The explosive success of the Roundhill Memory ETF (NYSE:DRAM) has turned memory semiconductors into one of the hottest themes in the ETF industry this year.
Riding that momentum, Tuttle Capital Management on Tuesday launched the Tuttle Capital Concentrated Memory Stack ETF (CBOE: HBMX), an actively managed fund targeting companies across the memory semiconductor ecosystem. The fund debuts as investors increasingly look beyond AI chipmakers and toward the memory technologies that have become critical to training and deploying next-generation artificial intelligence models.
The launch reflects a growing belief that memory, not compute power, is emerging as the key bottleneck in AI infrastructure. As demand surges for high-bandwidth memory (HBM), DRAM, advanced packaging, and semiconductor testing technologies, memory-related stocks like Micron Technology Inc (NASDAQ:MU) have become major beneficiaries of AI-driven capital spending.
Tuttle Capital CEO Matthew Tuttle argues that investors need more targeted exposure to these areas, positioning HBMX as a concentrated play on companies helping solve one of AI’s most pressing infrastructure challenges.
DRAM Vs. HBMX: Similar Theme, Different Playbook
HBMX enters a category already dominated by DRAM, which has amassed more than $14 billion in assets under management within two months of launch. While both ETFs are built around the AI-driven memory boom, their portfolio construction and investment mandates differ substantially.
Portfolio Concentration
DRAM holds 15 stocks and takes a highly concentrated approach centered on pure-play memory manufacturers. HBMX currently holds 23 stocks and is expected to maintain between 20 and 35 positions, providing broader exposure across the memory supply chain.
Top Holdings Exposure
Micron Technology is the largest holding in both funds, but the weighting differs sharply. Micron accounts for nearly 29% of DRAM’s portfolio, reflecting the fund’s concentrated nature. In HBMX, Micron carries a weight of approximately 9.14%, reducing single-stock concentration risk while still maintaining significant exposure to one of the industry’s largest memory producers.
Investment Universe
DRAM requires portfolio companies to derive at least 50% of revenue from memory semiconductor manufacturing or development. HBMX uses a lower 25% threshold, allowing it to invest in companies with substantial exposure to memory technologies even if they are not pure-play manufacturers.
As a result, HBMX can invest more broadly across memory chipmakers, advanced packaging firms, outsourced semiconductor assembly and testing (OSAT) providers, substrate manufacturers, materials suppliers, and semiconductor equipment companies involved in memory production.
Active Vs. Rules-Based Management
HBMX is actively managed, giving Tuttle Capital flexibility to adjust holdings based on thematic and fundamental analysis. DRAM follows an index-based methodology with predefined eligibility screens tied to revenue exposure, market capitalization, and liquidity requirements.
Market-Cap Flexibility
DRAM limits its universe through minimum market-cap and trading-volume requirements, resulting in a portfolio dominated by larger companies. HBMX can invest across all market capitalizations, potentially allowing exposure to smaller companies positioned to benefit from growth in the memory ecosystem.
Fees
HBMX charges an expense ratio of 0.95%. While higher than many passive ETFs, the fee reflects the fund’s active management approach and specialized thematic focus. DRAM, meanwhile, carries a lower expense ratio of 0.65% typical of index-based strategies.
For investors seeking exposure to the AI memory boom, the choice may come down to concentration versus breadth. DRAM offers a focused bet on a small group of memory manufacturers, while HBMX seeks to capture opportunities across the broader memory stack, from chip production and HBM to packaging, testing, and supporting infrastructure.
Bottom Line
With DRAM’s rapid asset growth proving investor appetite for the memory trade, HBMX arrives as a differentiated alternative rather than a direct copycat. The fund’s broader mandate and lower concentration levels could appeal to investors seeking exposure to AI’s memory bottleneck without relying as heavily on a handful of dominant chipmakers. As AI infrastructure spending expands, the competition between the two ETFs may become a test of whether the biggest gains lie with memory manufacturers alone or across the wider ecosystem that supports them.
Photo: Shutterstock
Login to comment