for the markets." — Tom Lee: His core thesis that AI is important but not singularly responsible for the stock market’s strength. "you sell the buildup, but you buy the invasion" — Tom Lee: His geopolitical market rule-of-thumb for trading around conflict risk and the U.S. strike on Iran.

Implications: Investors should treat AI as one powerful theme, not the whole market story. The bigger lesson is to watch liquidity, credit, earnings, and supply chains; long-term, AI may reshape labor and valuations, while Bitcoin/stablecoins may become more embedded in financial infrastructure.