Micron: What is a commodity?

Investment Analysis Clubs Saul’s Investing Discussions
1

I thought I would take a closer look at Micron today. Ant has brought it up a couple of times, and Putnid and others have chimed in.

This is a great thread from a couple months ago: http://discussion.fool.com/micron-mu-er-coming-watch-the-story-3…

One of Putnid’s contributions to the thread was especially helpful to me: http://discussion.fool.com/can-you-say-a-little-about-their-comp…

Especially this part: NAND and DRAM chips are commodity products. They’re in high demand because they’re used in all manner of electronic applications. When demand outstrips supply, chip prices skyrocket. Eventually, one of these companies builds another plant (even though it’s a huge investment).

It’s amazing to see how this plays out. Micron describes themselves in 4 segments on slides 11-14 of their recent investor deck: https://seekingalpha.com/article/4109704-micron-technology-i…

For brevity’s sake, I’ll call them:

DRAM
NAND
SSD
Auto

DRAM was the biggest by far, accounting for almost 50% of revenue in the quarter. It’s really incredible to see the numbers. Short version:

In Sep16Q they sold $1.25B worth of DRAM for a profit of $11M.
In Sep17Q they sold 2.85B for a profit of 1.6B.
So they went from a 0.9% profit to to 56%.

I have seen no clearer example of how commodity products work. That was just their largest (DRAM) segment, to be clear…but the overall company gross margin went from 18% to 51%. It’s truly all about demand.

Another way to look at it:
Sep16Q rev: 3.2B
Sep17Q rev: 6.1B

Sep16Q exp (incl COGS): 3.2B
Sep17Q exp (incl COGS): 3.7B

So basically, they need the demand or else they’re just covering costs. But when demand goes up almost all of it drops to the bottom line.

So what?

Short answer: I don’t know.

Are higher prices here to stay? Putnid’s point above makes me think probably not. But then, that’s why Micron is sporting a PE of 10.

It does look like, by all estimates I’ve seen, demand will continue into next year. Does that make Micron a short term buy? How much will it drop if another player announces plans to “build another plant?” And how likely is it they will?

I’ve heard a lot of predictions, but I’m afraid we will all have to decide for ourselves what’s likely to happen.

Bear

4 Likes
2

It does look like, by all estimates I’ve seen, demand will continue into next year. Does that make Micron a short term buy? - Bear

A quick check of my investing history reminded me that I have been buying MU in discrete lots every month throughout most of the year (Channel-surfing quite a bit). Indeed, I already bought more in November.

Bear, you are absolutely correct in noting that commodity plays bring with them their unique considerations. For example, Commodity plays aren’t my favorite candidates for simply buying and holding. Commodity plays fall more in the realm of “bottom fishing”.

At any rate, I still believe that MU has significant upside. Definitely in 2018 and, quite possibly, for several more years beyond.

The great unknown is the DEMAND. MU has fared well since memory chip demand outgrew supply. However, it’s not at all clear to me that demand growth will be orderly. Memory chips go into every conceivable device that needs some sort of memory. That’s a huge demand in and of itself. But we’re on the cusp of unfathomably greater demand: autonomous cars, robotics, the whole IoT phenomenon. Growth won’t be linear. It will be geometric and it will be astounding. I believe MU will be a direct beneficiary. There will be other winners, to be sure, but I’m still evaluating the candidates.

7 Likes
3

Guys

Ok 3 points to note here…

  1. The cycle in memory chips (price, volume and margin) has been asymmetric and trending towards an upwards direction. There will still be a cycle but every cycle has seen a higher peak and a higher trough so much so that the next trough might well still be positive - the closest I can describe it is an ascending triangle in chartest terms (BTW I’m no TA specialist). (This has been helped by a consolidation in players). There was a market realist series illustrating this that I posted way back when…

  2. Memory is going through a once in a lifetime maybe once ever transition where every single item becomes digital and smart/connected requiring memory at the edge as well as memory in the cloud to cope with the transmission, processing and storage of big data, similar to Cisco and the dotcom era.

  3. Potentially the best performers will be those that provide the least commoditised and most value add product - Micron is ahead of this curve with 3D point as well as other value added tech. Last week I met one of the guys that works for their innovation labs in Taiwan. He said they have plenty more innovation to come.

Micron is now a top 5 holding for me.

Ant

3 Likes
4

Potentially the best performers will be those that provide the least commoditised and most value add product - Micron is ahead of this curve with 3D point as well as other value added tech.

Problem is, nearly half of Micron’s revenue is from DRAM. Add NAND, and you’re looking at 65% of the company’s revenue. Automotive, which they’re touting as the next big thing, was only 13% of revenue this quarter. Hard to know how big 3d xpoint will be, but for the present I’m looking at all these things as “other bets.”

Will still be interesting when they report next month.

Bear