These Were the Best-Performing Asset Classes in May. Can They Continue to Lead?

May was broadly positive for risk assets, with U.S. stocks up 5.2%, the S&P 500 up 5.3%, and the Nasdaq-100 surging 10.6% as mega-cap tech and AI-linked names led gains. Developed international stocks rose 4.3% and emerging markets gained 1.6%, while commodities were the laggard, falling 7.5% as crude oil dropped 12.2% and gold declined 1.5%. The article frames AI infrastructure spending and semiconductor supply-chain leaders as the main drivers of continued tech outperformance.

Analysis

This is less a broad risk-on tape than a continuation of a narrow factor regime: liquidity and capex intensity are concentrating returns into the AI hardware/software stack. The second-order effect is that the winners are no longer just model builders; it’s the picks-and-shovels cohort with pricing power or supply bottlenecks, which is why semicap and memory names are inflecting faster than the broader index. That dynamic can persist for months if hyperscaler budgets stay intact, but it also makes the trade increasingly self-reinforcing and crowded.

The market is underappreciating how asymmetric the supply-chain beneficiaries are relative to the headline AI leaders. Memory and lithography exposure can re-rate quickly because investors are finally pricing a multi-quarter earnings revision cycle, while the large-cap platform names are already partially owned for the AI story and may have less near-term multiple expansion left. Intel is the odd one out: it benefits only if foundry credibility improves, otherwise it risks being a lagging beneficiary of the same capex wave rather than a true winner.

The commodity weakness matters more as a macro signal than as a directional call on oil or gold. A softer commodity complex eases input-cost pressure and supports margin duration for tech/industrial end markets, but it can also signal that growth expectations are being pulled forward by a handful of AI-related capex names rather than a clean global acceleration. If the AI spend cadence slows, this entire leadership stack can de-rate quickly because the market has already front-loaded much of the good news.

The contrarian risk is consensus overconfidence in a one-way AI buildout. Any evidence of hyperscaler capex discipline, export restrictions, memory pricing normalization, or delivery bottlenecks in advanced packaging would hit the most levered winners first and fastest. The setup favors tactical participation, not passive chasing, because the leadership group is strong but increasingly sensitive to even small revisions in forward spend assumptions.

Key Decisions for Investors

• Long MU / short QQQ for 3-6 months: MU has the cleanest operating leverage to AI memory pricing while QQQ remains dominated by names already priced for perfection; target 15-20% relative upside if memory ASPs hold, stop if hyperscaler capex commentary turns cautious.
• Buy AMD on pullbacks for a 1-3 month trade, but size it smaller than MU: AMD has strong AI optionality, yet the stock is more vulnerable to narrative compression if expectations outrun near-term deployment; favorable risk/reward only on dips of 5-8% from recent highs.
• Long ASML vs short INTC over 6-12 months: ASML is the direct toll collector on advanced-node investment, while INTC needs execution proof before the market pays for its foundry ambition; this is a cleaner way to express capex durability with asymmetric downside in the short leg.
• Avoid chasing ORCL upside after the recent move; instead consider selling out-of-the-money calls against a long position if held: the stock can keep working on AI infrastructure demand, but upside is likely to be slower and more linear than MU/AMD, making premium collection attractive.
• If commodity weakness deepens, add a tactical long tech / short energy pair for 1-2 months: lower input costs and easier financial conditions support AI capex beneficiaries, while energy loses on both price and sentiment; use XLE as the hedge if sector exposure is needed.