Invezz

Dan Loeb reveals a stake in a semiconductor ETF: find out more

Dan Loeb reveals a stake in a semiconductor ETF: find out more
Wajeeh Khan
May 18, 2026, 23:51 PM

powered by

Invezz
SMH

Buy SMH. Loeb is adding a concentrated, low-fee way to own the AI compute winners (notably Nvidia and Broadcom) while avoiding single-company operational risk. The setup is a hyperscaler capex wave plus “AI roadkill” dispersion—SMH captures the winners even if weaker chip names get punished. Key risk: AI chip demand growth slows or Nvidia/Broadcom earnings disappoint, and the market de-rates SMH’s ~54x trailing P/E fast.

Key Risk: AI earnings disappoint and the market de-rates SMH’s high valuation quickly.

ASML

Buy ASML. Loeb also increased exposure to chip-equipment leaders, and ASML is the choke point for advanced lithography—so it benefits even when the market rotates within semis. This is the “picks-and-shovels” version of the AI buildout: more leading-edge chip demand forces more equipment spending. Key risk: customers delay/trim leading-edge capex (or export restrictions tighten), cutting ASML order visibility.

Key Risk: Leading-edge capex gets delayed or export limits hit ASML demand.

  • Dan Loeb loaded up on the VanEck Semiconductor ETF in Q1.
  • Here's what might have attracted the billionaire to the SMH fund.
  • VanEck Semiconductor has already rallied about 50% since late March.

Billionaire Dan Loeb’s flagship hedge fund, Third Point, recalibrated its portfolio in the first quarter of 2026, leaning heavily into the AI trade by initiating a new position in the Vaneck Semiconductor ETF (SMH).

Loeb simultaneously increased exposure to individual chip-equipment makers – including ASML, Lam Research, and KLA Corp. Other notable investments he made in Q1 include Broadcom and AI infrastructure name Hut 8.  

This strategic realignment underscores the billionaire’s conviction that the market will sharply bifurcate between structural artificial intelligence beneficiaries and corporate casualties, or “AI roadkill” as he famously framed it.

What made Loeb invest in the VanEck Semiconductor ETF?

Investing in SMH offers Third Point a “concentrated exposure” to structural shifts dominating the 2026 chip landscape.

In Q1, Loeb trimmed his exposure to individual semiconductor names like TSMC, shifting capital toward a vehicle that captures the broader $650 billion (at least) hyperscaler capex wave this year.

The VanEck Semiconductor ETF currently holds a concentrated 17% weight in Nvidia and roughly 7% in Broadcom.

SMH – therefore – aligns the Third Point portfolio with dominant chip designers, while bypassing the specific operational risks of selecting individual stocks.

In short, the VanEck fund tracks a “market-cap-weighted index” heavily skewed towards industry leaders, which perfectly fits Loeb’s broader thesis on the global AI infrastructure buildout.

Significance of Loeb’s investment in the SMH fund

Third Point’s stake is significant because it serves as major institutional validation for the VanEck Semiconductor ETF – underscoring confidence in a fund now central to the AI‑infrastructure trade, with over $40 billion in net assets currently.

Loeb’s “vote of confidence” highlights the fund’s operational efficiency, maintaining a low 0.35% expense ratio, which explains why institutions continue to favor SMH as a “cost‑effective” vehicle for semiconductor exposure.

Simply put, what his investment signals is: hedge funds willingly justify VanEck Semiconductor’s trailing price-to-earnings (P/E) multiple of about 54x due to rapid growth across the enterprise AI hardware supply chain.

Note that the SMH has delivered impressive returns this year, currently up a whopping 50% versus its year-to-date low in late March.

Is it worth investing in VanEck Semiconductor today?

While the SMH find has already delivered a solid rally this year, the question for investors now is whether it still offers attractive risk‑reward after a 50% rebound from its March lows.

The ETF’s appeal rests on its concentrated exposure to companies driving the global AI‑compute buildout – Nvidia, Broadcom, and the broader chip‑equipment complex.

But its trailing P/E near 54x reflects how aggressively the market is already pricing in that growth.

For investors, the decision ultimately hinges on time horizon: SMH remains a clean way to express the AI‑infrastructure theme, but its elevated valuation means future returns may be more sensitive to earnings execution and macro conditions.