Should You Buy Micron Before Sept. 30?

One of the biggest beneficiaries of the artificial intelligence boom has been Micron Technology (NASDAQ: MU). Among the "big three" global players that can produce advanced DRAM memory, Micron's financial profile has improved significantly, thanks to the insatiable memory appetite of AI systems. And as the AI era moves from the generative to the agentic phase, these demands seem only set to increase.

Of course, that has also been reflected in Micron's stock price, which has delivered a 1,300% total return over the past five years -- that's a fourteen-bagger .

Missed AI's "Act 1"? Act 2 Could Be 15x Bigger. Most investors think they missed the AI boat because they didn't buy Nvidia in 2005. But according to our analysts, we're only at the end of "Act 1"—the R&D phase. "Act 2" is the global rollout. Continue »

After that big a run, is this high-flying stock still a buy? Micron reports its fiscal fourth-quarter earnings on Sept. 30. Here's what to look out for.

Today, Micron currently trades at 22.9 times trailing earnings but just 6.6 times the earnings estimates for the upcoming fiscal year , which ends in August of 2027. The reason for the low forward P/E ratio is that Micron has traditionally been a highly cyclical business, and investors fear that next year's huge earnings estimates will mark a peak in this cycle.

There are a few reasons for this. First, memory cycles typically only last two years or so, and the monster up-cycle for memory really began about a year ago, making next year's earnings the culmination of the past two-year cycle. So, investors scarred by past memory cycles are wary of bidding Micron higher.

Micron and its major competitors have also invested heavily in new memory plants slated to come online in 2028, increasing supply. Micron, for its own part, has six new fab projects in development in addition to expansion plans at its current facilities. Its massive new Idaho fab will come online next year and begin producing DRAM wafers in 2028. Micron's main Korean competitors are also building new memory fabs, which will begin coming online in 2028 and 2029.

That incoming supply is one concern, especially if AI companies figure out a way to use less memory going forward. Already, it appears that Nvidia (NASDAQ: NVDA) has "down-spec'ed" the memory requirements for its upcoming Rubin systems to 4-hi high-bandwidth memory (HBM), down from the originally planned 8-hi or 12-hi. This is because HBM is becoming very expensive, and higher-stack sizes may become uneconomical for many workloads.