Hi friends, I’d love to bring $MU back to the board’s attention. $MU was up 239% in 2025, but I believed it could still generate great return in 2026 and therefore just started a starter position. While I’m still doing research, here’re some high level reasons of why I decided to invest in it.
Jensen Huang has confirmed in the 2026 CES that memory is one of the biggest bottleneck for AI right now as the number of tokens today’s models have explodes. Micron ($MU) is one of the three major players in the DRAM market. While Micron ranks third in terms of market shares, it’s the only US company. (The other two are SK Hynix and Samsung)
While demand continues to surge, the entire DRAM market, especially HBM, will be supply constraint for the foreseeable future (at least for the entire 2026). This gives the chip providers huge pricing power. According to a recent news from a Korean media, Samsung and SK Hynix are discussing potentially increasing price of server memory by 70%.
The advantage can already be seen in $MU’s earning results. In the most recent quarter, gross margin surged to 56.04%, from 44.67% the previous quarter, and from 38.44% year ago. Net margin surged to 38.41% from 21.47% the year ago quarter.
YoY revenue growth has been on a up-trend. 36.56% → 46% → 56.65% → 132.22% (!!! next quarter, based on the midpoint of guidance)
Gross margin is guided to be 68% for the next quarter, another huge improvement. non-GAAP EPS is guided to be a record $8.42, compared to $4.78 in the recent quarter.
$MU has two major revenue streams. In the recent quarter, total revenue was $13.64B, among which $10.8B was DRAM revenue (+69% YoY) and $2.7B was NAND revenue (+22% YoY). Among DRAM revenue, HBM was still a small part and was the major driver of expanding gross margins. The strong guidance of next quarter is partially based on the expected ramp-up of HBM4.
Despite the significant growth of share price, $MU’s forward PE only sits at 10.83 based on consensus estimation today.
The risks?
While $MU’s forward PE is significantly lower than many peers in the chip industry (e.g. AMD, Intel, etc), memory was historically a cyclical industry and the DRAM price could drop significantly in the future if the supply constraint resolves. Jensen did mention that memory will be part of the architecture design of AI infra though, so the cyclical nature may change as the industry evolves.
Despite the exploding demand, $MU’s growth will still depend on its execution to solve supply chain challenges. Their future earnings may come below expectation if the execution fails.
Yes, I think this is a Saul type stock with metrics and growth like the ones you’ve posted above. Ive been thinking about getting into MU for the past few months. If I had I would have enjoyed its recent continued parabolic rise.
It’s actually looking cheaper on a valuation basis than it was earlier last year, in spite of the SP increase. It’s Fwd PEG is currently 0.21.
My problem is I don’t want to sell any of my other holdings to buy it, and like always, I’m fully invested. Unless I halved my Nvidia holdings to buy it, but then with Vera Rubin this year I’m reluctant…. Decisions, decisions…
But thanks for bringing MU back to the board. I am considering it carefully again.
Jonathan
Edit, 3 hours later… I decided this was the prompt I needed, so I have just this moment sold half of my Nvidia and have used the proceeds to purchase Micron.
Yep MU has just reached a top 5 holding position for me. I used to hold Micron during a previous couple of upswings and sat out the last cycle but re-entered between $86-100 between Nov 2024 - Jan 2025 at roughly at or slightly below my previous exit levels a cycle earlier.
I bought back in and continue to hold for 2 reasons…
The HBM story attached to AI as summarised by the OP and seeing something between another AI infra opportunity play but with an established reasonable priced player and the potential for a repeat of the Nvidia inflection moment. Clearly HBM attached to Nvidia chips in AI inference seemed to be a runaway demand opportunity and their 5 year forecasts for HBM was incredible. It had helped that at the time one of the HBM competitors had taken a misstep and was struggling with getting their HBM3/HBM4 issues resolved, certified and to market.
As a holder of Pure Storage, I was also familiar with the crossover scenario that the world of DC memory storage faces where effectively the performance v TCO of flash is overtaking that of spinning disks and effectively on a go forwards basis all new memory storage installations as well as end of life replacements are going to see flash replace spinning disks - this is only being accelerated by the hyperscaler demands for AI inference which should see a massive uplift in demand for NAND.
Throw in a couple of other variables including: i) this is a structurally conducive supply side industry set up with only 4 main players with massive Capex barriers to entry then you have relatively strong/safe market opportunity (leaving aside the cyclical nature of memory) and ii) Micron is the ONLY US memory supplier so the only one to receive any trade/tariff protectionism benefit under a Trump 2.0 system then you have a solid investment opportunity.
One other opportunity I don’t see discussed at all in terms of potential memory deployment - similar to embedded memory in autos with FSD cars; is the explosion of potential smart devices needing processing and memory attached - I’m thinking robots and drones in particular.
