ETFs to Buy as NVIDIA Marches Toward $6 Trillion Market Cap

NVIDIA reached a fresh intraday high of $239.86, putting its market capitalization at an estimated $5.78 trillion, after raising buyback authorization by $150 billion to $235 billion through fiscal 2028. Fiscal Q2 2027 revenue rose 106% year over year to $96.2 billion, while management expects fiscal 2028 revenue growth of about 70%, constrained by supply. The article highlights diversified ETF options with NVIDIA weights of 15.52%–21.98%; the cited funds gained 36%–76% year to date.

Analysis

The ETF framing overstates diversification: SHOC and SMH remain concentrated semiconductor-beta vehicles, while IYW and XLK still carry substantial exposure to the same mega-cap AI trade. They reduce single-name risk, not the risk that AI infrastructure spending, expectations, or multiples reset together. Recent strength also raises the bar for incremental upside; the key question is whether customer capex converts into sustained deployments and returns, not whether demand is currently strong. The announced repurchase capacity is not equivalent to executed buying and should not be treated as a near-term price floor.

Over the next 1–3 months, monitor hyperscaler capex commentary, NVIDIA supply/ramp delivery, and evidence that inference and software usage create recurring economics beyond initial hardware sales. Over 6–18 months, custom silicon from Amazon and Alphabet could shift some workloads away from NVIDIA while supporting Broadcom’s design role; that substitution is gradual and does not automatically eliminate demand for GPUs. TSMC, ASML, and Micron may benefit from continued infrastructure buildout, but their exposure is not a hedge against a broad semiconductor drawdown. The contrarian risk is that supply constraints are read solely as upside: they can defer revenue and expose execution or customer-concentration risk. A thesis weakening signal would be capex moderation, missed delivery/ramp milestones, or lower forward growth expectations; sustained deployment and recurring software monetization would falsify the bearish scenario.

Key Decisions for Investors

• Do not chase SHOC or SMH solely on NVIDIA momentum. For new AI exposure where lower single-name concentration is the priority, favor XLK over semiconductor-heavy funds, while recognizing its meaningful NVIDIA and mega-cap exposure.
• For existing NVDA holders seeking to reduce idiosyncratic risk without exiting the theme, trim selectively into strength and rotate part of exposure into a broader technology basket; avoid treating ETF ownership as downside protection from a sector-wide valuation reset.
• Watch AMZN and GOOG custom-chip adoption and Broadcom’s related design activity as a 6–18 month relative-winner test. Consider a measured long AVGO versus short NVDA only if evidence of workload substitution or NVIDIA delivery slippage emerges; otherwise the pair risks fading durable GPU demand.
• Verify actual repurchase execution, hyperscaler capex guidance, and NVIDIA delivery trends before adding. Reassess the bullish thesis if customer spending slows or growth expectations are revised down; reassess the cautious stance if recurring software revenue becomes material and deployments broaden beyond initial infrastructure purchases.