Crowded AI Trade Pushes Retail Toward Semiconductor ETFs

The AI trade has grown crowded enough that retail traders may need to look past the most obvious chip names, according to Financemagnates, which reported that this week’s sell-off hit semiconductor stocks hard as investors reassessed the sustainability of AI spending.

The report cites one figure that captures the speed of the repricing: SMH, the VanEck Semiconductor ETF, fell almost 6% on Monday alone. Financemagnates frames that move as evidence that AI spending expectations, not any single company’s results, are now the dominant driver of these shares.

Also read: S&P 500 Climbs 1.49% as Crude Plunges and Chips Surge

Key facts

• SMH (VanEck Semiconductor ETF) fell almost 6% on Monday, while SOXX (iShares Semiconductor ETF) was cited as a comparable vehicle.
• Goldman estimates that Meta, Microsoft, Amazon and Alphabet could collectively spend $5.3 trillion in capex between 2025 and 2030.
• Morningstar describes Broadcom as Nvidia’s key secondary AI compute vendor as hyperscalers seek greater control over their silicon.
• More than 1,000 ETFs have launched in the US in 2026, including numerous leveraged, inverse and single stock products.
• VanEck’s AI infrastructure research identifies power as an emerging bottleneck and highlights nuclear, utilities and grid infrastructure alongside semiconductors.

Where the money is shifting instead

The report’s core argument for retail traders is that AI spending is not confined to GPUs. It is also flowing through memory, networking, foundries, semiconductor equipment and custom accelerators, and VanEck is cited as saying the semiconductor opportunity has broadened considerably as AI infrastructure develops.

Rather than attempting to identify the eventual winner among Nvidia, AMD, Broadcom and Micron, Financemagnates suggests a semiconductor ETF as the better risk-adjusted route. That still carries valuation and concentration risk, which the SMH move illustrates.

Also read: Fidelity Small-Mid Multifactor ETF Sees $577M Weekly Outflow

Broadcom gets separate attention because it appears at several points in the AI infrastructure chain. Financemagnates points to its exposure to custom AI accelerators, networking and the connectivity required to link very large AI computing clusters, and suggests a basket spanning Nvidia, Broadcom, Taiwan Semiconductor Manufacturing Company and Micron over a single oversized Nvidia position.

Power, monetisation and cybersecurity

The report calls AI power infrastructure potentially the most interesting second-order opportunity, covering electricity generation, nuclear power, transmission, transformers, electrical equipment and data centre construction. Dedicated AI infrastructure ETFs now target energy, grid equipment, data centres and materials rather than AI software alone.

A separate angle is AI monetisation — backing the companies that ultimately earn revenue from AI rather than those that build it. Financemagnates notes the large hyperscalers have existing businesses that can fund AI investment, so standalone AI revenue does not have to be spectacular immediately for them to benefit.

AI cybersecurity is presented as a contrarian idea. During this week’s rotation, CrowdStrike and Palo Alto Networks benefited while Nvidia and other infrastructure names were hit. Financemagnates argues that security, identity, endpoint protection and cloud security providers could gain whether AI development accelerates or becomes more regulated.

The report also warns about the 2026 ETF boom, noting the volume of leveraged, inverse and single stock launches and the danger of holding those products for a multi-year AI thesis. It flags evidence of concern that AI capex is growing faster than AI revenue, and cautions against companies whose investment case rests on the claim that AI will transform the world and therefore their share price should rise.

Why it matters

For retail investors, the practical shift is from stock-picking to structure: exposure to a supply chain rather than a single vendor. The same broadening that makes ETFs attractive also spreads the risk — when capex expectations move, the whole chain reprices together, as Monday showed. Investors who arrived late to Nvidia-heavy portfolios now face a market where the AI theme and AI equity returns have started to separate.

What to watch

Hyperscaler capex guidance is the variable the report repeatedly ties to these moves, alongside whether power constraints become a binding limit on data centre build-out. Any further reassessment of AI spending sustainability would hit semiconductor ETFs first, given their concentration.

This is not financial advice, and the assets discussed here are volatile and uncertain.

Written by

Benjamin

Benjamin Carter covers business, finance, and the stock market for StockPil, focusing on the trends and data that matter to everyday investors.

Source: Finance Magnates