Published on: 2026-07-29
RAMZ appeared on 28 July and immediately landed among the market’s biggest movers. It rose because it is designed to move against a group of memory-chip stocks, which had one of their worst sessions in weeks. The unusual part is that the memory business itself was still producing strong demand, higher prices and record profits.
RAMZ is a fund that aims to move twice as far in the opposite direction of the Roundhill Memory ETF, known as DRAM, for one trading day.
DRAM fell 8.89% on RAMZ’s first day, giving the new fund the exact kind of decline it was created to benefit from.
Micron, Samsung Electronics and SK Hynix make up almost 75% of DRAM, so those three companies drive most of RAMZ’s movement.
RAMZ can rise even when the memory-chip industry is doing well because share prices depend on what the market expects, not only what companies deliver.
The fund starts its calculation again after every trading day, so a rough series of falls and rebounds can produce a loss even when DRAM eventually ends lower.
RAMZ is a stock-like fund created by the T-REX brand. It lets the market make a double-speed, backwards bet against companies producing computer memory chips.
It also moves against another fund called DRAM, which owns shares in memory and storage businesses.
The basic relationship is easy to follow.
If DRAM falls 5% in one day, RAMZ aims to rise about 10%. If DRAM rises 5%, RAMZ aims to fall about 10%.
The calculation only covers one trading day. After the market closes, RAMZ starts again from its new price.
That explains the sudden attention. DRAM fell 8.89% on 28 July and closed at $47.77. RAMZ rose because the fund underneath it had a brutal session.
RAMZ closed at $29.60 compared with its official starting value of $25, producing the widely reported 18.4% increase. Its first market trade was already at $28.81, so someone buying at the open experienced a much smaller gain by the close.
RAMZ merely rose because memory-chip shares fell.
DRAM holds companies involved in computer memory and data storage. It owned 17 positions at the end of June, though almost three-quarters of the fund sat in only three companies.
Company |
Share of DRAM |
|---|---|
Micron Technology |
25.81% |
Samsung Electronics |
25.04% |
SK Hynix |
23.99% |
Together, Micron, Samsung and SK Hynix made up 74.84% of DRAM.
When all three companies fall together, DRAM can suffer a large daily loss, and RAMZ can rise sharply, which explains why RAMZ can move so quickly. Smaller companies inside DRAM, including SanDisk, Kioxia, Western Digital and Seagate, have less influence on their own.
Samsung and SK Hynix also trade in South Korea before the US market opens. A major move in Seoul can already be reflected in RAMZ’s opening price by the time trading begins in New York.
Much of the RAMZ trade depends on what happens to three of the world’s largest memory-chip producers.
The memory-chip business was not collapsing when RAMZ launched. Industry forecasts still pointed to higher memory prices during the third quarter. Demand for chips used in AI servers also remained strong.
SK Hynix reported quarterly revenue of KRW79.3 trillion and operating profit of KRW60.5 trillion. Both were records, and operating profit was more than six times higher than a year earlier.
However, the shares still fell. Share prices do not react only to whether profits grew. They react to whether profits were better or worse than the market had already expected. A company can report record earnings and still disappoint when its share price had assumed even faster growth, higher profit margins or a longer-lasting chip shortage.
RAMZ does not require the memory business to enter a recession. It can rise when the market begins lowering its expectations for future chip prices, AI spending or company profits.
RAMZ is a bet that excitement around the memory boom will cool faster than the companies can produce better results.
RAMZ aims to deliver twice the opposite of DRAM’s move for one day. It does not promise twice the opposite result over several days.
A two-day example shows why.
Assume DRAM falls 10% on the first day. RAMZ aims to gain 20%, taking a $100 position to $120.
DRAM then rises 11.11% on the second day. That increase takes DRAM almost back to where it started.
RAMZ aims to lose 22.22% on the second day. The loss is calculated from the new $120 value, reducing the position to about $93.33.
DRAM ends almost unchanged. RAMZ loses roughly 6.7%.
The problem came from the order of the moves. The second-day loss was taken from a larger amount than the first-day gain. A steady fall in DRAM can work well for RAMZ because each gain builds on the previous one. Repeated falls and rebounds can slowly wear the fund down.
RAMZ can be correct about the general direction of memory shares and still produce a poor result. The decline also needs to happen without too many violent recoveries along the way.
The fund charges a 1.50% yearly fee, although large daily price moves will usually have a much greater effect over short periods.
RAM aims to move twice as far in the same direction as DRAM. RAMZ aims to move twice as far in the opposite direction. Both restart their calculations after every trading day.
Not in a fixed basket. RAMZ moves against the DRAM fund, whose value is heavily influenced by those three companies.
RAMZ is built around a one-day target. Keeping it for longer periods means the result will depend on every daily rise and fall, not only where DRAM eventually ends.
Yes. If DRAM rose more than 50% in one session, a fund aiming to deliver twice the opposite move could lose its full value. Such a move would be extreme, though the maths allows it.
DRAM had already fallen about 41% from its June peak before RAMZ launched. That leaves room for further losses, while also increasing the chance of a sharp rebound after so much damage.
The clearest signal is whether DRAM falls below its 28 July low of $46.09. A clean break would extend the decline that powered RAMZ’s debut. A rebound led by Micron, Samsung and SK Hynix would quickly work against it.
