Friday, February 06, 2015

When Energy EPS Drops....

According to the Investment Strategists at Well Fargo there is hope for the IT, Healthcare and Consumer Discretionary sectors when energy prices drop.   Here's an excerpt from this piece:

When Energy-sector earnings declined by 32 percent in the one-year period ending on March 31, 1998, the best earnings-growth rates over the next 12 months were generated by the Information Technology, Health Care, and Consumer Discretionary sectors. The comparisons for first-quarter 1998 through fourth-quarter 1999 are as follows:

Click the image to see a larger view.

Related, here's a clipping related to the current state of earnings releases from StockTwits:

Last but not least, a nice set of charts about the oil industry:

Wednesday, February 04, 2015

The ITRS is Now Focusing on Systems

Unreal...   I have been talking about this transition for years and, lo and behold, the ITRS (International Technology Roadmap for Semiconductors) is now doing the same:

Focus areas:

In an IEEE paper published late last year titled “ITRS 2.0: Toward a Re-Framing of the Semiconductor Technology Roadmap,” the roadmappers explain why it’s time for a change. “As new requirements from applications such as data center, mobility, and context-aware computing emerge, the existing roadmapping methodology is unable to capture the entire evolution of the current semiconductor industry. Today, comprehending how key markets and applications drive the process, design and integration technology roadmap requires new system-level studies along with chip-level studies.”

The ITRS roadmapping committee has already been reorganized to focus on ITRS 2.0. There are now seven groups focused on what ITRS chairman Paolo Gargini calls the seven “building blocks.” 
  • System Integration—studies and recommends system architectures to meet the needs of the industry. It prescribes ways of assembling heterogeneous building blocks into coherent systems.
  • Outside System Connectivity—refers to physical and wireless technologies that connect different parts of systems.
  • Heterogeneous Integration—refers to the integration of separately manufactured technologies that in the aggregate provide enhanced functionality.
  • Heterogeneous Components —describes devices that do not necessarily scale according to “Moore’s Law,” and provide additional functionalities, such as power generation and management, or sensing and actuating.
  • Beyond CMOS—describes devices, focused on new physical states, which provide functional scaling substantially beyond CMOS, such as spin-based devices, ferromagnetic logic, and atomic switch.
  • More Moore—refers to the continued shrinking of horizontal and vertical physical feature sizes to reduce cost and improve performance.
  • Factory Integration consists of tools and processes necessary to produce items at affordable cost in high volume.
I have met ITRS Chairman Paolo Gargini several times during trips to the Industry Strategy Symposium and Semicon West. To say he is distinguished is an understatement:   (34 years at Intel speaks volumes) 

Here is a link to a presentation that is quite technical but gives some insight about industry transitions and how long it takes to get them to production-worthy levels.   On page 18 of the presentation there are some pictures of papers depicting FinFet research from '99. '00 and '01.    Page 28 gives an estimate of the incubation time for these transitions (~12 to 15 years!)

You don't have to listen very hard to hear the noise about who is going to be first with 14nm/16nm production.  From there the crowd shouts about who will be the first to 10nm and then 7nm and if you don't like that you can debate who will be the first to produce 3D NAND.   To me, this is really unfortunate because if one were to take the time to go through the confessionals of the past few weeks it would become clear that the bleeding edge, while certainly glamorous, is not the only segment of the chip industry that is enjoying growth.

Investors should make note of this......

Vital Stats from IC Insights

Foreword:  This post is going to need some editing/reorganizing but I wanted to get the charts up here for starters.  

The following charts are from the 1/2 day semiconductor seminar produced by IC Insights:

IC Insights is expecting 7% semiconductor industry growth in 2015.   

Plenty of food for thought here - much more to add.....

Wednesday, January 21, 2015

Knee Jerk Reactions

I'll get to the Sandisk earnings call this evening but for posterity I wanted to preserve this snapshot because it speaks volumes about members of the financial media.  

First, the news flow from Yahoo Finance:

The headline from The Street is absolutely priceless. 

Next, the after hours quote on the stock:

Yes, this was all on the same page. 

When Sandisk pre-announced a disappointing quarter a few weeks ago the stock got hammered.  Today the final tally was disclosed and even though they lowered the bar they still tripped over it.  

Clips from ASML's Earnings Presentation

I have yet to read the transcript of the ASML conference call but these clippings from the Earnings Presentation provide us with a few insights:

Memory spending ended this year with a bang.  Looks like the Korean memory manufacturers were pretty aggressive in Q4 of this year.   The Q4 '14 vs. Q3 '14 drop in IDM and Foundry purchases, as measured by End Use is striking - over 50% in both instances.   Extending this to other capital equipment players would suggest that good results are on tap for Lam Research and Applied Materials.     


Again, memory dominates the landscape in 2014.  705 million of the 2.225 billion spent on memory happened in Q4 or, 31% of 2014 spending by memory companies happened in the last quarter.  

