Generated July 2026 from current fund data.
Overview
HBMX and KMEM are both memory-semiconductor-focused equity ETFs launched in mid-2026, but they pursue the exposure differently. HBMX is an actively managed, concentrated fund covering the full memory stack—DRAM, NAND, HBM producers, plus the packaging, testing, and equipment vendors behind AI infrastructure. KMEM takes a simpler basket approach, targeting memory chip producers directly. The key distinction: HBMX casts a wider net across the supply chain; KMEM narrows to the semiconductor manufacturers themselves.
How they differ
HBMX's strategy is broader and more thematic. It includes not just memory chip makers but also the equipment manufacturers, advanced packaging specialists, and test providers that support AI-driven memory infrastructure—essentially a full-stack bet on the memory ecosystem. KMEM, by contrast, focuses directly on memory semiconductor stock producers without explicitly including ancillary vendors.
Cost differs materially: KMEM charges 0.65% versus HBMX's 0.95%, a 30-basis-point spread that compounds over time. HBMX distributes annually; KMEM's distribution frequency is not specified in the data. HBMX is actively managed, giving the fund manager discretion over holdings and positioning; KMEM is described as a basket strategy, suggesting a more passive or rule-based approach. Both are extremely new funds (inception within weeks of each other in mid-2026), so neither has a track record.
Who each is best for
HBMX: Fits investors who believe the memory semiconductor tailwind extends beyond chipmakers themselves—that equipment vendors, packaging specialists, and test providers represent equally compelling long-term opportunities in the AI infrastructure buildout, and who are willing to pay for active management and accept a more concentrated portfolio to pursue that thesis.
KMEM: Fits investors seeking direct, cost-efficient exposure to memory chip production capacity without paying for active management or supply-chain expansion; designed for those who view the memory producer stocks as the core holding and prefer a simpler, lower-cost implementation.
Key risks to know
- Concentration and single-theme risk: Both funds bet heavily on the memory semiconductor sector and the AI infrastructure cycle supporting it. A slowdown in AI capex, a memory chip glut, or cyclical pricing pressure in DRAM or NAND would affect both simultaneously, with limited diversification offset.
- Extreme newness and limited operating history: Both funds launched in June 2026 and have no meaningful performance track record, making it impossible to assess how holdings, fee structures, or strategies behave through market stress or sector cycles. Trading liquidity and fund viability are unproven.
- HBMX's active-management risk: Concentrated, actively managed equity funds depend heavily on manager skill and decision-making. HBMX's relatively high 0.95% expense ratio reflects that overhead; underperformance relative to a simpler memory basket (like KMEM) would compound the drag.
- KMEM's strategy opaqueness: The "derivative overlay" description and "basket" methodology lack specificity in the data; the mechanics of KMEM's exposure—whether it holds stocks directly, uses swaps, or employs other derivatives—are unclear and carry tail risks (counterparty, leverage, path dependency) that aren't detailed here.
Bottom line
If you believe the memory tailwind encompasses the full supply chain—equipment makers, packagers, testers—and value active stock selection, HBMX's broader mandate may justify its higher fee. If you want straightforward access to memory producers at lower cost and without active-management overhead, KMEM's simpler structure and 30-basis-point fee advantage stand out. Both are infant funds with no proven track record, so neither has demonstrated whether its chosen strategy delivers in real market conditions.
AI-generated analysis for educational purposes only. Verify important details independently; past performance does not guarantee future results.
