Micron Technology, My new pick for year 2018


Today, MU looks good with down price. This has grown 100% this year and still low P/E 6.95 and PEG 0.25.

Bought today, planning to hold as long as economy & Company are doing good.

Your comments and reviews are welcome !


I think it’s a solid pick. It should outperform the market.


It’s my 5th biggest position. Just added a bit more this morning. I totally expect it to be my top 3 if not top 1 poising jn 2018.



AI and China are my two biggest investment themes. To just pick one stock for each doesn’t do it justice. I tend not to touch biotech but I will take a look at Editas to see what it does.


I am very skeptical about china these days.

That said, MU +1.


Crowd says buy, do opposite or follow?


for MU? PE seems very low, so probably ok?


You followed the crowd into SHOP. What was the consideration there?


Jil, marcus, manch and tomato say buy MU. Crowd says buy.
marcus says buy SHOP. Jil and manch say no. Not sure about tomato. So crowd is not buy.


I recalled everybody said buy SHOP except me. Was @Jil against? I don’t recall.


Correction. I do not say BUY, but I said that I bought as it is attractive. Sharing the details to forum members. Individuals/they need to analyze and decide.


I guess MU can get some benefit from AI demand on memory but this is No3 company out of 3 DRAM companies with 20% DRAM 13% NAND market share.
There is nothing truly innovative in terms of DRAM technology except fab technology and their fab technology/capacity is usually 1-2 generations behind Samsung.
They tried to come up with new stacked memory (HMC) to be a market leader but I am not sure how well it goes.
All in all, I guess the company would be OK but not sure about great stock growth.


There’s an industry wide memory shortage going on right now (primarily in flash) and it doesn’t look like the shortage will end in the next 2 quarters (has been going on for 2 quarters now). So in the short term, i concur nothing dramatic will happen (short of a geopolitical event)


AI also requires hard drive storage - WDC might be another play.


Memory requirement is cyclical i.e. go through a cycle of shortage and excess.
I recall the thesis to buy memory makers now is that the shortage would last a lot longer than previous cycle because the industry underestimates the demand of memory from things going AI, hence under supply. This thesis has been going around for awhile (about 1-2 years), how much is already accounted for in the price of the stock? Still worth to buy now to catch the remaining upside?


Yah we were expecting to be out of the woods by this time of the year on the flash shortage. But nope… lead times are at 24 weeks for many components. Looks like all the Fabs are running full out (semiconductors are doing great this year after years of over supply). Probably it will correct sharply with the next downturn.


I think we are on a secular uptrend. That is the old cycles don’t apply anymore.

On the demand side we have AI and VR that’s driving huge demand for DRAM. We are just getting started to deploy GPU’s in datacenters. The even bigger demand driver will be AI at the edge devices like cars. VR hasn’t even started.

On the supply side we have a lot of consolidation in the last decade. We are now down to 3 players. An even bigger constraint is the death of Moore’s law. We used to be able to double DRAM bits every 18 months by moving to a new node. Now that 18 months is stretching out to 24 and 36 months. We are hitting physical limit. The new fabs are also super expensive and that means barrier to entry is much higher than in the past. Each new fab easily costs 10b.


My guess is all the 3 players are thinking alike and could have started building fabs to catch the increased demand. You would end up in excess supply when those fabs are built :grin:
Since you’re an ex-INTC and could likely still have contact in the semi industry, tell us how many semi companies are planning to build or already building new fab plants. When would these fabs be ready.


You can read the earnings call transcripts of the big 3 to know their capex plans. It takes 10s of billions and multi years to build a new fab.


Dont bet on vr.