Dave Mazza runs a boutique asset manager in New York with a dozen employees. On April 2, his firm Roundhill Investments launched an exchange-traded fund called DRAM, holding a handful of memory-chip stocks. It had no seed investor, no model-portfolio placement, no wirehouse deal.
Ten trading days later, it has $1.5 billion in assets — a milestone that typically took the most successful thematic ETFs of the past decade between three and eight years to reach.
DRAM is a fast-start outlier in an industry where newcomers tend to struggle to gain assets and attention.
For two decades, the ETF business ran on a simple logic: the biggest managers won because they controlled distribution. Scale begat seed capital, seed capital begat model-portfolio placements and those placements begat the sticky institutional flows that built a $15 trillion ETF industry in the US. A boutique upstart was expected to have a harder time competing in that game.
The game has changed. Momentum-chasing retail traders, armed with commission-free brokerage apps, deep options markets, and the attention economies of Reddit and X, have become a distribution channel powerful enough to push a product to a billion dollars without a single wirehouse meeting.
Those investors are finding smaller firms, not giants — because the giants, for all their thematic lineups, rarely build products as narrow as a memory-only ETF.
"You have a highly concentrated ETF hitting at an important moment in the macro cycle, alongside a much bigger role for retail and options-driven trading," said Strategas Securities' Todd Sohn. "That's a very different form of distribution than what we've seen in the ETF world over the past 10 or 20 years."
Shelf space, compliance and distribution cost roughly the same for a small thematic product as a large one. For Roundhill, which manages $12 billion across some 50 ETFs, a billion-dollar smash hit can be transformative.
Samsung Electronics and SK Hynix, two of the largest memory companies in the world, do not trade on U.S. exchanges, meaning that American traders may have missed out on their triple-digit rallies over the past year.
Those who wanted clean exposure to the companies making the chips that run the AI build-out had few ways to get it. Korean ETFs may dilute the bet with financials and industrials. Broad semiconductor ETFs may bundle memory in with equipment makers.
DRAM holds Samsung, SK Hynix, and Micron, via swap positions. They make up the majority of the fund, at more than 20% each.
"Some of these South Korean names have been on a tear. Many American investors aren't able to own the local names, and now they are able to buy the ETF as a proxy instead," said Mohit Bajaj, managing director of ETFs at WallachBeth Capital. "Timing is everything."
Within two weeks, DRAM ranked in the top 75 US ETFs by options contract volume, said Mazza. Options liquidity is the flywheel, and he argues that once that flywheel takes hold, competitors could struggle to dislodge him with a lower expense ratio.
The pattern is not unique to Roundhill. Defiance, YieldMax, Tidal and Amplify have all built billion-dollar products on narrow themes many of the big players ignored. The pattern is consistent: sharp idea, good ticker, retail discovery, options liquidity, flywheel.
This works until it doesn't. Memory stocks appear stretched. Big tech companies are signing long-term supply deals, and each new generation of AI chips needs much more memory than the last.
"The tricky part will be what happens during the next drawdown in those names — how sticky the assets and trading volume actually are," said Sohn of Strategas.
Mazza acknowledges the risks: the boom-bust history of memory, geopolitical exposure, unhedged Korean won denomination and the possibility that a new compression technique like Google's TurboQuant algorithm reduces memory demand at the edges.
If the trade cracks, the billion dollars that arrived in ten days can leave in five. Boutique-driven flows have no sticky institutional base to cushion the unwind.
Still, "while the stocks have had an incredible rally, the fund is positioned for long-term growth," he says.
(Credit: Adobe Stock)
Copyright 2026 Bloomberg. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.