While memory dominated actual spending in Q4, bookings shifted to the foundry sector.  The ramp of 14nm/16nm production   Flip-flop, flip-flop.

In the tech trade rags EUV gets a lot of attention even though it is unlikely it will be put to work before the 7nm node (even if then - see the earlier comments about thechatter from ISS).   In the interim, immersion lithography has much more work to do. 

The full presentation and some other bits are available at the ASML website:

Outlooks and Forecasts 2015

I am going to accumulate forecasting bits and post them here.  Once I've got a reasonable sample, I'll consolidate them so we can keep a tally.

Over at the Solid State Technology site there is a must read: 2015 Tech trends and drivers

From the Inbox: A complete copy of the forecast bulletin from IC Insights:

IC Market CAGR Rebounds But Growth Remains Below 30-Year Average
Strong mobile market, IoT not enough to spur stronger IC growth rate increases.

Figure 1 shows the IC market size at the beginning and end of each decade since the 1980s and the corresponding CAGR for each 10-year period. IC Insights recognizes the importance of the endpoints chosen for each CAGR and how they can easily sway the data to look very good or simply mediocre. For this figure, end points were selected based solely on the start and end of each decade.

• The 1980s were characterized by the relatively new but rapidly growing personal computer market and strong demand for computer memory like DRAM. IC market CAGR was nearly 17% in the 1980s.

• In the 1990s, Intel and AMD battled in the microprocessor space to provide the greatest PC MPU horsepower. Meanwhile, Microsoft introduced new operating systems every two to three years and each upgrade required a big boost in DRAM to run the systems efficiently. CAGR averaged 13.6%.

• The first decade of the new century was not kind on the IC market. Things started great in 2000 with 35% IC market growth, but two global recessions—one in 2001 due to the bust and the other due to the severe financial meltdown that started in 2008—prevented the market from gaining any traction. Continued flat or falling IC ASPs later in the decade resulted in the IC market showing a paltry 0.5% CAGR from 2000-2009. These setbacks started a trend of significant company consolidation within the IC industry.

• Following the recession in 2009, the IC market rebounded with 33% growth in 2010. Fewer suppliers and modest capital spending have had a positive impact on IC ASPs since then. While demand for standard PCs (desktops and notebooks) has cooled, the market for tablet computers and smartphones has remained hot. Sharing data on mobile/wireless devices and growing interest in the Internet of Things—connectivity to everything—will give a boost to IC market. Total IC market growth this decade is forecast to average 4.1% per year.

Over the past 30 years, the IC market has averaged 9% annual growth, but over the past 10 years, it has exceeded that 30-year average growth only one time (in 2010). Annual IC market growth rates are forecast to remain at or below the 30-year average (9%) through 2019.

Additional details regarding IC Insights’ 2015-2019 IC industry market forecasts are provided in The 2015 McClean Report.

Report Details: The 2015 McClean Report

The McClean Report—A Complete Analysis and Forecast of the Integrated Circuit Industry (released in January 2015) is IC Insights' flagship market research report, with more than 400 charts and graphs providing data on all aspects of IC industry. A subscription to The McClean Report includes free attendance to an upcoming McClean Report half-day seminar that will be held in the following locations:

• Sunnyvale, California on Thursday, January 22
• Boston, Massachusetts on Thursday, January 29

In addition to the seminar, a subscription to The McClean Report includes free monthly updates from March through November (including a 250+ page Mid-Year Update), and free access to subscriber-only webinars throughout the year. An individual-user license to the 2015 edition of The McClean Report is priced at $3,590 and includes an Internet access password. A multi-user worldwide corporate license is available for $6,590.

To review additional information about IC Insights’ new and existing market research reports and services please visit our website:

More Information Contact
For more information regarding this Research Bulletin, please contact Bill McClean, President, at IC Insights. Phone: +1-480-348-1133, email:

Rules for Technology Stock Investing (GS - Dan Benton)

Yesterday I stumbled across a page from a Goldman Sachs Investment Research Report containing Dan Benton's Rules for Technology Stock Investing

From what I can tell this was picked out of an older report.  If my sleuthing is accurate, Dan Benton was a top-ranked PC analyst at GS from 1988 to 1993.  From there he went to work at Dawson Samberg Capital Management which eventually became Pequot Capital Management.  I recall representatives of Pequot showing up on a few panel discussions I moderated when I was active with SEMI.  After Pequot, Mr. Benton founded Andor Capital Management.  Andor operates a long/short hedge fund.  

Suffice to say the history of Andor is interesting - it appears as though they shutdown in '08 and then reopened in '11.  This year it looks like the sailing has been rough.  Here's a link to a story from the WSJ:  Andor Capital Management fell 18% in March  A quick perusal of the news shows that Andor likes Tesla and Twitter....  FWIW

One thing about hedge funds seems overwhelmingly clear, they never seem to die - they open and close just like window shades (unless they are busted for violating the rules).

