关于DRAM,感谢网友提及,简单查了一下

来源: 2026-04-14 13:13:31 [博客] [旧帖] [给我悄悄话] 本文已被阅读:

这支roundhillETF有点吸眼球,因为我有一只Roundhill Generative AI & Technology ETF (CHAT),因该是同一家公司的产品。然后看到日涨跌可以大到5%,这种少有的标签在ETF里唯roundhill所有。兴许是CHAT比较成功,现在就又出了一个DRAM。

 

下面的文字是抄来的:

 

The Roundhill Memory ETF (DRAM) is a newly launched, "pure-play" fund that debuted on April 2, 2026. It is designed to capture the "supercycle" in memory semiconductors, specifically driven by the massive hardware requirements of AI data centers.

Core Strategy & Market Context

The fund targets companies that derive at least 50% of their revenue from memory products like DRAM (Dynamic Random-Access Memory), HBM (High Bandwidth Memory), and NAND flash.

  • The "RAMmageddon" Narrative: The ETF's launch coincides with a significant global memory shortage projected to last through 2028. This is fueled by manufacturers shifting capacity to HBM to meet AI demand, which has restricted the supply of standard DRAM and sent contract prices surging.

  • Performance: In its first two weeks of trading, the fund has shown high volatility and growth, jumping over 18% in a single day recently following strong earnings guidance from Samsung.


Top Holdings

The portfolio is highly concentrated, with the "Big Three" memory makers accounting for roughly 72% of the total weight:

Company Ticker Weight (Approx.)
SK Hynix 000660 (KS) 24.7%
Samsung Electronics 005930 (KS) 24.1%
Micron Technology MU 23.9%
Kioxia Holdings 285A (JP) 5.9%
SanDisk SNDK 5.0%
Seagate Technology STX 4.7%
Western Digital WDC 4.7%

Key Statistics & Risks

  • Expense Ratio: 0.65% (Actively managed).

  • Structure: It uses Total Return Swaps for some exposure, meaning the fund's risk profile is more complex than a standard "physical" stock-holding ETF.

  • Concentration Risk: Because just three companies dominate the fund, any regulatory or earnings setbacks for Micron, Samsung, or SK Hynix will cause significant swings in the ETF's price.

  • Geopolitical Factor: SK Hynix has recently applied for a U.S. listing (ADR). Analysts suggest this might impact Micron’s status as the primary U.S.-listed memory play.

Note: Analysts generally view this as a "satellite" investment—meaning it's intended to be a small, high-growth slice of a portfolio rather than a core holding, due to its extreme sector concentration.

Are you looking at this as a tactical play for the AI infrastructure build-out, or are you more interested in the specific technical differences between these memory makers?