No Dementia for Chip Stocks

Mike Horwath, CFA | Chief Investment Officer • August 20, 2026

Everyone wants to talk about memory. It’s become the old SNL Christopher Walken skit where you need to replace cowbell with ::drumroll:: memory stocks. How did a once commoditized business turn the stock market, and all South Korea, upside down?



In his most recent quarterly letter, Matt Topley wrote “NVDA was the most important stock in the world; now MU has taken the lead. The Mag 7 used to drive the entire market. Now semiconductors are driving returns.” I know I’m getting old when my first reaction to hearing memory storage is a 1GB flash drive. Leaving my boomer-ish qualities aside, let’s dig into what has happened underneath the hood in the technology sector.

Memory chips take center stage

For most of the last decade, memory chips were the semiconductor world's discount rack — a commodity that ran in brutal boom-and-bust cycles driven on slow changes to supply and demand that ebbs and flows with phones and PCs. In 2026 that script broke. 


The culprit is AI: the data centers racing to build it are ravenous for a specialized, high-speed variety of memory (the industry calls it high-bandwidth memory, or HBM), and the handful of companies that make it — Samsung, SK Hynix, and Micron — have redirected the bulk of their most advanced production toward it. The result is a classic supply/demand imbalance: contract prices for standard memory jumped roughly 90% in the first quarter of the year. This graphic does a good job of explaining some of the financial implications. We have stable input costs, soaring prices, and gross margins that even Nvidia would be jealous of.


Unlike past booms, the manufacturers have largely resisted the temptation to flood the market with new capacity — which is a big reason shares of memory makers like Micron and Western Digital have posted triple-digit gains. Look at the stocks of the three main memory players in the US over the last one year, namely Micron (MU), Sandisk (SNDK), and Western Digital (WDC). As Larry David would say, pretty pretty pretty good.

For most of the last decade, memory chips were the semiconductor world's discount rack — a commodity that ran in brutal boom-and-bust cycles driven on slow changes to supply and demand that ebbs and flows with phones and PCs. In 2026 that script broke. 

Investment Ramifications – One Giant Casino

Where this really affects markets and portfolios is through the combination of semiconductor indexes/weights and South Korea.



The clearest place to see the boom is in the semiconductor indexes. Chip funds and ETFs have been leaving the broad market in the dust: at one point this year, the iShares Semiconductor ETF was up roughly 108% on the year while the S&P 500 had climbed about 10% — a tenfold gap. The single name doing the heaviest lifting is Micron, the lone major U.S. memory maker. It has ranked among the very best performers in the entire S&P 500 in 2026, with the stock up around 200% and its market value swelling past $1 trillion. That kind of performance now puts Micron as the 10th largest holding in the S&P 500. How quickly things can change.

The wildest version of this story is playing out in South Korea, home to the two dominant memory makers, Samsung and SK Hynix. The pair loom so large that they and the ETFs tracking them have recently accounted for more than 70% of daily turnover on the entire Korean market. What turned a hot rally into a white-knuckle one was leverage. More than ten single-stock leveraged ETFs tied to Samsung and SK Hynix launched this year — products that aim to deliver twice a stock's daily move — and retail investors hold roughly 90% of them. Assets in these funds ballooned from about $3 billion at launch to over $9 billion within weeks, even as retail margin debt hit a record near $39 billion. The mechanics are unforgiving: because leveraged ETFs must rebalance daily, they become forced buyers into rising markets and forced sellers into falling ones, amplifying moves in both directions. 


It has become so extreme that the exchange has been forced to halt trading with circuit breakers seven times through mid-July, versus none in all of 2025. Regulators have since scrambled to respond, suspending new single-stock leveraged listings and tripling the cash investors must put up to trade the existing ones. It's a vivid reminder that when leverage piles onto an already-concentrated bet, the ride gets violent in both directions. The good news is that emerging markets have contributed to diversified portfolios this year. The bad news is that it hasn’t come in the healthiest of ways.

Final Thoughts: Is this sustainable?

The responsible answer here…no one knows. On one hand, maybe this time is different. Maybe memory companies and their chip demand aren’t as cyclical as they once were. My opinion leans the other way where I tend to think chasing this trade is dangerous. These are big, important, profitable businesses so owning them in a diversified portfolio is a perfectly reasonable, and prudent, thing to do. I believe overconcentrating your portfolio or using the stock market casino that is leveraged ETFs is a dangerous game to play. It’s dangerous in general, let alone doing so with companies who have likely seen their profit margins max out in the extreme demand we’ve seen for their products.

How Well Do You Play Bad?

By Patrick Greenhalgh, Investment Advisor Rep • July 30, 2026
How Well Do You Play Bad? By Patrick Greenhalgh, Investment Advisor Rep There is a delusion every golfer carries to the course: that we, too, could have saved par from that fairway bunker, if only the bounce had gone our way. Golf does that to a person. It is the only game I know that can make you feel like a genius and a fraud inside of about twenty minutes. Here is what I have come to believe after enough rounds to know better. Golf is not really a test of your swing. It is a test of you.

Should We Party Like It’s 1999?

By Mike Horwath, CFA · Chief Investment Officer • July 23, 2026
Should we Party Like It's 1999? By Mike Horwath, CIO  For the inaugural post, I thought an ode to our founder, Matt Topley, was in order. Matt’s quarterly commentary — which takes its title and direction from a song — is a genuinely informative read, and one I’d recommend to anyone interested in markets. The ’90s gave us the buildup to Y2K, the explosion of grunge (Alice in Chains’ 1996 MTV Unplugged set is a personal favorite), and a frenzy of technology IPOs (initial public offerings). With the recent IPO of SpaceX — and the anticipation that Anthropic and OpenAI follow this year — we’ve fielded plenty of questions about what it all means for individual portfolios, and whether it’s worth buying shares if the opportunity arises. There’s a lot of noise out there, so I thought it would be worthwhile to dig into what this means, and to share our perspective here at Lansing Street.