Let’s say we’re in another positive cycle of better times for Micron. We’ve been through them before; the most notable and mind-altering one was in the mid-1990’s. Although it got eclipsed by the Internet bubble a few years later, it was a heady time and everyone wanted to own Micron.
The problem for me as a growth investor is that Micron seems to be just a new-age commodity company. Memory sizes are growing so Micron will ship far more “bits” every year. But thanks to advances in density they are basically shipping the same number of chips. (Micron fans will point out that they are expanding lines and shipping more chips but let’s just think about it like the “same store sales” model. In that analogy Micron grows by adding stores but the stores themselves don’t grow.)
To put a consumer perspective on it I buy a USB key or two every year for the family and pay around $10. So it’s the same $10 per year over many years. Not a bad business if you consider the millions sold but not growing either. It’s true I get a 4GB or 8GB key now versus the 128MB I was getting a few years ago but Micron had to bear the capital investment and development cost to get my $10 again.
Of course Micron sells other types of memory but it’s all the same. You get one or two sticks of RAM for your laptop and they just have twice as much memory on them every year. Same two sticks, same price as long as you don’t try and buy at the tip top of the density curve.
In some ways it reminds me of EMC. A little less than a decade ago EMC was trading at book value and generating huge amounts of free cash flow (FCF) It was the one and only time I bought a disk drive stock. However, EMC has acquired many software companies to improve the dynamics of their business. They were very fortunate in their purchase of VMware which now is talked about more as a value driver than the company itself.
Sometimes it doesn’t work out so well if the acquisition comes at too high a cost (Veritas comes to mind) but that certainly doesn’t preclude the acquisition strategy from starting to improve the dynamics of the business. Of course some companies, like Dupont and 3M, have figured out how to innovate enough around pseudo-commodity markets to enjoy brand-like margins and better growth.
It appears that Micron is headed for a great year ahead with improved demand, stable to improved pricing with limited planned capacity additions. This will allow the company to generate close to $2B of FCF.
But the up cycle has always been followed by the down cycle where capacity comes into the market, usually just when demand is starting to ebb and prices are falling. So most investors discount the down cycle on the way up. (This is the same way the auto companies used to trade. As senior analyst Dave Healy used to say “you buy them when their PE is infinite and sell them when the PE hits 4x.”)
Most of the semiconductor companies we follow have substantial IP and software in thematic areas we invest in like the Cloud, Mobile Internet or RealVR. This year may be the last good chance Micron has to do something very strategic and transition the company to a model that attracts long-term growth investors.
Should they buy a GPU company? That’s an interesting thought.
[Disclosure: I own a bit of Micron here because I think it can move higher given the fundamentals and market conditions, but it’s not and will never be a long-term holding unless something changes.]