If the AI Bubble Bursts, These Stocks Will Be Winners

The 10-year Treasury yield fell 3.8% to 4.95%, its first close below 5% in a month, and was down 7% over two weeks as investors reassessed rate and AI-spending risks. The Nasdaq Composite dropped 2.6% Thursday and Meta fell 7%, while the SPDR S&P Homebuilders ETF rose 3.8%; Pool gained 14.4%, RH 9.5%, and Lennar 9%. The article points to a possible rotation toward housing stocks if yields continue to decline, while noting that AI compute demand remains strong and future interest-rate direction is uncertain.

Analysis

The key transmission is not simply “AI out, housing in”: a fall in Treasury yields helps housing only if it reaches mortgage rates without signaling a damaging deterioration in household income or confidence. Mortgage rates also depend on MBS spreads, so the 10-year move alone is an incomplete catalyst. If yields are falling because investors are marking down growth, weaker affordability and buyer traffic could offset lower financing costs; builders may then defend volume with incentives, pressuring margins. The one-day strength in LEN, POOL, and RH is therefore a flow signal, not confirmation of a housing recovery.

The AI selloff is more discriminating than a broad demand rejection. Funding-sensitive compute providers such as CoreWeave face a higher cost of capital and a tougher refinancing narrative; Meta’s risk is instead whether incremental AI capex earns adequate returns, with cash-flow conversion the near-term test. Memory and other constrained suppliers could remain beneficiaries if compute demand holds, though the article provides no company-level evidence to size that exposure.

Over days, rotation and positioning can extend the housing rebound. Over 1–3 months, mortgage-rate pass-through, builder orders/cancellations, incentives, and Meta’s capex/FCF trajectory matter more. Over 6–18 months, housing gains require sustained affordability improvement, while AI infrastructure valuations require demonstrated monetization. Contrarian risk: the market may be treating lower yields as an unambiguous housing positive when they could instead reflect weaker growth; a one-session rally in beaten-down cyclicals is not proof of durable leadership.

Key Decisions for Investors

• Do not chase the single-session XHB/LEN move. Consider a staged, modest long in LEN or XHB only if mortgage rates follow Treasury yields lower and builder orders hold; cut the thesis if orders weaken or incentives rise enough to pressure margins.
• Within AI exposure, differentiate funding risk from platform risk: avoid adding to CRWV until its financing terms and liquidity/refinancing runway are verified. For META, track capex guidance and subsequent free-cash-flow conversion rather than treating one negative quarter as proof of failed returns.
• Use mortgage rates and MBS spreads—not the 10-year yield alone—as the housing trigger. A reversal in yields, widening mortgage spreads, or worsening housing demand would falsify the rotation thesis; monitor builder cancellations, order trends, and incentives.
• Treat the housing rally as a relative-value watch, not a confirmed sector regime change. If AI weakness broadens alongside deteriorating growth data, reduce cyclical housing exposure even if Treasury yields fall.