On Wednesday, memory-maker Micron Technology reported another beat-and-raise quarter, with 379% sales growth and an astounding 87% gross margin. The stock rose 3%.

Come Thursday morning, my inbox was full of questions that all boiled down to, "What gives with this stock?"

Micron stock is trading at only 6.4 times its projected earnings this fiscal year, which began a month ago, well below either the S&P 500 or PHLX/semiconductor, or SOX, indexes. It is particularly strange when held up against the Wall Street analyst consensus estimate for earnings growth of 126% in fiscal 2027.

More than any other tech stock, trading in Micron is about psychology, and all the abnormally large numbers in the world can't overturn some long-held beliefs. The stock has become a battleground between "this time is different," and "no, it isn't."

Single-digit forward price/earnings ratios are nothing new for Micron. Over the past 10 years, 54% of days saw its P/E under 10, and 27% were lower than today.

It all stems from the commoditized nature of the memory business. There have been wild swings in inventories and prices, which get reflected in Micron's financials as sharp fluctuations in sales growth and gross margins.

Micron has been subject to a phenomena known as the bullwhip effect. First identified by Massachusetts Institute of Technology business school professor Jay Forrester in 1961, the bullwhip effect is that the further upstream a supplier is from the end-user sale, the less visibility it has into demand, and the wider the oscillations are in inventories and, consequently, prices.

During the last cycle-from 2019 to 2023-most of Micron's sales were to consumer electronics and vehicle manufacturers. These products go through multiple distribution layers after the memory leaves Micron's warehouses, and the demand signals get more opaque with each step.

In the last go-round, the inventory situation was further complicated by the supply-chain chaos and fierce consumer goods demand during the Covid pandemic lockdowns. When that demand flagged, it was just as new memory manufacturing capacity was ramping up. Inventories at Micron and its customers shot up, and prices collapsed. With sales contracting and inventory write-downs piling up, Micron had a negative gross margin in all the calendar-2023 quarters.

The stock peaked in early 2022, just a few weeks after Micron reported its first fiscal quarter of that year, which saw slight declines in sales growth and gross margin. Micron's forward P/E was 9.7 that day, not much higher than it is today.

The old Wall Street maxim played out: The best time to buy cyclical chip stocks is when things go from horribly bleak to merely awful, and the best time to sell is when growth and margins are peaking. Many tech investors still feel pangs of regret from a cycle they mistimed, and have either sworn off the stock or are ready to jump ship at any sign of a cycle peak.

For those so inclined, they got that signal with Micron's guidance for the first quarter, with projected slight declines in earnings growth and gross margin. Déjà vu all over again.

But there a lot has changed in the memory market since that last cycle. Memory is still a commodity product, but it is also one of the key chokepoints in the generational investments into AI data centers, which show no signs of slowing down. Micron and its peers-SK Hynix and Samsung Electronics-were initially reluctant to invest new capital to meet data-center demand, and new manufacturing capacity isn't expected to begin providing supply-side relief until the middle of next year, with more slated for the following years.

Instead of riding fickle consumer demand, now data-center customers predominate, representing 63% of sales in the fourth quarter, rising from about a quarter of revenue in the last upcycle. There are fewer layers of distribution between Micron and the end customer in this market, and Micron has better demand signals. In this week's earnings call, CEO Sanjay Mehrotra said that fiscal 2027 would have an even tighter market for memory than 2026, with more price increases, and that the following year would be tighter still. "We do not have a line of sight to when supply and demand will get in balance," he said.

The stock was unmoved by that declaration.

Micron has its best visibility ever into what long-term demand looks like. Like its competitors, Micron is using the extraordinary market power it has to push customers into five-year agreements with price floors and high price ceilings, and take-or-pay terms. Micron now has 26 of these contracts which constitute over a third of sales through 2030. The more it can add to that, the softer an eventual downcycle will be.

And that may be what it takes to convince some investors that Micron is worth more than a single-digit valuation multiple: a downcycle in which the average sales price doesn't collapse, revenue doesn't crater, and gross margins bottom at respectable levels. If Mahorta is right, that may be years off.