Why do I still have extreme faith in Micron $MU?

Micron is currently one of the companies in my portfolio that I really believe in a lot. And despite the significant rise in the stock price recently, I still think its potential is far from exhausted.

Personally, I would not be surprised at all if the stock managed to move another 50–60% higher in the coming years. Of course, that does not mean the path upward will be straightforward, because Micron is a very cyclical company and the cyclicality of the memory market is one of the main risks I count on with my investment.

But what has surprised me about Micron in recent months is its strength. The geopolitical situation remains tense, the macroeconomic environment is definitely not ideal, and yet the stock is holding near historic highs. In my opinion, that mainly reflects how fundamentally the situation in the memory market is changing.

Memory is becoming one of the key parts of AI infrastructure

Artificial intelligence does not only need powerful processors from Nvidia $NVDA or other chip makers. It also needs enormous amounts of fast memory.

As AI models, data centers, and inference demands grow, the amount of memory needed in individual servers grows as well. Micron itself pointed out this year that the memory volume per server has roughly doubled over the last three years and the demands on memory capacity and throughput continue to grow.

That is exactly why I believe that DRAM, HBM, and memory chips in general will be one of the most important parts of further AI infrastructure buildout.

And the problem is that their supply cannot be increased overnight.

Micron has previously stated that DRAM and NAND supply will not be able to satisfy all demand, and according to the company's latest outlook, tight market conditions are expected to continue even beyond 2027.

The memory shortage will not be short-term in my view

The market is increasingly talking about the current memory shortage possibly being much longer than originally expected.

Some estimates already work with the idea that supply-demand problems may extend all the way toward 2030. SK Hynix $SKHY also warns that 2027 could be one of the toughest years in terms of available production capacity.

And that is exactly where my investment thesis for Micron strengthened significantly.

I do not think we are just watching another short memory cycle. In my opinion, AI structurally increases the amount of memory the world will need.

Long-term contracts are a very important signal to me!

Another factor I watch is the behavior of the manufacturers themselves and their customers.

SK Hynix, for example, announced that it has already signed long-term agreements with roughly ten customers and is negotiating multi-year contracts with other large clients. The company is also investing tens of trillions of won in new production capacity precisely to handle the expected demand for AI memory.

We see a similar development directly at Micron.

The company has signed multi-year agreements with customers, with a typical contract running until the end of 2030. Already signed contracts represent approximately 20% of Micron's DRAM volume and a third of its NAND volume in that period.

That is an important signal to me.

Large customers would not, in my opinion, have such motivation to secure supplies years in advance if they expected the memory market to quickly return to a significant oversupply.

Risks have certainly not disappeared

At the same time, I do not want to present Micron as a risk-free investment.

The memory industry has historically been extremely cyclical. Periods of shortage, high prices, and huge profits can be replaced by new production capacity, falling prices, and a sharp deterioration in profitability.

Moreover, increasing competition is emerging, for example from China. CXMT is already expanding production of a new generation of DRAM and is trying to close the technology gap with the world's largest producers.

That is why I do not expect the current extraordinarily favorable conditions at Micron to last forever.

My thesis is more that this positive memory cycle could be significantly longer and stronger than previous cycles thanks to AI.

And if data center buildout continues at the pace we see today, I think the market still underestimates how much memory the world will actually need over the next few years.

That is why Micron remains one of the companies I believe in the most in my portfolio. Although the stock is trading near historic highs, I still see room for another 50–60% growth in the long term.

Of course, this is only my view and investment thesis, not investment advice.

💎 And if you are interested in my more detailed analyses, specific buys and sells, and other content I do not share publicly, you can join my membership here on Bulios. You will find many more benefits and content for those who want to follow my portfolio and investment decisions more in depth.