Stretched Tech Valuations Raise Questions on Capex, BofA Says

Bank of America strategist Sebastian Raedler warned that the tech rally looks overdone after Micron surged on a blockbuster sales forecast. He questioned who will ultimately fund the capital spending behind the demand, signaling concern that AI- and chip-related enthusiasm may be getting ahead of fundamentals. The remarks are likely to weigh on sentiment for semiconductors and broader tech, but do not constitute a direct company-specific downgrade.

Analysis

The key second-order issue is not whether AI demand exists, but whether the funding stack behind it is durable. If capex is being financed through a small number of hyperscalers and AI infrastructure vendors, earnings expectations across the semiconductor complex can remain elevated longer than fundamentals justify; once free-cash-flow discipline reasserts itself, the unwind tends to hit equipment makers and memory first, then broader semis with a 1-2 quarter lag. That makes the current setup more vulnerable to a sentiment air pocket than to an immediate demand collapse.

For MU, the market is implicitly pricing a clean multi-quarter demand runway, but memory is where supply response is fastest and pricing power is most cyclical. The risk is that the marginal buyer of AI capacity becomes more price-sensitive than investors expect, forcing a pause in orders before end-demand weakens—historically this creates a sharp multiple compression even while revenues are still rising. In other words, the stock can de-rate on capex digestion alone, without needing a full-blown downturn.

BAC is a weaker direct read-through, but the broader implication is that financials are competing with megacap tech for capital allocation and index leadership. If AI capex begins to look like a capital sink rather than an earnings engine, investor rotation could favor sectors with visible cash returns and less reinvestment intensity, which is modestly supportive for banks versus long-duration growth. The contrarian point is that the market may be underestimating how long the AI spend cycle can stay elevated if hyperscalers treat it as strategic infrastructure rather than discretionary capex.

Key Decisions for Investors

• Fade strength in MU via a short-dated bearish options structure: buy 1-2 month puts or put spreads on rallies into post-forecast enthusiasm; risk/reward favors limited premium outlay because semis often de-rate faster than earnings estimates reset.
• Pair trade: long BAC / short MU over the next 4-8 weeks to express a rotation out of capex-intensive growth and into cash-returning financials; the trade works best if rates stay rangebound and AI sentiment cools.
• If staying long semis, prefer equipment beneficiaries with backlog visibility over memory names; avoid adding to MU until investors get evidence of capex moderation from hyperscalers, which is typically a 1-2 quarter process.
• For portfolio hedging, consider a partial NASDAQ hedge against AI-capex disappointment using QQQ puts or a call spread structure; this is most attractive if implied vol is still below realized vol after recent momentum.
• Set a tactical stop on any MU short if hyperscaler capex commentary re-accelerates or memory pricing inflects higher for another quarter; the squeeze risk is highest when positioning is crowded and fundamentals are still improving.