The Roundhill Memory ETF (DRAM) is an actively managed exchange-traded fund that seeks capital appreciation through concentrated exposure to global memory-chip companies. Rather than tracking a broad semiconductor index, the fund targets producers and suppliers of DRAM, NAND flash, and high-bandwidth memory (HBM)—the storage technologies central to artificial intelligence (AI) infrastructure.
Launched in April 2026 on Cboe BZX, the fund carries an expense ratio of 0.65% and manages assets under management (AUM) of roughly $26 billion. Its portfolio is deliberately narrow, holding approximately two dozen positions, with a heavy emphasis on the three dominant memory manufacturers: Micron Technology (MU), Samsung Electronics, and SK hynix, which collectively account for more than 70% of assets. Additional positions include SanDisk (SNDK), Western Digital (WDC), Seagate Technology (STX), Kioxia, Nanya Technology, and Winbond, with some exposure achieved through total return swaps, including a position in China's CXMT.
This concentration in a single cyclical theme explains the fund's outsized moves in both directions. When memory pricing and AI-driven demand are rising, DRAM tends to outperform broad technology benchmarks; when sentiment shifts, the same focus magnifies the downside. I also checked this using Tickeron’s AI Screener to see how the fund compares to other semiconductor holdings.
Over the trailing 30 days, DRAM fell from roughly $50 to the low-$40 range, a decline of approximately 17%. The move was not uniform: the fund traded in a volatile range through much of the period before an abrupt multi-day drop in early November accelerated the losses.
The quarterly picture is steeper. Measured against a closing level near $57 roughly three months earlier, DRAM has declined approximately 27%. This longer horizon captures a sustained de-rating of memory equities that began as investors questioned how long elevated memory prices—and the resulting profit margins—could persist.
The sharpest catalyst was political. Following U.S. midterm elections, investors reassessed the outlook for AI data-center construction, a primary source of demand for HBM and enterprise DRAM. Calls from newly empowered lawmakers for a pause on advanced AI development, combined with state-level proposals to restrict data-center approvals, raised the prospect of delayed or deferred infrastructure spending. Because memory demand is tightly linked to hyperscaler capital expenditure, the sector sold off sharply in response.
A second driver was the deceleration in memory pricing. Industry data showed general DRAM contract-price increases narrowing from roughly 58–63% quarter over quarter earlier in the year to a projected 10–15% in the fourth quarter. With operating margins near record levels, investors increasingly focused on the rate of change rather than the absolute level of earnings. Weak smartphone demand—global shipments fell about 20% year over year in the third quarter—and cautious guidance from key customers reinforced concerns that consumer-end memory demand was softening.
The fund's largest holdings bore the brunt. Micron Technology (MU), Samsung Electronics, and SK hynix each fell by double digits in early November, and the concentrated portfolio translated those single-stock moves directly into fund-level losses.
The three-month decline reflects a broader rotation out of the AI-infrastructure trade rather than a collapse in current fundamentals. Memory producers continued to report strong, even record, quarterly results, but the market began pricing a longer-term question: whether a wave of new production capacity arriving in 2027 and 2028 would eventually outstrip demand.
Analysts trimmed price targets on leading Korean memory makers, and institutional positioning shifted toward a more cautious stance on semiconductor exposure. Slowing capital-expenditure momentum at major cloud providers and weaker free cash flow (FCF) generation added to doubts about the durability of AI-driven memory demand. Within this environment, DRAM's concentration in the most cyclical names left it more exposed than diversified technology funds.
Several factors are likely to shape DRAM's path in the coming months. First, the trajectory of memory contract prices and any signs that supply growth is catching up with demand will be critical, given the fund's sensitivity to pricing momentum. Second, policy developments around AI data-center construction and power infrastructure could influence capital-expenditure plans at major cloud providers, directly affecting HBM and enterprise memory demand.
Investors should also monitor the performance of the fund's largest holdings, particularly whether HBM order visibility extends further into 2027 and whether consumer-end markets—smartphones and PCs—stabilize. Capital flows into thematic ETFs, analyst revisions, and broader shifts in semiconductor sentiment will also matter. The fund remains a high-conviction, high-volatility vehicle; its concentrated structure means that both recoveries and further drawdowns can occur quickly. From what I see, this is important because any shift in those contract prices could move the fund quickly in either direction.
Tickeron’s AI Screener has been helpful in my process for scanning memory-related names and comparing them to broader industry benchmarks. It lets me quickly filter by technical signals, fundamentals, and volatility to spot where momentum may be shifting without manually reviewing dozens of charts each time.
The information on this webpage is provided for general informational and educational purposes only and is not intended as investment advice, a recommendation to purchase or sell any security, or an offer or solicitation related to investments. It does not consider your personal financial situation, goals, or risk profile, and all investing carries inherent risks, including the possibility of losing your entire investment. For more details, please review our full disclaimer.
DRAM saw its Momentum Indicator move below the 0 level on November 02, 2026. This is an indication that the stock could be shifting in to a new downward move. Traders may want to consider selling the stock or exploring put options. Tickeron's A.I.dvisor looked at 7 similar instances where the indicator turned negative. In 7 of the 7 cases, the stock moved further down in the following days. The odds of a decline are at 90%.
The Moving Average Convergence Divergence Histogram (MACD) for DRAM turned negative on November 04, 2026. This could be a sign that the stock is set to turn lower in the coming weeks. Traders may want to sell the stock or buy put options. Tickeron's A.I.dvisor looked at 4 similar instances when the indicator turned negative. In 4 of the 4 cases the stock turned lower in the days that followed. This puts the odds of success at 90%.
DRAM moved below its 50-day moving average on October 08, 2026 date and that indicates a change from an upward trend to a downward trend.
The 10-day moving average for DRAM crossed bearishly below the 50-day moving average on October 15, 2026. This indicates that the trend has shifted lower and could be considered a sell signal. In 1 of 1 past instances when the 10-day crossed below the 50-day, the stock continued to move higher over the following month. The odds of a continued downward trend are 90%.
Following a 3-day decline, the stock is projected to fall further. Considering past instances where DRAM declined for three days, the price rose further in 50 of 62 cases within the following month. The odds of a continued downward trend are 76%.
The Aroon Indicator for DRAM entered a downward trend on November 12, 2026. This could indicate a strong downward move is ahead for the stock. Traders may want to consider selling the stock or buying put options.
The RSI Oscillator points to a transition from a downward trend to an upward trend -- in cases where DRAM's RSI Indicator exited the oversold zone, 1 of 1 resulted in an increase in price. Tickeron's analysis proposes that the odds of a continued upward trend are 90%.
The Stochastic Oscillator suggests the stock price trend may be in a reversal from a downward trend to an upward trend. 4 of 4 cases where DRAM's Stochastic Oscillator exited the oversold zone resulted in an increase in price. Tickeron's analysis proposes that the odds of a continued upward trend are 90%.
Following a +12.89% 3-day Advance, the price is estimated to grow further. Considering data from situations where DRAM advanced for three days, in 38 of 43 cases, the price rose further within the following month. The odds of a continued upward trend are 88%.
DRAM may jump back above the lower band and head toward the middle band. Traders may consider buying the stock or exploring call options.
The average fundamental analysis ratings, where 1 is best and 100 is worst, are as follows