YieldMax launched the Memory and Storage Portfolio Option Income ETF (YRAM) on August 24, 2026, and I think it might be one of the more interesting income vehicles to come along in a while. It’s not because YieldMax funds are exotic anymore, but because of what this one sells options on. Memory is arguably the most profitable, most supply-constrained, and most volatile corner of the entire market right now. When you run an ultra-high income strategy, hot markets are where you want to be. If NAV erosion is a deal breaker, markets that are on fire in a good way will be the only markets where you will be happy with this product as an income generator.
With YieldMax, Implied Volatility=Yield
The first thing to understand about the YieldMax family is that implied volatility is the raw material. Their own prospectus language says securities are selected “primarily based on implied volatility levels,” and YRAM specifically evaluates holdings on liquidity, share price, and IV. The fund is engineered to harvest whatever the options market is charging for fear and greed in memory stocks, and the options market is currently charging a lot.
Where does memory IV sit right now? Micron’s 30-day implied volatility has been reading in the mid-50s in quiet stretches, and Marketchameleon has it around 59, and that is only the 17th percentile of its own one-year range. In other words, “calm” for this sector would be a fire drill anywhere else. In hotter moments this summer, Trefis put Micron’s IV at 82%, and around June earnings CNBC reported 111% — the highest implied volatility in the entire S&P 500.
Historically, memory names tend to trade somewhere in the 45–65% IV range, with spikes toward 80% and beyond around earnings and cycle news. So the base case for premium generation is rich, and the spike case is very rich. If the market cools and IV mean-reverts toward the low end of that band, the yield compresses. That’s a trade-off you’re accepting if you invest in this fund.
Why the Underlying Can Support This for Years
An options-income fund is only as good as the thing underneath it, and this is where the memory thesis gets fun.
The AI buildout has turned memory into a structural shortage rather than a normal cyclical upturn. Data centers are expected to consume more than 70% of high-end memory chip output in 2026, per TrendForce. Micron’s HBM capacity is sold out through 2027. Kioxia committed its entire 2026 NAND output back in January, with hyperscalers asking for supply agreements stretching into 2027 and 2028. SK Hynix’s chairman has said global memory supply will likely run roughly 20% below demand through 2030. And after Micron’s latest results, BofA pushed its supercycle timeline out to the end of 2027, with a scenario extending to 2030.
As for margins, SK Hynix posted a record 72% operating margin in Q1 2026. Micron just printed a quarter with an 85% gross margin and signed sixteen multiyear take-or-pay agreements with binding commitments that management says put a floor under margins “well above our peak quarterly margins in any past cycle.”
These are, at the risk of being blunt, obscene numbers for a historically boom-and-bust commodity business. Whatever happens to any individual AI player, cloud provider, or gadget maker, almost every plausible future involves more compute, and more compute means more DRAM, more HBM, more NAND. Fast-access memory and long-term storage are consumed by every compute paradigm such as, AI training and inference today, cloud storage expansion, edge devices, and whatever comes after that such as space-based infrastructure, autonomous systems, things that don’t have names yet. Memory suppliers profit at the layer below the winners-and-losers game. You don’t need to know which AI company wins; you need to know that somebody, somewhere, is buying RAM and storage hand over fist. Right now, everybody is. I thought things were crazy when my friend and I built my gaming PC in 2024, but now things are crazier and it will get more crazy.
What the Fund Does (And When It Lags)
Mechanically, YRAM holds the underlying memory and storage stocks (its sample portfolio is roughly 40% the Roundhill Memory ETF, plus Micron, SK Hynix, SanDisk, Seagate, Western Digital, and smaller names) and sells call spreads on select holdings to generate premium, paying out weekly.
The call spread structure is deliberate, and YieldMax has been explicit about the design target. Chief strategist Michael Khouw has said the firm uses covered call spreads rather than traditional covered calls specifically to target roughly 80% participation in outsized upside moves. The spread sells a call and buys another at a higher strike, so gains don’t cap out entirely when a stock makes a huge run. On YRAM’s first day of trading, Khouw explained on Schwab Network the trade-off in one line: by selling covered call spreads, “we’re getting a little bit less premium week to week. But we get more participation when the stock really rips.” That behavior shows up in the capture ratio YieldMax itself has described on the order of 80% of the upside and 80% of the downside.
That structure has a specific behavioral fingerprint:
Big rallies: the fund catches roughly 80% of the move, plus the premium. This is where it earns its keep.
Mild bull markets and chop: the caps on the call spreads start to hinder performance. The fund lags the underlying because it’s giving up the upper portion of ordinary-sized rallies. These are the moves that don’t blow through the spread, and this will result in the fund getting capped.
Down markets: The fund’s NAV takes the hit, softened on a total return basis but not eliminated by the premium.
So, YRAM is a vehicle for harvesting rich volatility while staying long in a sector you believe in. It is not a total-return substitute for just owning memory stocks or a standard memory ETF like DRAM or SOXX, and, on NAV performance, it will feel disappointing in a slow grind higher.
The Stress Test: What If We Revert to the 2014–2024 World?
Here’s the question that matters for any options income fund: what happens if the sector goes back to “normal”? A decade ago there was no such thing as a memory fund or even a memory trade. Until very recently, memory was never distinct from semiconductors. Micron traded inside the broad semiconductor indices, the Korean giants moved with the same cycle, and the sector lived and died as one block. It’s only the HBM/AI supercycle that carved memory out as a recognized standalone category with its own funds and its own narrative. That’s why I think SOXX is the right benchmark for the pre-2024 period. It’s not an approximation of memory’s behavior; it was memory’s behavior because memory was part of it.
