Small-cap stocks enjoy their best first half in 35 years. Here's what's driving it

The Russell 2000 jumped more than 21% in the first half of 2024 (best since 1991), driven by AI infrastructure spending that is rippling through semiconductors and their equipment suppliers. LPL expects 2026 earnings growth for Russell 2000 companies to rise to 38% from ~23% earlier in the year, supporting a potential broadening beyond tech. The key risk is higher-for-longer rates: Bank of America estimates each additional 25 bps hike would reduce Russell 2000 operating earnings by ~2% (amplified by refinancing risk), even as investors increasingly think the tightening cycle is nearing peak.

Analysis

This is less a classic small-cap macro beta trade than a supply-chain torque trade: the names with the most upside are the ones where incremental AI capex hits a tiny revenue base and expands margins fastest. That favors AEHR, ICHR, and MXL over the broad Russell 2000, because their earnings power is more sensitive to order inflections than to the index-level valuation catch-up narrative. NVDA remains the upstream reference point, but the better risk/reward is often in the second-ring vendors where revisions can outrun the stock moves if bookings stay strong for another quarter or two.

The near-term risk is rates, not AI. A 25bp hike meaningfully hurts levered small caps and refinancing-heavy balance sheets before it shows up in consensus, so the first losers would be lower-quality industrial and consumer small caps rather than the AI suppliers themselves. Higher-rate volatility should also support CME and, to a lesser extent, banks with strong funding bases like BAC, while leaving rate-sensitive asset gatherers and weaker lenders exposed. If September hike odds keep rising, I would expect the broadening trade to narrow back to quality within 4-8 weeks.

The contrarian miss is that the market may be extrapolating AI capex breadth too far into 2026. Stocks that have already rerated several hundred percent are pricing in uninterrupted order momentum; any pause in hyperscaler spending or a softer book-to-bill could compress multiples faster than earnings grow. The thesis is falsified if AI supplier backlog and guideups stay intact through the next two earnings cycles and the Fed clearly pivots dovish; otherwise, this looks more like a tradable wave than a durable regime change.

Key Decisions for Investors

• Go long AEHR/ICHR/MXL as a basket vs short IWM or RTY futures for 1-3 months; this isolates AI-supply-chain operating leverage from the broader small-cap rate overhang. Risk/reward is attractive if order momentum persists, but cut the trade if Russell breadth holds while these names stop raising guidance.
• Buy CME on any pullback as a 1-2 month hedge on higher-rate volatility; the setup improves if September hike pricing rises and front-end realized vol picks up. Falsify if the Fed turns explicitly dovish and rate-vol collapses.
• Avoid adding NVDA here; it is the crowded upstream beneficiary and already has the cleaner consensus path. If forced into the AI theme, own the suppliers rather than the leader for better asymmetry over the next quarter.
• Watch BAC as a relative winner versus rate-sensitive small financials over the next 3-6 months; it should be more resilient if higher-for-longer pricing returns. Reverse the view if yield-curve steepening and credit deterioration become the dominant macro signals.