Shares of Micron (MU +9.17%) have been on a veritable tear over the last several years, topping $26 and bursting to new multi-year highs. The DRAM business has consolidated nicely, and the remaining major players have done a fantastic job of keeping the supply/demand balance quite favorable for what is essentially a commodity product. That being said, with predictions from the analyst community that Micron's DRAM margins may hit 50%, one thought that's impossible to ignore is, "If DRAM gross margins hit 50%, it's time to short Micron!"
Upfront disclaimer: NOT calling for a Micron short today
In the most recent quarter, Micron reported a gross margin of just 34%, a far cry from firm Drexel Hamilton's prediction that DRAM margins will eventually hit 50%. At these levels, margins are still healthy, but well below the margins of semiconductor logic fab, Taiwan Semiconductor (TSM +2.79%), which reported a gross margin of 47.5%, and guided to a range of 47.5%-49.5% for the coming quarter.
For now, the margins for commodity DRAM, which are much less difficult to design/manufacture than more complex logic semiconductors, are still well below the logic manufacturers, which implies that there's fairly muted risk of a "gross margin" bubble (i.e. the margins aren't yet so obscenely high so as to invite companies to build capacity/enter the game). However, near the 50% level, things get riskier.
Why's that?
Imagine for a moment that DRAM margins eventually hit 50%, predicated on the notion that DRAM prices are just going to continue to rise essentially in perpetuity. Here we have a market that is more than $30 billion in size, whose major barrier to entry is capital expenditures, and not really R&D. The good news is that, for a roughly 30%-35% gross-margin business, laying down billions for manufacturing plants is a difficult proposition. But at a 50% gross margin, which is more than a highly-sophisticated logic foundry gets, it's a no brainer. Eventually -- and probably quickly -- somebody will break and begin to add DRAM capacity, eventually driving margins back down.
We're still far away from that
The good news is that with Micron's margins still in the 34% range, there's probably quite a bit of runway before we need to worry about margin pressure. Further, given the very clearly positive secular trends in the market (mobile DRAM consumption is likely to continue to see robust growth as phones/tablets become more functional and as the high-end phones move to 64-bit, enabling greater than 4GB of RAM per device), the demand growth is not likely to tumble just yet.
Things will start to get less fun for investors in Micron when mobile devices, even the lowest-end ones, start to feel "good enough." At that point, the amount of DRAM/NAND content within them will probably flatline (most consumer PCs today have 2GB-4GB of RAM) and – in the race to the bottom – there will probably be a shift to lower-priced, slower DRAM. That said, this is probably not a phenomenon to worry about until probably 2016-2017, so there's still – again – some runway there.
Foolish bottom line
Micron is cheap today, trading at about 11 times trailing-twelve-month earnings with robust participation in a secular-growth market within which there is limited competition. There's probably still some really nice growth to be had over the next year or so, but if margins expand to the point that companies with the money would be foolish not to try to get a piece of that 50% gross margin action for fairly limited R&D, then that'll probably signal the top for industrywide profitability as competition/capacity expansion would probably drive margins back down.
