BCTK is an actively managed ETF from Baron Capital that owns about 41 technology stocks, heavily concentrated in AI chips, cloud, and semiconductors. It holds about 6% in SpaceX, bought at the June 2026 IPO. You pay 0.75% a year for professional stock picking in the hottest corner of the market.
Scale is 0–50%/yr. The shape of the verdict: a high-octane tech satellite whose base case clears neither hurdle. Only the bull case clears the 12% growth bar.
This is not diversified tech. It is one big view expressed three ways: the chips (Nvidia, Micron, AMD, Broadcom, about a quarter of the fund), the machines that make the chips (Lam Research, TSMC, ASML, about a fifth), and the hyperscalers buying it all (Alphabet, Amazon, Meta, about a seventh). Then a 7% kicker in SpaceX and another 7% in the power chain that feeds the data centers. If you believe the AI buildout runs for years, this is that belief in a single ticker, managed by people who keep rotating toward wherever the bottleneck moves.
About 7% of the fund sits in SpaceX, a position Baron initiated at the June 2026 IPO when the company went public. It is a top-five holding now. This is the active management you are paying for: a concentrated bet no index fund would make on its own, and the only public-market space exposure in the portfolio.
The most interesting 7% of this fund is the part no AI narrative mentions: Bloom Energy's fuel cells, Monolithic Power's power chips, Quanta's grid contracting, Forgent's data-center electrical gear (with a $3 billion backlog), and Fervo's geothermal. Data centers are running into the power grid, not into demand, and these are the companies that unblock it. It is the binding constraint on the whole AI trade, owned here almost by accident.
Two named managers run the book: Michael Lippert, Baron's head of technology research with 24 years at the firm, and Ashim Mehra, on the strategy since day one with about 26 years of experience. They run it the Baron way, hunting for stocks they believe can double in four to five years and sizing positions by conviction, not index weight. The 0.75% annual fee is 25 times what SCHX charges. That fee is worth paying only if the stock picking beats the market by more than the fee, every year, for years. Most active funds do not.
The ETF listed in December 2025, but the strategy is a conversion: it began as the Baron Technology mutual fund in late 2021, run by the same two managers ever since. That means almost five years of real history, including 2022's tech selloff: about 17% a year since inception, a touch behind its tech benchmark. The last three years read better, about 37% a year and ahead of the benchmark. The ETF itself is still small at around $250 million, and small funds can trade with wider spreads.
Our separate research on AI infrastructure ranked the five best ways to own AI demand: the companies that get paid on every token and every training run, no matter which AI model wins. Four of the five sit in this fund: Nvidia, Micron, Meta, and Broadcom, about 20% of the fund combined. So this ETF is quietly a concentrated bet on the AI toll collectors, and Baron picked them independently of our research. That is the bull case for the fee in one paragraph: active managers who landed on the same names our own work did.
This is a satellite, not a core holding: a broad-market beta of 1.64 means it moves about two-thirds faster than the market in both directions. The base case (~10% a year) is decent compounding that clears neither hurdle. It earns a real look on a pullback toward $28, or if Baron's picking proves itself over a full cycle.
55% of the fund sits in 10 names riding the same AI cycle. If AI capex disappoints, there is nowhere to hide inside this portfolio. Concentration is the return engine and the risk, in equal measure.
0.75% a year sounds small until you compound it: over a decade it eats roughly 7% of your wealth versus an index fund. The managers have to be genuinely good, not just lucky, to earn it back.
Launched in late 2025, this fund has only known a friendly market. Nobody knows how the strategy, the SpaceX position, or the fund flows behave when tech falls 30%. You would be finding out with your own money.
At ~$260 million with thin daily volume, the bid-ask spread is wider than a big index fund, so use limit orders. One thing you don't have to worry about: unlike a mutual fund, redemptions don't force the manager to sell holdings, because authorized participants swap ETF shares for the underlying stocks. The real small-fund risk is closure. If it never gathers assets, Baron liquidates it, you get NAV back, and you owe taxes on any gain. A hassle, not a loss of principal.
In a market selloff this falls roughly two-thirds faster than the S&P 500. A 20% market drawdown could mean a 30%+ drawdown here. Satellites need position sizes that survive their own volatility.