Forget SMH and Its 87% Year: Three Semiconductor Funds Did Much Better
SMH delivered a jaw-dropping year, yet three semiconductor funds quietly left it in the dust by betting on corners of the chip market that most investors completely ignored.
If you own the VanEck Semiconductor ETF (NASDAQ:SMH), the past twelve months delivered a spectacular ride. SMH returned 87.86% over the trailing year and 54.32% year to date, powered by its 17.55% weight in NVIDIA and 9.29% weight in Taiwan Semiconductor. That concentration is exactly why investors bought SMH in the first place, and over five years it has been a winning bet. But three lesser-known semiconductor funds beat SMH on both the one-year and year-to-date scoreboards, and the reason traces back to how 2026’s rally has broadened beyond a handful of mega-caps.
One-Year Scoreboard
Here is how each semiconductor ETF stacks up through August 26, 2026:
| Fund | 1-Year | YTD |
|---|---|---|
| SMH | 87.86% | 54.32% |
| Invesco Semiconductors ETF (NYSEARCA:PSI) | 114.47% | 72.98% |
| Xtrackers Semiconductor Select Equity ETF (NASDAQ:CHPS) | 134.83% | 74.47% |
| First Trust Nasdaq Semiconductor ETF (NASDAQ:FTXL) | 123.50% | 72.93% |
| iShares Semiconductor ETF (NASDAQ:SOXX) | 106.41% | 71.33% |
SOXX is the fund most SMH holders already know. It landed ahead of SMH but behind the three challengers. The story of 2026 is dispersion: names that SMH underweights (Micron, Intel, Marvell, memory, equipment, packaging) drove a larger share of the return than NVIDIA did. The same broadening is showing up outside the chip space too, in the power, cooling, and networking suppliers behind the data-center buildout (we profiled seven of them in a free report here: 7 Stocks Powering the AI Boom (That Aren’t Chipmakers)).
PSI: Tiered Modified Equal Weighting
PSI tracks the Dynamic Semiconductor Intellidex, and its tiered weighting caps single-stock dominance. That structure meant less exposure to NVIDIA and more to mid-cap chip designers and equipment makers when the rally broadened. Over the past year, PSI returned 114.47% versus SMH’s 87.86%. PSI is smaller and less liquid than SMH, but for investors who want semi exposure without a single stock driving nearly a fifth of the fund, that is the design tradeoff working in your favor.
CHPS: Global Reach, Different Top Names
CHPS holds 54 positions spanning U.S., Taiwanese, South Korean, Japanese, European, and Israeli semiconductor firms. Its largest positions are Micron at 6.82%, SK Hynix at 6.80%, and AMD at 6.46%. NVIDIA is just 3.40%. That memory-heavy tilt paid off. CHPS returned 134.83% over the past year, the best in the group. One caveat to note: CHPS launched in 2023, so it has no true five-year record and net assets of only $89.6 million. Therefore, it is a small fund with a short history.
FTXL: The Non-NVIDIA Semiconductor Fund
FTXL is a modified factor-weighted fund with 35 positions, and its top holdings look nothing like SMH’s. FTXL puts 13.28% in Intel, 12.14% in Micron, and 7.91% in Marvell, with NVIDIA at only 4.79%. The fund leaned into the exact names SMH underweighted, and FTXL returned 123.50% over the trailing year. With $2.75 billion in net assets, it has the scale that CHPS lacks.
Five-Year Counterweight
Before switching positions, look at the longer record. Over five years, SMH returned 321.11%, ahead of SOXX at 241.58%, FTXL at 231.07%, and PSI at 216.19%. CHPS lacks a five-year track record. The concentration that hurt SMH in 2026 is the same concentration that made it dominant from 2021 through 2025. This is a 2026 dispersion story about which corners of the chip market led the rally.
Tradeoffs and Tax Reality
SMH’s $77.2 billion in net assets means the tightest spreads and deepest liquidity in the category. PSI, FTXL, and especially CHPS trade thinner. And if you hold SMH in a taxable account with large embedded gains, selling to swap creates a capital gains bill that a modestly higher expected return may not cover.
What This Means for Your Position
If you are near or in retirement and hold SMH for semi exposure, the practical move is a partial trim. Consider keeping a core SMH position and pairing it with FTXL or PSI to reduce the NVIDIA/TSMC single-name risk that now drives roughly a quarter of SMH’s return profile. Direct new contributions into the broader-basket fund rather than selling. If concentration risk keeps you up at night, that split gets you diversified exposure without triggering the tax bill that a full swap would. If you specifically want the NVIDIA thesis to keep working, staying in SMH is the straightforward answer.
Contact [email protected] for any questions or corrections.