Anyhow whilst certainly under normal conditions buying a cyclical stock with a low P/E tends to be the top of the cycle and the fastest way to lose your shirt but this feels like an Nvidia type inflection point and a super cycle right now. (We forget that Nvidia had also been operating in cyclical industry conditions previously).
The guidance Micron gave was incredible across all three areas they track. Projecting the following for the next quarter,
Revenue 18.7B +132% yoy, +38% qoq
Gross margin 67%, +3000 bps yoy, +1100 bps qoq
GAAP EPS 8.19 +481% yoy, +78% qoq
Some other stand outs from their latest call,
Demand well outpacing their supply
CapEx increase from 18B to 20B next year supporting New York, Idaho, India and Hiroshima buildouts (Already have production in Virginia, China, Singapore, Taiwan, Japan)
Higher pricing, “robust bit shipments”
Paying down debt, buying back shares
HBM4 Industry leading product with highest performance over 11 gigabits per second
“Significant upward revision on ASIC XPU volume shipments next year”, Google TPU, AWS Trainium run on HBM memory
Tight supply environment “gap between the demand and supply for all of DRAM, including HBM is really highest that we have ever seen”
“Very, very tight supply environment”
“We are not really able to meet the demand for customers across any segment”
Medium term: “We are only able to meet 2/3 of the demand from our several key customers”
HBM sold out for calendar year 2026
DRAM pricing +20% qoq, “NAND was up pretty nicely too”
“We do believe that margins can go up” (Curious if they mean even further than their 67% GAAP margin guidance")
I started a small position again in Micron after having a look over this report. The risk reward looks compelling to me as it’s guaranteed next quarter’s revenue numbers will be spectacular. Of course it will depend a lot of where they guide to next.
Their supply being sold out I see as positive. There is simply no way to scale up to be able to match the demand for their product, even with spending 20B on CapEx next year. It seems like whatever they are able to produce will be easy to sell and at a price of their choosing.
100% agree WPR. Honestly - for anyone who believes in the higher for longer scenario with Nvidia, one can only come to the same conclusion with Micron, except: you are getting in earlier in the inflection story, supported by an even better trailing and forward valuation and benefiting from an early warning system in Nvidia as the secondary order beneficiary with HBM and DC memory.
It’s a compelling entry opportunity especially if you pay no attention to the share price over the last year.
This is a good point here on the market cap comparison. Next quarter, Micron will have roughly 1/3rd of the revenue and similar margins to Nvidia. I do think Micron can triple or more in value over the next 1-2 years which is one of main criteria I am looking for. It seems like a layup that this company will become a 1T+ market cap company.
It may be tempting to think there could be a repeat of 2023 where Micron’s revenue halved going from 8.7B down to multiple quarters below 4B. The situation for the RAM market was dramatically different back then though. The three main producers created an oversupply as they anticipated upgrading cell phone cycles happening faster. My understanding is that people were upgrading their phones less and phones were being equipped with less memory than anticipated. Micron along with the other memory producers had excess inventory to process through.
The situation now is that Micron has a competitive edge with the best speeds on their HBM and the best energy efficiency. Not only do Nvidia based systems need HBM, but the hyperscalers ASIC build outs are centered around HBM for their memory component.
There are countless reasons for an officer of a company to sell stock in that company. But there’s only on reason for that same officer to buy stock in the company - Yes, very bullish.
Does anyone have insights into long-term prospects for Micron and other hardware vendors benefiting from the infrastructure demand to support the growing AI sector?
My initial impression is that many of these companies (not Micron in particular) look like amazing investments at the moment purely because they exist, not because they are particularly good companies for investment outside the current AI infrastructure boom. Even in the best case scenario of AI use continuing to grow, at some point hardware infrastructure will catch up. When that happens, is there a growth driver to companies such as Micron to keep their valuation from crashing down?
I am reminded of two other investment markets: The pandemic and the migration to cloud applications / SaaS. In both cases, it was trivial to find companies growing fast at the moment. Far more challenging was identifying the companies with long-term potential.
While I have nothing against being an opportunist in my investments, those companies come with significant risk in determining the proper time to exit.
MU, and quite a few AI infrastructure related companies, are in the sweet spot of being Saul’s Investing Discussion type companies. They are exhibiting accelerating growth, increased gross margins, increased free cash flow all driven by a market that has accelerting growth. Saul’s Investing Discussion companies rarely pass muster beyond a 3-year window because growth starts to decelerate leading to smaller multiples leading to a lower stock price.
The long term potential companies? I hate sounding like a shill, but TMF Rule Breakers has been pretty dang good at identifying these types of companies.
I submit that “determining the proper time to exit” is always important. Only two of these (CRWD and NET) have gone on to be long term winners. But even more important – there was an almost unfathomable opportunity cost to holding even the long term winners here in 2022:
Lastly, while both are “long term winners” (at present) and are beating the market since the end of 2020, neither has tripled. You could have done much better elsewhere.
