You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!

Semicap Summary- December Quarter & 2016 Outlook- Where to from here?

Robert Maire

Moderator
The expected recovery in 2016 started early for some... Q1 weak- Big H2? - Betting on the come for 10nm? Capex zero sum game= -Samsung +TSMC +Intel

View attachment 16533

With most companies in the Semicap space reporting by now we see some obviously common themes- A bounce in the December quarter off the bottom- A soft Q1 followed by an assumption of a strong H2 driven by 10nm and 3D NAND spending. Who benefits in a "Zero Sum" Capex game?

2016 is shaping up to be the third year of flat capex for the industry with companies talking about roughly $33B in industry spending. We are likely looking at spending being down for Samsung while up for Intel and TSMC (though TSMC is being coy at this point by not talking about its capex plans giving it perhaps more leeway).

So from an investor perspective the things we have to figure out is who will benefit and who will suffer by the change in customer spending mix and who will gain or lose market share in a stagnant market. Market share gains can either be organic or through acquisition.

KLAC is the poster child for customer mix shift...

KLAC has suffered for the last few years as memory spending has ruled the roost. DRAM has died and everyone wants to get to 10nm logic soon so KLAC is an obvious beneficiary as logic/foundry need more yield management than memory. Yield management & metrology usually are one of the first things bought by chip companies that need to develop the next new process. This explains why KLAC is knocking the cover off the ball on orders especially in Gen 5 tools.....its a no-brainer

EUV continues to be late so Dep & Etch continue to benefit...
Every extra dollar that companies don't spend on the never to be ready EUV tools is another dollar going into the pockets of the manufacturers of dep and etch tools (like AMAT & LRCX) that support the multi-patterning alternative to EUV. This is yet another zero sum game in capex....

Arms merchants benefit in the uptick...
Sub suppliers to the industry such as UCTT, MKSI & AEIS are generally the first to benefit and were the first to report positively such news. To some extent they are also a bit agnostic as to who wins, be it Lam or Applied as they sell to both and will see orders no matter who wins the end customer (with some mix pluses and minuses)

Bigger companies tend to see benefits first...
In all the upturns we have been witness to, the larger companies see their orders uptick before the smaller companies see it. In this case, although they had good quarters, companies such as NANO and RTEC didn't knock the cover off the ball like their big competitor KLAC.
From an investor point of view this could potentially be an opportunity as the uptick is yet in front of them and the stock hasn't traded up on the anticipation.

Investors are concerned about Samsung exposure...
Given the crappy report by Samsung, with the exception of their semiconductor division, exposure to Samsung as a customer is seen to come with some risk. With Samsung capex looking to be down and DRAM down for the count the only bright spot for Samsung spend is 3D NAND. We think that there is likely downside to Samsungs foundry spend as there appears to be a reported tilt by Apple towards TSMC. If thats the case Samsung will spend to get to 10nm but not build it out for high capacity.

ACLS was likely one of the victims of an association with Samsung as investors took the stock out and shot it even though it had a good quarter and good guide and continues to make nice progress. In our view likely an over reaction that likely creates some opportunity. If Mattson were not being acquired and eaten by the Chinese dragon its likely their stock would have otherwise suffered given their huge exposure to Samsung and memory (out of the frying pan and into the fire??)

Speaking of the Chinese...who's next on the acquisition menu?
We have thought about this for a while and think that one of the prime candidates to be acquired by the Chinese is Ultratech (UTEK). Why you ask? A number of reasons..... First, we think the Chinese decided to start with Mattson as it is a small, not strategically critical to US interests and not a particular threat while still having three technologies; strip, etch & thermal. Hmmmmm, that sounds a lot like UTEK, only UTEK is a better company. They are small enough that it likely wouldn't raise government eyebrows and the Chinese would get a stake in thermal, lithography, metrology & ALD technologies. UTEK has a pile of cash so the enterprise value is low. Art may be getting tired of activist investors and this would be a great exit as well as a way to shut up the activists. Sounds like a potential winner......

They say timing isn't everything....its the only thing...
If there was an award for timing in the industry, Lam would be the winner. Lam started talking to KLAC back around Semicon West when things weren't all that hot for KLAC and cut a deal. Now after the deal is announced and the two companies are working towards the integration KLAC comes in with huge upside all to the benefit of Lam.

Semicap companies remain the Caboose on the train...
Although almost every semicap company is reporting better times ahead we would remind investors that semicap companies are the caboose on the technology train. The locomotive pulling the train are companies like Apple selling smart phones, tablets and computers to consumers and industry. Tim Cook is at the controls and he is reporting cows on the tracks. The locomotive and subsequent cars on the train can be off the tracks careening down the embankment long before the caboose ever sees it coming. They are always the last to know.

We remain concerned about the macro economic environment and the slowing of China and other issues that eventually will trickle down through the semiconductor industry into the semiconductor equipment industry. No matter how much the industry wants 10nm and 3D NAND, if consumers aren't buying , no one is going to be buying equipment to produce what no one is buying.

We do think that Intel has to spend on 10nm much as we believe that TSMC has to get to 10nm for 2017's Iphone, but there can be a large difference between buying enough tools to support high volume and enough tools to support softened demand. In other words...we would remain very cautious and not assume anything.

The stocks have been a mixed bag...
For the most part, semicap stocks seem to have a higher correlation coefficient to the broader market than to industry specific expectations. Most of the stocks haven't traded out of their "normal" range and most stocks are down over the last three months much like the broader market despite the better news flow.

We would suggest that many of the stocks remain "cheap" relative to their "intrinsic" value as technology companys but that doesn't mean they are great buys at current valuations.

We think that there will likely be a cap on the appreciation that many of these stocks can achieve and post earnings season in the absence of good news they will likely drift with the market.

We have to be very selective as to who will benefit from the mix shift we are seeing over the next year. We also need to pay attention to any signs of changes in investments or 10nm timing that could push back or pull in related tool purchases.

We think its going to be harder to get stock home runs in this group in 2016 and will have to settle for singles and doubles by picking stocks that are unfairly beaten down or have better longer term trends or unique circumstances.

This seems to imply trading around positions rather than settling down for long term marriages to stocks.

This lack of "beta" probably makes the group less attractive as many investors played it due to its high beta and frequent "10 babgers"......which will likely not be the case....

Semiconductor Advisors LLC
 
Last edited:
Top