If You Put $500 a Month into This Tech ETF Since 2016, You Could Retire Tomorrow

Drop $500 into the VanEck Semiconductor ETF (NYSEARCA:SMH) starting in June, 2016, add another $500 every month, and the math pegs the account at $495,000 this week. Total contributions across those ten years come to $60,500. Everything else is what a single-sector ETF did to your money while you ignored it.

The math holds up against the underlying price history. SMH closed at about $26 on June 13, 2016 and printed $609 on June 11, 2026, a 2,269% decade on a price basis. A lump sum at the start would have done dramatically more. The dollar-cost averaging trade-off, smoother entries in exchange for buying a lot of shares at much higher prices in 2023, 2024, and 2025, is the reason $60,500 of contributions compounds to ~$495,000 rather than seven figures. ~$495,000 is enough to retire on with Social Security coming in and a reasonable cost-of-living zip code.

What actually did the work

SMH holds 25 stocks, one sector, and a market-cap weighting that tips the fund into whichever names happen to be winning. The entire AI capex food chain, from chip designers to the foundry casting them to the Dutch lithography monopolist whose machines etch the transistors, lives in one ticker.

Concentration is the mechanism. When semiconductors are the trade, SMH is the cleanest expression of the trade, and the underlying numbers have been ridiculous. Worldwide semiconductor revenue reached $298.5 billion in Q1 2026, up 79.2% year over year, while average selling prices jumped 57%. SMH itself returned 399% over five years and 133% over the trailing twelve months. The 0.35% expense ratio matters here because the compounding window is long, and 35 basis points of annual drag over a decade is the difference between funding the kitchen remodel and funding the kitchen.

What you would need to see for a repeat

The honest forward look is that this run is regime-dependent, and the regime is starting to wobble in places worth watching. The fund is up 69% year to date through June 11, which pulls forward a lot of earnings expectations on its own. SMH then dropped 9% from its June 3 record high, and put-option flow has run above 70% of total volume, which is what hedging looks like when professionals do want to sell outright. Leopold Aschenbrenner reportedly built a roughly $8 billion put position against the chip sector, while Dan Loeb’s Third Point disclosed a long position in SMH itself. Sophisticated investors are taking opposite sides.

The variable that matters most is hyperscaler capital expenditure. Microsoft (NASDAQ:MSFT | MSFT Price Prediction), Alphabet (NASDAQ:GOOG), Meta (NASDAQ:META), Amazon (NASDAQ:AMZN), and Oracle (NYSE:ORCL) fund NVIDIA’s (NASDAQ:NVDA) order book, Broadcom’s custom silicon business, and Micron’s HBM ramp. If their next round of capex guidance tightens, the multiple on the entire fund compresses fast. Memory pricing is the second tell. HBM3E supply has been the bottleneck Micron and the Korean memory makers have been earning outsized margins against, so any sign of inventory builds at the hyperscalers shows up in Micron gross margins within a quarter. The third indicator is ASML’s order book, because EUV equipment bookings are the leading edge of every fab plan two years out.

Stack those together and the read becomes straightforward. The mechanism that turned $500 a month into ~$495,000 is intact, expensive, and visibly contested. Another decade of the same outcome would require the same buyers spending at the same pace into a sector already trading at much richer multiples than it did in 2016. The playbook keeps working until the capex line bends.