I know it’s a bit anathema to say the long term doesn’t matter. But Saul-style investing has a lot to do with what’s growing fastest NOW, as @Buffjan2 points out. Sure seems there’s huge alpha to be had using this method (just watch @wpr101), and frankly in simply ignoring the long term. There’s always been a speculative nature to what we do here, but it’s a speculation based on how the market usually treats a particular type of growth company.
But again, there’s alpha to be found here, so…why? Why can we ignore the long run? One, because we have to accept there will be periods like 2022 where nothing we do works, anyway. (FWIW, this is why I don’t hold 15% or 20% positions anymore!). But also, because no one can predict what things will be like in 5 years – truly any company that could 5x or 10x also has the potential to fail. (GOOG and AMZN will be fine in 5 years, but they won’t triple.)
So Saul’s “Modified buy and hold” just works better.
Bear, I agree with you completely. But let me rephrase my question as I clearly was not clear (my apologies).
How likely are these hardware companies to have their stock price fall off a cliff?
There are many scenarios to what can happen to stock price as a company evolves, but take three scenarios for the sake of discussion:
2022 type event where the entire market drops
Growth slows (but continues at a lower rate)
Growth stops (possibly suddenly)
Each of these requires a different exit strategy.
A 2022 type event that effects the entire market is unpredictable and, in the grander scheme of things, not worth worrying about. There are ways to mitigate that risk, but it has nothing to do with any individual company.
Growth slowing is an inevitability of all companies. In a well run company with a good place in its market, watching the financials can give clear warning that growth numbers are changing. This covers many of the stocks we have talked about here. There is time to look at quarterly reports and decide what action to take. I am not too worried about this scenario as there are often warning signs even before growth slows significantly. Many of these companies remain valid investments, just not for the Saul-style approach we look at here.
Growth stopping, potentially suddenly, is somewhat common, but an indication of either a flaw in the company or its position in the market. I consider this scenario a far greater risk than growth simply slowing simply because the price drops can be huge and unexpected with no recovery. Far more challenging to determine an appropriate exit strategy.
What I am debating with Micron and other hardware vendors rising dramatically at the moment: Are these companies going to see growth slowing or suddenly stopping? Will MU suddenly drop back to the $50 - $100 price range that has been its norm for most of the past decade? Will it drop lower than that if there is a massive oversupply in its manufacturing capability after the current AI fanaticism fades? Could that drop be sudden and rapid? Is there an enduring growth driver beyond the current push for rapid AI infrastructure creation? Is there any reason to think Micron might be a long-term market winner with enduring (slow) growth, or is this a one-shot anomaly?
The above types of questions are the ones I am asking myself to help identify the risks involved with investing in companies such as MU, and identifying appropriate exit strategies. Some of the above questions are from a “devil’s advocate” perspective and highly unlikely scenarios, but still important to ask in order to understand the potential risks.
I am not looking to determine if these companies have long-term potential as a Saul-type stock (or a TMF Rule Breakers stock). I am looking to determine what, if any, growth potential these companies have after this unique situation passes. To help determine both expectations and exit strategies.
At least for my interpretation of Saul-style investing, I am far more successful overall when I have somewhat of an understanding of more long-term potential in addition to current growth trajectories.
Examples like CRWD and NET apply here, and I agree. When growth goes gradually from 50% or 60% to 40% and then 30% and then 20%, unless the stock was way overvalued, you’re probably dealing with opportunity cost more than anything.
These days, I would put APP and RBRK and MELI and others in this group. Maybe even NVDA. Unfortunately, those are probably not where the 3x or 5x returns will come from (although with RBRK I believe there’s hope to 3x if ARR can accelerate).
Agreed…this would be the AEHR, ELF, ZM, UPST, PTON scenario. Strangely, some of those did so well befor they crashed back to earth, that it was possible to get out with gains. You’re right – this is hard (Saul getting out of Zoom with amazing timing was one of his finest moments) and you need to know whe you’re in this territory. I think ALAB and CRDO, MU, and most hardware companies are.
These are good questions…some people think the AI infrastructure push is enough, and that it will last for many years, and that Micron will continue to get stronger throughout. The reason I haven’t bought MU is that a large part of their massive growth is that they’ve been able to raise prices massively. (Grok: “Industry-wide DRAM contract prices jumped 45-50% in some periods of late 2025.”) I think those are going higher short term, but I don’t think anyone believes it will last forever. That’s just not how RAM works.
Here I’ll shut up and see if anyone else wants to jump in on Micron pricing issue specifically, or on how you think about exiting these hardware companies at the right time. It’s hard, and I think Saul showed (with Zoom) that sometimes you have to look ahead a bit.
