Tom Lee Just Warned of a 20% Crash… While Stocks Are Ripping

The speaker opens by framing the market as strong but uncomfortable: stocks are back near record highs, AI names and semiconductors are surging, and momentum is broadening, but the NASDAQ and semiconductor moves look stretched and overbought. The core tension is that the rally appears to be supported by real earnings and AI demand, yet the speed of the move makes chasing riskier. The speaker repeatedly notes that markets can keep going higher, but that momentum often increases volatility and makes the next negative headline more dangerous. A major theme is that the rally is not “fake,” but it is being driven by a mix of fundamentals and FOMO. The speaker points to strong participation across technology, communication services, industrials, and some financials; then argues that underexposed investors who waited for a pullback are now being forced back into the market. …

1. The rally is broad and real, but it has also become stretched and more vulnerable to surprise headlines.
2. AI and semiconductors are the main momentum leaders, but the speed of the move has raised the risk of late chasing.
3. Tom Lee’s view is basically bullish now, cautious later: strong earnings and AI demand support prices, but a 15%–20% drawdown could still arrive later this year.
4. Elevated oil, inflation expectations, and geopolitics remain the main macro threats to high-growth valuations.
5. Stock selection matters more than index exposure right now; the speaker favors businesses with real earnings and cash-flow improvement.
6. Uber is presented as a high-quality example: strong demand, operating leverage, and nearly $10B of trailing free cash flow.
7. The transcript repeatedly warns that underexposed investors and FOMO can push markets higher even when valuations and sentiment are getting hot.

• The transcript leans on technical overbought signals as evidence of risk, but those indicators do not by themselves establish a bearish turning point.
• The call to a 15%–20% later-year drawdown is presented as a possibility, not a well-quantified forecast, so the timing and trigger confidence are limited.
• The speaker treats possible US-Iran de-escalation as supportive for risk assets, but that geopolitical path is highly uncertain and may be overstated.
• The DCF-based valuation discussion for Uber depends heavily on assumptions about long-term free cash flow growth, which are inherently fragile.
• The argument that the market is both expensive and still reasonably attractive relies on mixing index-level momentum with stock-specific value, which can be directionally true but analytically loose.