MU - Micron Technology Inc.

[yadayada]

If you look into the history of Intel... it started off as a memory company.  Then they starting losing a lot of money in the business.  Andy Grove took a long time to get Intel out of the memory business... but fortunately for Intel he did.  Intel subsequently morphed into a very successful microprocessor company.

 

The memory business is scary and ugly.  The main competitive advantage is scale.  Unfortunately, it means that many industry players will overbuild capacity because they are all trying to get bigger.  Then everybody loses a lot of money.

 

The market leaders in memory are no longer the market leaders today.  Many of these companies have gone out of business over the years.  Historically it has been a crazy and ugly business.

Yeah but dynamics can change. There used to be alot more players, and now there are only a few. WIth more smaller players you have more risks like this no?

 

But yeah i really liked this 10-12$. But now it doesn't seem so safe anymore.

 

It seems when there is even 1 weak player competing with them (or more in this case), then the stronger players have a very large incentive to drive down prices to drive out the weak players and try and get some of that market share. But when you have only a couple large financially strong players left, that incentive is no longer there right"? Im just thinking out loud here. So now the rational thing to do is not really to try and drive out any of those large players,  because in the long run that probably wont get them much market share and kill profits for everyone.

 

This is pretty informative, there were 6 players only a few years ago, and now there are 3.

http://thememoryguy.com/dram-consolidation-in-2012/

 

http://www.theregister.co.uk/2013/03/19/micron_flash_numero_uno/

 

This is also interesting. I think it is worth looking into.

[yadayada]

oh and whatever your opinion is, just  look at the LEAPS here? you get almost 2 years for the market to get rational.

 

http://finance.yahoo.com/q/op?s=MU&m=2016-01

 

With only 2.10$ in EPS, that is a 25$ share price. What about the 20$ calls here? at 25.80$ you break even. Doesnt that look pretty cheap?

 

Depending how high conviction this is. What I like is, that depending how solid the thesis is here, I dont really see the market stay irrational and at a multiple of like 6-7 here. They are also buying back shares this year.

 

with 3.50$ in earnings that is more then 42$  a share I think.

[dbuch]

Thanks for the links, they give a good history of the DRAM business.

 

I actually do own several of the 2016 LEAPS and the stock. I think without subsidized players and Samsung (40%), Hynix (24%) and Micron (25%) controlling the market, the supply will be much more rational. If you listen to the latest Hynix, Micron or Samsung conference calls they all talk about being more rational with supply, more stable pricing, slower transitions to the next node etc.

 

On Samsung's last CC

 

<Q>: I have a question regarding the Memory business. The companies have been trying to restrain from supplying or increasing the supply of DRAM, but we’re now seeing some of the companies actually increasing their CapEx as well as increasing their capacity. Does this imply that you may, from Samsung Electronics’ perspective, see more than a growth – a growth higher than the mid-20% for in terms of bit shipment?

 

<A>: As we’ve mentioned in previous earnings conference calls, in 2014, we have no plans of increasing physical capacity on the DRAM side. The other companies are investing currently in terms of capacity and CapEx, but I don’t think those investments will lead to capacity increases during this year. The investments that are made – being made currently by the other companies will probably come online in H2 next year or after that, which leaves us the only option of migration. But in order to migrate, there is a limit in terms of the physical space additionally that we can use. And in addition to the investments, migration itself takes time. So due to all of these considerations, in 2014, we still believe that the potential bit growth would be limited.

 

The only additional supply planned is from technology shrinks and is expected to grow below or in line with overall demand. This means prices will remain stable and revenues growth will equal demand growth.

 

Samsung has another reason to keep prices higher. Their major competitor is Apple and Apple must buy DRAM for their devices from Micron or Hynix. The higher the price of this input the more pressure on Apples margins giving Samsung, who is vertically integrated, more competitive leverage. So all three have huge incentives to act rationally.

 

It also seems incredibly difficult for new entrants. There is technology know how, patents, not to mention the cost of building a state of the art fab is probably $5B. Also any new entrants would know the minute they enter the market and try to build market share, prices would drop ruining the economics they sought after.

 

The one risk with all technology companies is obsolescence but DRAM technology has been around for 40+ years as a way of storing temporary data so doesn't seem like technology will completely wipe out the need for memory devices any time soon.

 

Just a back of the envelope calculation: If I assume $21B in revenues in 2016 with a 45% gross margin and $2.7B in SG&A and R&D, I get $6.6B of Operating Income. Subtract interest of $250M and 35% taxes (that's GAAP not actual which is likely much less) gives you $4.2B of Net Income on 1.1B shares gets you $3.8 EPS. If I throw a 13x multiple on that I get a $50 stock price or 31% annual return over the next 3 years. Lots of assumptions there but I don't think that's too crazy.

 

 

 

 

[dbuch]

Drexel Hamilton Raised price target to $50. Does anyone have their thesis?

