SMH ETF rallied to $602.14 on May 26, posting a 4.48% single-session gain as investors piled back into semiconductor stocks. The move followed a breakout session for Micron Technology and renewed concern over global memory-chip supply after labor disruption at Samsung.
The jump matters because SMH is one of the market’s largest and most liquid vehicles for semiconductor exposure, with $66.86 billion in assets under management. Its advance signals that investors are again favoring the AI infrastructure trade, especially companies tied to DRAM, NAND, GPUs and data-center buildouts.
At the center of the move is a simple market message: memory pricing power is improving again just as demand from advanced AI workloads accelerates. That combination has strengthened the case for higher earnings across major chipmakers in 2026 and beyond.
Key Facts
- SMH closed at $602.14 on May 26, up 4.48% in one session, after trading in a recent range of $558.64 to $603.21.
- The ETF manages $66.86 billion in assets and charges a 0.35% expense ratio, with an annual dividend rate of $1.10 for a yield near 0.19%.
- SMH is up roughly 60% year to date in 2026, making it one of the strongest-performing sector ETFs in the U.S. market.
- Micron, a top holding in the fund, surged about 18% to $886.60, helping drive nearly 2 percentage points of SMH’s gain.
- Samsung’s 18-day labor disruption reportedly cut foundry output by 58% and memory fabrication by 18% during affected shifts.
SMH ETF
The latest move in SMH reflects a sharp re-rating of the memory segment of the semiconductor market. Investors have shifted attention from chip demand in broad terms to a more specific thesis: AI inference is consuming far more memory and storage than many supply chains were built to support. That has lifted sentiment around companies exposed to DRAM, NAND and high-bandwidth memory, especially as cloud operators continue spending heavily on AI capacity.
Micron’s outsized rally gave this theme immediate market validation. As one of SMH’s largest positions alongside Nvidia and other semiconductor leaders, Micron’s advance had an amplified effect on the ETF. The broader sector also benefited, with the Philadelphia Semiconductor Index reaching fresh highs and memory-related names such as Seagate and Western Digital gaining in sympathy.
What makes the story more consequential is that it combines both demand strength and supply friction. Semiconductor rallies are often powered by one or the other. In this case, investors are seeing a market where AI model complexity is increasing memory consumption just as one of the world’s largest memory suppliers has suffered operational disruption. That raises the prospect of firmer pricing, wider margins and stronger earnings revisions across the space.
When rising AI memory demand meets even modest supply disruption, semiconductor pricing can tighten quickly and pull the entire sector higher.
Why Memory Has Returned to the Forefront
The memory trade had already been building momentum after component prices climbed sharply from mid-2025 levels. DRAM and NAND prices were reported to have risen as much as four to five times from the lows seen in July and August 2025, reflecting the strain created by AI servers, high-bandwidth memory demand and expanding enterprise storage needs.
That trend appeared to be moderating in early second quarter 2026 as supply improved. The Samsung labor dispute changed the near-term picture. Estimates that the disruption could trim global DRAM supply by 3% to 4% and NAND supply by 2% to 3% may sound limited, but even small shortages can have an outsized effect in a tightly balanced semiconductor market.
Demand is also becoming more memory-intensive. New generations of AI models are being built with context windows near 1 million tokens, up sharply from prior generations measured in the low hundreds of thousands. In practical terms, longer context windows and heavier inference workloads require more bandwidth and storage per query, supporting demand for the kinds of memory products manufactured by the largest names in SMH.
Implications for Investors
For investors, SMH remains one of the clearest listed expressions of the AI infrastructure theme. The ETF offers concentrated exposure to companies at the center of the buildout, including leaders in compute, memory, networking and wireless chips. Its scale and liquidity also make it a preferred vehicle for institutions seeking broad semiconductor exposure without taking single-stock risk.
The bull case rests on earnings growth that still appears unusually strong for a mature hardware segment. Forecasts cited in the market point to portfolio-level earnings growth above 100% in 2026, followed by roughly 36% growth in 2027. If those expectations hold, valuation may remain supportable even after a 60% year-to-date rally. Investors betting on SMH are effectively wagering that AI capital spending remains resilient and that memory pricing stays firm through the second half of 2026.
The risks are equally clear. Semiconductor stocks are cyclical, and the market often discounts the peak well before capital spending actually slows. If hyperscale data-center demand moderates, if major AI deployments are delayed, or if signs emerge that infrastructure spending is outpacing monetization, the sector could correct sharply. Drawdowns of 20% to 30% are not unusual in semiconductor cycles, even when the long-term trend remains intact.
There is also a technical factor to watch. SMH has now pushed above the upper end of its recent consolidation near $603, with prior support around $560 to $570 and medium-term trend support near the 50-day moving average around $550. Momentum remains strong, but overbought conditions can leave the ETF vulnerable to fast reversals if a near-term catalyst disappoints.
Relative positioning versus peers also matters. SMH charges the same 0.35% expense ratio as SOXX, while lower-cost alternatives such as SOXQ offer cheaper access at 0.19%. Even so, SMH’s large asset base and deep liquidity make it especially attractive for investors who prioritize tradability and concentrated exposure to the biggest semiconductor names.
The next phase for SMH will likely depend on whether the memory rally broadens into a sustained earnings cycle or fades as supply catches up. Investors should watch Micron’s pricing commentary, Samsung’s production normalization, hyperscaler capital-expenditure plans and any signs that AI inference demand is accelerating faster than chipmakers expected.
VIP Trading Signals
