Is This ETF the Smartest Investment You Can Make Today?

The author recommends gaining AI exposure via the VanEck Semiconductor ETF (SMH), which tracks the MVIS US Listed Semiconductor 25 and has returned 38% year-to-date in 2025 versus 17% for the Nasdaq, with Nvidia the largest holding at an 18.5% weight and the five biggest names (Nvidia, TSMC, Broadcom, Micron, AMD) comprising roughly half the fund. SMH provides diversified exposure to critical AI infrastructure—GPUs, CPUs, AI accelerators, DRAM and foundry capacity—with recent catalysts such as Broadcom’s OpenAI accelerator deal and Grand View Research’s projection of AI market growth from $279bn in 2024 to $3.5tn by 2033 underpinning demand. The ETF charges a 0.35% expense ratio (competitive for a focused thematic fund) but carries concentration and cyclicality risks tied to a few large caps; the author discloses a personal Nvidia position and notes Motley Fool’s own recommendations differ.

Analysis

The author recommends the VanEck Semiconductor ETF (SMH) as a diversified way to capture AI infrastructure exposure, noting the fund tracks the MVIS US Listed Semiconductor 25 and has returned 38% year-to-date in 2025 versus 17% for the Nasdaq Composite. SMH holds 25 of the largest U.S.-listed chip companies with no single holding above 20%; Nvidia is the largest position at 18.5% (34,986,193 shares) and the top five names—Nvidia, TSMC, Broadcom, Micron, and AMD—account for nearly half the ETF. Specific stock moves cited include Micron up 172% YTD, AMD up 91.4%, Broadcom up 46.8%, and TSMC up 41.3%, illustrating concentrated upside within a handful of names.

The note cites structural demand drivers for semiconductors tied to AI, referencing Grand View Research’s projection of global AI market growth from $279 billion in 2024 to $3.5 trillion by 2033 (CAGR 31.5%), and highlights product-level roles: Nvidia GPUs, AMD CPUs, Broadcom accelerators (including a cited Broadcom–OpenAI 10-gigawatt accelerator deal), Micron DRAM, and TSMC foundry capacity. The fund’s 0.35% expense ratio (about $35 per $10,000) is positioned as reasonable for a focused thematic ETF but higher than broad-market index funds.

Material risks called out include concentration risk (five stocks ~50% of assets), inherent semiconductor cyclicality and the likelihood of intermittent corrections despite secular AI tailwinds, and potential author bias (disclosure of a personal Nvidia position and differing Motley Fool recommendations). Investors should weigh long-term AI upside against single-stock exposure, sector volatility, and the fund’s concentrated structure when sizing positions.

Key Decisions for Investors

• Consider a tactical allocation to SMH to obtain diversified exposure to AI infrastructure given SMH’s 38% YTD performance and the cited AI market tailwind, but cap position size to limit single-stock and sector concentration risk from Nvidia’s 18.5% weighting
• Compare total cost and exposure versus building a basket of the top holdings (Nvidia, TSMC, Broadcom, Micron, AMD) before buying SMH given the fund’s 0.35% expense ratio and the fact these five stocks comprise roughly half the ETF
• Monitor catalysts and risks highlighted in the article—Broadcom’s OpenAI accelerator deal, DRAM demand (Micron performance), and signs of semiconductor-cycle corrections—and be prepared to trim or hedge the position if concentration-driven volatility or sector weakness emerges