[dbuch]

Hynix reported.  It sounds like they plan on remaining rational with respect to supply.

 

On the supply side, factors like suppliers switching their DRAM capacity to NAND, process migration to 1X nano and

increased share of TLC and start of 3D product production may have negative impact to the overall supply-demand

situation. However, since most of the capacity conversion have been completed and since the suppliers will continue to

make efforts to maintain supply-demand balance in case of weak demand, the overall supply-demand situation will

improve starting in the second quarter.

 

Under such market circumstances, SK Hynix will focus on profit-oriented business management and try to achieve

qualitative growth rather than quantitative growth. At the same time, we will make our best efforts to preparing

next-generation products to take a great leap forward and to building the growth infrastructure to generate stable profit

for a long time.

 

If you look at other players like [ph] Intel or Micron (32:38), overall, there is some shortage in the overall industry capacity. However, due to technology migration efforts and – they're not really – the bit shipment is currently not very high. When about do you believe that they're going to start investing in the capacity expansion, because right now their free cash flow level seems to be relatively high? So we want to know at the company level, what's your planning to do?

 

Regarding your second question, which is about the capacity increase for DRAM in the long-term. Well, on the

short-term, we believe that the overall players are likely to maintain their current capacity, just maintaining the

operation at the maximum level, and the wafer capacity will slightly drop. However, if you look at the long-term

outlook, while it's difficult to forecast the future, but we believe that in the next two, three years there won't be any

major capacity expansion.

Everything will depend on how people view about the demand growth. However, we believe that as a Hynix point of

view, we are not really planning on any capacity expansion in the next two to three years. Everything actually depends

on when the new generation technologies will settle in.

 

 

[constructive]

Micron is down 19% today on weak earnings and guidance, accompanied by lots of downgrades. Rolling out their latest process technology is not going smoothly.

 

This looks really cheap to me. The memory market is a lot less competitive than it has ever been before, with Samsung, SK Hynix and Micron manufacturing 90% of DRAM. If you look at the analogy to hard drives, Western Digital and Seagate have done very well since they consolidated their market down to 3 companies.

[AzCactus]

This looks crazy cheap to me too.  However, I haven't taken a deep dive into the company. 

[Jurgis]

Although I have invested in the past in WDC/STX on the same premise: that the competition is down because of "oligopoly", I think it's a bit risky premise.

 

In case of memory, you have Samsung who does not have to behave rationally. (Or they might behave rationally from their side, dumping memory to get marketshare and make money from other chips).

 

It still might be that MU is cheap enough here now. I'll have to look.

 

I think this is also a risky premise for airlines.

 

Disclosure: no positions in any of the above.

[constructive]

Although I have invested in the past in WDC/STX on the same premise: that the competition is down because of "oligopoly", I think it's a bit risky premise.

 

In case of memory, you have Samsung who does not have to behave rationally. (Or they might behave rationally from their side, dumping memory to get marketshare and make money from other chips).

 

It still might be that MU is cheap enough here now. I'll have to look.

 

I think this is also a risky premise for airlines.

 

Disclosure: no positions in any of the above.

 

There is plenty of evidence that the memory market is more profitable than it has ever been before. I believe the industry as a whole was not earning its cost of capital for 30+ years, and now they are earning well above that mark and should continue to.

 

Samsung is not necessarily irrational. Their short term growth plan is overaggressive, but they can't change capex strategy on a dime to be responsive to short term pricing. Long term bit growth is still over 20%, so each of these companies obviously needs some kind of growth plan.

 

Airlines are completely different since there are still dozens of international competitors, high levels of regulation, and low barriers to entry.

[RadMan24]

Drexel Hamilton Raised price target to $50. Does anyone have their thesis?

 

Well..a year later and they cut to sell and a $20 price target. Citing: "Predicated upon a view Samsung’s strategic interests bring willingness to accept lower memory margins, DRAM price drops are liable to continue," said Whittington. "PC DRAM price weakness has spread to mobile, reducing margins and earnings estimates as Samsung brings process cost drops to customers."

 

This was when Micron was at $25. Also believe DRAM and eventually NAND gross margins will fall for Micron.

 

But the company has executed very well over the past few years, starting from 2010 to now, management has done a very good job. 

 

 

[jawn619]

Stock is looking kinda cheap

[dbuch]

It appears the market doesn't think the oligopoly is working. ASPs fell as expected due to PC weakness but costs were higher. I think the reason is the transition to 20nm. They are running parallel lines with higher WIP, moving to DDR4, I'm sure there are some hiccups in the process as well. Micron still thinks they will be getting 20-25% cost reductions and pricing should largely stabilize.

 

Fundamental technology migrations yield fewer bits per dollar so the costs to oversupply the market are getting higher and should be a governor on irrational oversupply. Samsung isn't just going to build fabs to gain market share when it continues to get more expensive to do so.