And from 2014 through 2024, SOXX compounded at 23.4% per year (a +907% cumulative run). A 20–25% average annual return is a good decade for semis.
Now the options math. In that environment, semiconductor IV ran much lower, roughly 30%. A fund like YRAM would have been selling for significantly less premium compared to today’s environment, so call the yield around 30%. Run the two against each other:
Underlying return: ~20–25% per year
Fund distribution: ~30% per year
NAV erosion: ~5–10% per year
So in a good market (20-25% average return per year), the fund’s share price bleeds 5–10% annually while paying you 30%. That may sound bad to some, but it isn’t. This is the part most income investors get wrong.
The reinvestment math is the whole game. If you take the distributions in cash and spend them, your income stream shrinks with the NAV: over those ten years, a $50,000 position paying 30% would see its annual income fall 37–61% depending on the erosion rate. But if you reinvest, the fund’s total return is still positive, roughly 20–25% per year, and that’s what compounds:
Here’s what I want you to see from these hypotheticals. With a 30% yield and 5–10% NAV erosion, you only need to reinvest 17–33% of each distribution to keep your income stream growing. The other 67–83% which is roughly 20–25 percentage points of yield relative to the value of the investment, is genuinely spendable without shrinking next year’s paycheck (this only pertains to the example environments). The income machine compounds as long as the underlying sector’s total return stays positive. And really, it’s considered to be bad financial management to spend all of your paycheck without saving it; there is no reason why collecting dividends would be any different. So, when people complain about having to reinvest in weaker markets, I just don’t get it.
In this case, the failure isn’t the fund. It’s the holder. Morningstar estimated that the average dollar invested in YieldMax funds lost 11.2% per year from December 2022 through July 2025, not because the strategy math failed, but because most investors spent every distribution while NAV eroded underneath them. If you treat a distribution like a bond coupon and never replace the eroded principal, you’ve built a slow-motion return-of-capital machine sans the potential returns that you are spending. Reinvest a third of the paycheck in this environment and the picture inverts completely. That’s the discipline this product demands, and, aside from picking the fund with exceptional underlying assets, it’s the difference between the people who lose money in these funds and the people who don’t.
A Potential Strategic Pairing: YRAM + DRMY
I think YRAM pairs well with the XFUNDS Memory Income ETF (DRMY), which launched a few weeks earlier in July 2026.
DRMY is what we may call a more aggressive sibling. It’s an actively managed basket of memory companies, currently concentrated in Micron, SK Hynix, and Samsung, that primarily holds the underlying stocks, with the ability to build exposure synthetically through options where that’s more efficient (handy for names like the Korean giants). Its income engine is mainly put-spread based: it sells credit put spreads to collect premium, with call spreads available as a secondary tool.
That put-spread tilt is what defines DRMY’s behavior. Selling put spreads doesn’t cap upside at all. Rather, the position just collects premium and sprints like a bat out of Hell when its underlying rips. So, DRMY catches the upswings, full stop. The cost is on the other side: short put spreads add more downside exposure below the strikes, so in a memory selloff, DRMY falls harder than the stocks alone. It’s growth-first, income-second. Though, I will note DRMY retains some of its premium instead of paying out all of it. This will help cushion the NAV to an extent.
Together they form a barbell on the same thesis:
DRMY is your direct engine. Full participation in memory’s upside, amplified exposure to its crashes. The current yield of this is around 35%.
YRAM is the larger paycheck provider at the cost of upside participation. It catches roughly 80% of the massive upward swings (the ones that blow through the spread caps), throws off weekly income at a rate driven by the sector’s fat IV, and its distributions give you a steady stream to redeploy into DRMY, at the bottom, when memory is on sale.
In a euphoric melt-up, both perform and YRAM pays you fat stacks of cash to hold through it. In a grinding bull market, DRMY outpaces YRAM. In a crash, YRAM’s higher premium and yield soften the amplified drawdown you’re taking in DRMY. The pairing doesn’t remove the sector risk, but it lets you size memory exposure like you mean it while still getting paid to wait. So, basically, buy YRAM if you are bullish but memory stock prices are inflated. Buy DRMY in a down market or in a crash. That way, when the sector recovers, DRMY will likely catch more than 100% of the upside and that will help recover your capital more quickly than holding YRAM alone.
What Could Go Wrong?
There is always risk:
IV Lowers And Lackluster Underlying Performance. If memory stocks go sideways-to-down while IV compresses toward 45%, the yield shrinks and the NAV still bleeds.
A classic memory bust. The sector’s history is brutal cycles. If memory stocks fall, DRMY and YRAM would fall sharply along with it. In the 2008 Financial Crisis, SOXX suffered about a 70% drawdown. Even in today’s market, SOXX crashing would likely take the memory stocks down with it.
The behavior risk. As shown above, this fund can punish spend-everything holders. You can spend everything when the underlyings are doing exceptionally well, but in a down market or even in a mild or standard bull market, reinvestment may be required to keep your capital and income stream stable. DRMY can help offset YRAM’s lag in bull markets but it will not pay as much as YRAM in most markets. You have to decide how much income is acceptable and blend the two funds (or not) accordingly.
The Verdict
My opinion is that we’re in the early-to-middle innings of a genuinely structural memory supercycle, supply short of demand for years, margins at levels the industry has never sustained, and demand drivers (AI, cloud, whatever’s next) that all consume memory by definition. YRAM is a way to get paid handsomely for holding that view. Overall, I think YRAM’s yield can be supported for years; just be mindful of the risks and what could happen should the sector cool down in the future.