so micron has a record ER, and guides higher but the market dumps.
Anyone who has held Micron who a couple years knows this story.
For those who lost their ass on weeklies – this is why I say dont do weeklies on micron.
Why is it dumping? Partially because of people getting margin called and closing out their weeklies. STOP DOING IT.
So anyway on to the report:
▪ Expect CY-21 DRAM industry bit demand growth to be somewhat above 20%, with DRAM industry supply below demand; there is currently unmet demand for DRAM due to end market strength and we expect the DRAM market to remain tight into CY-22
▪ Expect CY-21 NAND industry bit demand growth in the mid 30% range, with NAND supply below demand; we also see the NAND market as tight into CY-22
▪ Long-term DRAM bit demand growth CAGR of mid-to-high teens
▪ Long-term NAND bit demand growth CAGR of approximately 30% Micron
▪ Targeting to align our long-term bit supply growth CAGR with the industry bit demand growth CAGR across DRAM and NAND
▪ CY-21 DRAM and NAND Micron bit supply growth expected to be below industry demand growth
▪ Placed orders for multiple EUV tools from ASML as part of long-term agreement
▪ FY21 Capex revised to somewhat above $9.5 bn; mostly from areas that don’t impact CY21 and CY22 bit growth such as EUV pre-payments, construction spending and other R&D and corporate items
Now during the call, analysts seemed to really dig in deep into Sanjays comments re purchasers shifting from “just in time” models to “just in case” models. They think this indicates previous purchases were above actual consumption and intended to raise inventory. As such they expect weakening demand and pricing moving forward.
I disagree with their fears. Sanjay guides in a way so that he can smash at ER. He has done this for Micron and he has done with Sandisk before they got bought out at WDC. Have a look at previous earnings and whether they beat or not.
expand table to see more dates on links below
[https://www.streetinsider.com/ec_earnings.php?q=sndk](https://www.streetinsider.com/ec_earnings.php?q=sndk) (Sanjay for all of it)
[https://www.streetinsider.com/ec_earnings.php?sort=earning_date&q=mu](https://www.streetinsider.com/ec_earnings.php?sort=earning_date&q=mu) (Sanjay from Q3 2017)
Sanjay knows markets dont like surprise, so he guides low. Hes also been in Semi industry forever. Hes not an idiot. So when he says guidance is X, I take guidance to be X + extra for the beat. I dont freaking think I need to discount the guidance because purchasers increasing inventory. Sanjay has priced that shit in.
So thats inventory and guidance.
What about the sale to Texas Instruments? Thats not linked to RAM or NAND really. Kinda but not really. Thats them giving up on (3dx point) Intel Octane. There was no market for it so keeping the plant was costing them 500 mil a year since it was under-utilised. However they kept all the IP from Octane in case it comes back to life later on. Its basically non volatile RAM which can be used as storage – sounds really cool but I am guessing its expensive AF so people dont want it yet. Again, selling this plant is basically good for micron. We want higher margins to run a slick operation.
I dont really like toms hardware but this article explains the sale the best. Other news stories on it say there was a sale but doesnt really tell you much about about the background.
What is driving increasing margin? Well Micron is using DUV (deep ultra violet) with computational lithography and multiple masks to achieve smaller resolution on DRAM. Samsung are achieving same resolution with EUV. EUV costs a bunch and ASML are the only supplier. You should probably get into ASML at some point as well WSB but thats a post for another day. As such, Microns margins for latest tech DRAM is way lower than Samsung. SK Hynix is the third player in RAM and they are playing catchup – seem a bit fucked to me personally. Nanya is the 4th player and realllllly small and think they are just screwed. Dont have the money to get into EUV so next few generations, they will be so far behind they just get bought out if any competition regulatory body allows it (at near bankruptcy stage is my guess).
Now Micron is starting to invest in EUV which is good and they have given themselves a lead time for 2 years which feels about right to me. In 2023 we will see if they are still at cutting edge of RAM vs Samsung who will have a 2 year head start into EUV production. This is the only concern to me. In this time, they will gain the benefit of reduced pricing for EUV tech and additional expertise from ASML who will be helping Micron to achieve results. I am hoping they can transplant their specialist know how from DUV to EUV (that everyone else didnt manage to do at this resolution), but at this stage, my science ability is woefully unable to track it meaningfully.
Lets look at Microns plan for EUV and the interim releases until then.
So you can see we are currently at 1 alpha node for 10nm in 2021. We have 1 beta and 1 gamma to go planned (still using DUV) before 1 delta which is when EUV should kick in. Thats a new node every year still using DUV. Thats sweet if it works out. In the interim Micron gets to print money like nobodys business (higher margin from DUV) and continue hiring R&D people with EUV knowledge from Korea and Japan.
I have a small long position in micron. I dont really care about short term price movements. I expect this to be higher in 3 to 6 months. Thats what I care about. I expect it to be higher in 12 months. Beyond that I want to see how their 1 beta and 1 gamma nodes are coming along with updates as to 1 delta using EUV.
We are trading at price levels from Jan/Feb this year. We have had 2 great ERs since then. I dont see TAM in RAM or NAND dropping any time soon or pricing (thesis of tacit collusion as to supply/pricing in oligopolies – we arent cyclical market anymore because 3 market players control like 98% of the DRAM production in the world – its why they are not dumping on capex – its intentional).
this is good price to buy, but dont expect results in a week or even a month. We might trade sideways for a while. Shake out the paper hands before funds buy back in at discount and ramp up for next ER. You cant time the ramp up, you just have to be in the stock at a decent price (like now) and then wait for it to happen.