 

What do they have to gain? They have 40% market share and trying to put Micron out of business wouldn't help them as Intel would likely just buy the assets and then Samsung is competing with Intel instead of Micron, they replace one competitor with another. Knowing this, they'll remain rational and limit supply to demand the best they can. It probably won't be perfect and their could be some ASP degradation but it should be substantially below any cost reductions which will increase gross margins.

 

On the NAND side they are transitioning to TLC and 3D which should be more competitive and reduces costs. The majority of the cost declines will be passed on to customers but margins have room to move up here as well since the others are already using TLC and current pricing allows for higher margins compared to what Micron earns.

[AJDelphi]

Einhorn pretty much lays out his thesis for Micron and his valuation a few years down the road in his Q2 letter.

 

http://www.valuewalk.com/2015/07/greenlight-q2-letter-netflix-nflx/

[Wilson-TPC]

He should've sold like Klarman when MU was at 35. That was 10x P/E which is pretty fair for a cyclical business.

 

Second half of this year should see some boost as window's 10 release will help PC sales.

 

I haven't looked at the projected earnings yet... maybe it's time I should.

[gurpaul88]

http://www.bloomberg.com/news/articles/2015-07-14/china-s-tsinghua-unigroup-bid-21-a-share-for-micron-wsj-says

[AJDelphi]

I certainly didn't expect Micron to get a buyout offer....

[dbuch]

Samsung never seems to answer the question directly but in the latest call they do mention they have nothing to gain by lowering DRAM prices. I still think the three will remain rational with respect to supply which means stable prices and margin growth. Micron should continue to be able to save costs via node transitions while keeping ASPs relatively stable. This means higher revenue and margins.

 

"First of all, on the PC side, yes, we’re not offering a positive outlook on the PC demand either especially given the fact there it seems to be a large finished PC that PC set inventory, that’s still there in Q2 and that will continue on Q3. So that would remain slow. But I don’t think it’s type of a market where we would gain any by lowering our DRAM prices. I don’t think, given the current PC DRAM situation, lowering prices on our PC DRAM would generate additional demand or create demand. And so given that market, it’s not the type of market where we want to drive down our prices for the sake of market share."

[abcd]

http://www.bloomberg.com/news/articles/2015-10-01/micron-s-fourth-quarter-sales-profit-top-analysts-estimates

 

[walkie518]

anyone have a feel for today's drop?  earnings looked fairly solid...guidance for bit supply growth?

[nkp007]

Anyone involved in this? Trading at 5-6x earnings, industry historically cyclical but some claim that cyclicality has been blunted because new supply is more expensive to bring online and much more diversified sources of demand (phones + data centers + autos + pc's).

 

 

[saltybit]

https://www.bloomberg.com/news/articles/2017-09-29/sun-co-founder-sells-secretive-hedge-fund-on-radical-chip-trade

[DTEJD1997]

Hey all:

 

I've been kicking myself for a while about MU.

 

I use an incredible amount of RAM in one of my businesses.  The price of memory has MORE than tripled in the past 18 months.  The price of memory is now starting to be a real problem.

 

There are a lot fewer memory manufacturers now...so the industry will probably act in a rational fashion.

 

Micron Memory is also very high quality.  I've been starting to use AMD Ryzen processors which are very good when it comes to price/performance ratio.  ONE BIG PROBLEM with them though is that they are VERY finicky about the memory that they take.  Some manufacturers have a problem with this...but not Micron product.  As a result of this, I'm upping the purchase of MU memory.  Only in certain oddball cases will I use another manufacturer going forward.  It will be easier for me to standardize on MU memory.

 

I am going to assume that I am not the only having this issue with memory compatibility.

 

I think MU will be making even more $$$ than what the analysts think.

[JRM]

I haven't invested in a semiconductor or memory company since SunEdison Semi was bought out by Global Wafers.  Back then the market hadn't 'turned' yet, but the bull thesis resolved around a demand wave on the horizon driven by smartphone refresh cycles and 'internet of things' adoption. 

 

I haven't looked closely at these companies, but is the recent demand due to blockchain demand instead (processors and GPUs for mining)?  If so, is the demand sustainable?

 

I think eventual adoption of 3D NAND and 3D Cross Point will be revolutionary when it comes down in price a little bit.

[nickenumbers]

All,

 

I am looking at Micron, and it looks like a good little performer trading at a mid 6 PE.  The revenue growth and EPS growth look very good.

 

But, this recent closing year 2017, had revenue growth that was just abnormal high.  I assume it was price based, demand driving by limited supply.  It could be blockchain mania and the underlying hardware demand, etc.  Is the revenue sustainable?  Was there another cause of the revenue pop, price pop?

 

I don't think I know enough about the underlying hardware technology.

 

Does anyone have a good bear or bull case for this one?

[brianr27]

Einhorn gave some insight on MU at the end of the last cycle. Although it's from Nov 2015, I think that it's still relevant. You can find it here:

https://www.gurufocus.com/news/373674/david-einhorn-comments-on-micron-technology