A Wall Street Analyst Just Placed a $3,000 Price Target on Micron: Here Are the Cases For and Against a 200% Surge in the Stock.

DA Davidson analyst Gil Luria shocked the market on Wednesday by setting a $3,000 12-month price target for memory maker Micron (NASDAQ: MU), representing nearly 200% upside from the stock's Tuesday close. The stock has already risen about 1,500% since last April and was up 266% year to date.

Micron is one of the big three memory makers, along with its South Korean counterparts Samsung (OTC: SSNLF) and SK Hynix (NASDAQ: SKHY). The company derives about three-quarters of its revenue from DRAM (dynamic random access memory) and a quarter from NAND flash memory. The company has been riding the memory supercycle, seeing a huge surge in revenue and gross margins as a serious shortage has pushed memory prices sharply higher.

Let's look at the cases for and against this AI stock delivering such a huge upside from here.

The bull case

Luria is now the biggest Micron bull on Wall Street, and his investment thesis is quite simple. He thinks this memory cycle is much different than prior cycles, and the current supply-and-demand imbalance will last significantly longer than expected. This, in turn, will lead the market to rerate the stock to a higher multiple.

The analyst called memory "a lever for better AI performance" and believes there has been a big structural shift in the market. He pointed to Micron's long-term supply contract agreements and the strength of the companies making these deals. Meanwhile, he thinks the trend of memory de-specing -- the use of less memory per AI processor -- could actually lead to more AI processor sales and boost overall memory sales.

I think Luria is correct when he says this memory cycle is different from past ones, as the AI infrastructure build-out has created what clearly looks like a structural tailwind. High-bandwidth memory (HBM) has become an essential part of the AI infrastructure story, as graphics processing units (GPUs) and other AI chips need to be packaged with copious amounts of HBM to rapidly feed them data to process -- or else they will spend too much of their time idle. Meanwhile, memory makers' ability to increase their production capacity is constrained by a host of factors, from a limited ability to secure new EUV (extreme ultraviolet) lithography machines to the time it takes to build new clean rooms. In addition, HBM requires roughly three times the wafer capacity of conventional DRAM, which is straining silicon wafer supply for all types of memory.

Image source: The Motley Fool

The bear case

That said, I think there are some flaws in Luria's argument. First, HBM is not the main direct driver of Micron's huge revenue growth and suddenly robust margins. In fact, the company derives the least amount of its revenue from HBM of any of the big three memory makers. While in Q2, the company had a 24% share of the global DRAM market, it held only an 18% share of the HBM market, compared to 33% for Samsung and 50% for SK Hynix. (Those numbers add to 101% due to rounding.)

As a result, Micron is enjoying more of the piggyback action from the current state of the memory market. With the big three memory makers focusing most of their resources on HBM to keep up with demand, prices for conventional DRAM and NAND have risen at a much quicker pace than HBM prices. This has benefited Micron more than its rivals.

This can be seen in the results of Micron compared to those of HBM leader SK Hynix, which has close ties to Nvidia. Last quarter, Micron's revenue increased nearly fivefold to $55.2 billion, and its gross margin went from 44.7% a year ago to 86.8%. This was largely due to price increases, as its shipments for DRAM rose by just a mid-single-digit percentage sequentially, and NAND shipments grew by 10%. By comparison, SK Hynix saw its revenue climb 257% year over year on sequential shipment growth (high-single-digits for DRAM and mid-teens for NAND), while its gross margin rose from 53% a year ago to 83%.

Why the HBM story is a good one

The simple fact is that if you back into the numbers, less than 10% of Micron's revenue is coming from HBM (roughly 20% market share of what is projected to be a $55 billion HBM market this year). It is much more reliant on conventional DRAM and NAND prices remaining high than on HBM demand, and that's why I wouldn't buy the stock now.

Geoffrey Seiler has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Micron Technology and Nvidia. The Motley Fool has a disclosure policy.