I'm a proud owner of Micron Technology (MU -0.18%), but it's not a stock for the faint of heart. Despite good long-term growth prospects for memory demand, Micron trades at an enormous discount to the market and especially other tech stocks – even peers in the volatile semiconductor sector. That's because of the historical high cyclicality of the memory business, which is prone to booms and busts.
Case in point: Micron's recent earnings report, in which the company posted stellar numbers across the board, while also giving solid guidance. Yet as always, investors chose to focus on the negatives, real or perceived. Here's what those concerns are, and why I'd be a buyer of this dip despite them.
Image source: Getty Images.
Costs and investments are going up this year and next
Memory crashes tend to happen when the industry overinvests in supply when demand is good and memory prices are rising, only to have too much capacity when demand hits an air pocket and pulls back. Currently, we're in the early stages of one of those good times. Last quarter, Micron's revenue surged 19% quarter over quarter and 36% year over year. Adjusted (non-GAAP) earnings per share nearly doubled quarter over quarter. Management projected the growth to continue next quarter, forecasting 10.5% sequential revenue growth and 22.3% sequential earnings growth.
Yet on the subsequent conference call, Micron explained it would be increasing costs more than the market may have expected. First, the company announced that it was ready to deploy extreme ultraviolet (EUV) technology into its DRAM production starting in 2024. But because of EUV supply constraints, Micron has to make early pre-payments to secure these expensive machines.
Second, although Micron is lowering costs through node transitions, it's also targeting more specialized, higher-value types of memory solutions -- but those come with higher costs. The worry here is that Micron will be producing more expensive chips in 2022, a time when some believe we could be seeing falling memory prices.
Perhaps adding to the concern was that Micron only repurchased $150 million in stock last quarter, or only about 10% of its free cash flow, below the company's long-term 50% target. That could indicate Micron is being cautious about its balance sheet for some reason.
Why I'm not worried
Given not one but several surprising wrinkles in the company's spending plans, so it's no wonder Micron investors -- already a skittish bunch -- sold off the stock, which is down some 20% from its April highs.
Still, I think these concerns are a bit overblown. Yes, the EUV costs are coming sooner than expected, but Micron said its capex would increase this year from around $9 billion to over $9.5 billion. Micron is flush with over $9.8 billion in cash, and it will likely be very profitable over the next couple quarters at least, so it should be able to afford these hundreds of millions in early payments.
Second, although costs for higher-value chips will be, well, higher, Micron should still be able to earn as good or better margins on those products, at least relative to lower-cost chips. After all, there's a reason management is targeting higher value solutions. Higher-value chips may also be less prone to price drops than more commodity-like memory chips. Zinsner noted on the conference call, "we are trying to drive toward higher value products, which, arguably, on a like-for-like — or at least on a comparable basis to other products would carry better gross margins."
Finally on share repurchases, Zinser also told investors not to worry:
I think you'll find in the fourth [fiscal] fiscal quarter that our buybacks are meaningfully higher than our third fiscal quarter. So nothing to read there, we do feel like this price is obviously a good price to be buying the stock back. And we are committed to what we've talked about previously, which is to return at least 50% of our free cash flow in the form of buybacks.
Micron's stock price is now lower than it was for much of the fiscal third quarter ending in May, making buybacks more attractive, so perhaps this was just a bit of savvy timing by management.
Another possibility is that Micron has set liquidity targets in the low 30% range of its revenue. Therefore, if Micron is expecting big revenue growth this year and next, it may have wanted to bolster its balance sheet to keep up with its internal model. Micron ended last quarter with $12.3 billion in liquidity, and guided to $8.2 billion in revenue next quarter, or roughly a $33 billion run-rate. However, Micron has a habit of guiding conservatively, and if revenue continues to grow, Micron would have needed to add liquidity to its balance sheet. Management still expects using 50% of cash flow to fund buybacks over the long-term.
Things aren't so bad -- in fact, they're really good
While nervous Micron investors tend to wring their hands at any imperfection, some may be overlooking how good conditions are for the memory industry right now. CEO Sanjay Mehrotra noted supply tightness in almost all of Micron's end markets, and sees that persisting at least into 2022.
That's due to a great combination of pandemic-fueled digitization across many industries, a strong economic recovery, and customers looking to hold higher levels of inventory going forward. Many businesses such as car manufacturers used to hold the bare minimum of chips in inventory in order to lower capital costs; however, that ultra-lean posture is costing those businesses in lost sales today.
Combined with the facts that chips are becoming more difficult to produce and the three major DRAM manufacturers have been prudent with supply growth, it's easy to see how supply may have trouble catching up with demand for a while.
The last downturn was spurred by the U.S.-China trade war, which caused a recession-like demand shock in the memory industry, as all clients pulled back on buying at once. While a big demand shock like that could happen again, it may not happen soon, or be quite as severe. That could make this memory upcycle stronger and more durable than some anticipate. If that's the case, then Micron is still woefully undervalued.