So, about those Rules.  The first six deal with Momentum. Rules 7 and 8 address Valuation. 9, 10 and 11 speak to Seasonality. 12 through 15 addresses Management.

  3. Positive earnings surprises occur when revenue and earnings growth are accelerating, when AVERAGE SELLING PRICES ARE RISING, and when gross margin and operating margin are rising.
  4. Most technology stock ideas are product-cycle stories.
  5. New product cycles often lead to earnings surprises; product cycle transitions usually lead earnings disappointments.
  6. Technology stocks also do well when COMPANIES REBOUND FROM POOR EXECUTION.
  7. Value investors don't make money in technology. There are few "cheap" technology stocks.
  8. Don't buy on relative P/E, P/B, P/R, particularly when estimates are falling (see Rules 1 & 2).
  9. Technology stocks perform poorly in the summer.
  10. Seasonal slowdowns cause secular concerns.
  12. Reorganizations without restructuring charges usually lead to earnings disappointments within two quarters.
  13. One quarter problems exist (but only if caused by supply constraints).
  14. Management usually appears weakest at the bottom of a product cycle.
  15. Insider selling doesn't matter; management gets new stock options every year.
Here's the actual image of the page (click it for a larger view):

Discussion is welcome and expected.  

Monday, January 19, 2015

The Upgrade Dilemma

One of my readers (there is more than one) sent me the link below because he knows I am in the market for a new laptop:

I replied:

Thanks for sending. It’s a formidable chip and a nice platform - even if it is from Dell. I don’t think it matters if someone – actually there is only AMD - tapes out a processor with similar or better specs because the issue of mass production looms ominously. Intel is the PC market.

To me, the most relevant comments are at the end of the article:

“But does CPU really matter at this point? What’s “good enough” in terms of CPU throughput? For that matter, what’s good enough for graphics? Intel and AMD seem to be dialing in on optimizing notebook performance well enough such that flat out performance isn’t the be-all and end-all metric anymore. It’s about performance-per-watt in most notebook designs, and at 14nm Intel’s design and manufacturing muscle is tough competition.”

I feel a serious case of Déjà vu. This debate, “How much CPU is enough?” has been playing out forever.

Yes, I am in the market for a new laptop. Up until this weekend I was seriously entertaining the idea of migrating to a Mac but there seems to be an increasing number of bugs proliferating in Mac software. Perusing the recent blogs of several Apple developers made it very clear that they are not happy campers. By no means am I suggesting that the latest version of Windows is anything to write home about but it is the platform I am the most familiar with.

The laptop I have today is a Sony Vaio running Win7 with an Intel Core3 inside. The machine, despite being ugly, bulky and loud, is almost six years old and still does the things I need it to do. The worst part about it is when it is unplugged the battery life is barely 2 hours – which is basically nothing. Before you ask, I have taken pains to back up the important files I have on this computer because I know it could crap out in a heartbeat.

I really am a cheap bastard when it comes to this stuff.

Saturday, January 17, 2015

Weekend Reading: ISS on EE Times

For many folks in the semiconductor business the year starts off with a trek to attend SEMI's annual Industry Strategy Symposium. I've lost track of how many times I have attended ISS. There are ISS presentations archived on my hard drive(s) extending back to the mid-90's. This year I chose not to attend - next year may be a different story. .

Rick Merritt of EE Times has penned a piece containing fodder that is worth a bit of discussion.

For those I communicate with frequently (you know who you are) you will find that many of the points made in this article germane to topics I refer to during our trading-day chat sessions.

Rick serves up his seven takeaways from ISS:
  • Samsung, GF winning 14/16nm race 
  • Slow Train to 3D Ics 
  • Slow Road to Vertical NAND 
  • Little said about EUV 
  • Little said about EUV Part II 
  • Living in the materials world 
  • Faster, higher, cheaper 
My $0.02:

In the semiconductor world nothing ignites a cat fight like the Samsung/GF/TSMC node race. Wall Street Analysts, Chip Industry Analysts, Industry Trade Rags, Self-proclaimed Pundits and Bloggers all know that a flashy headline - including exclamation point - touting a bleeding edge qualification/production order from Apple, Qualcomm or any leading-edge chip design house, will generate clicks and claws.

If you are investing in Samsung or TSMC knowing the winner of this every-other-year battle makes a difference. If you are investing in the capital equipment companies it doesn't matter because winner, and loser, still have to buy tools to stay in the game.

Yes, it’s that simple.

On 3D ICs: It is taking forever. By forever I mean way more than 10 years and I have the presentations from ISS and other events to prove it. <sarcasm on>

3D NAND : Today's aspect ratios (height and width - spaces and lines), as noted by my equipment industry colleagues, are outrageously challenging - hence the production delays we are seeing at the major memory houses. 

These are important comments from Micron's VP of process R&D:

“We don’t have the right characterization capabilities. [Currently] we have to use destructive techniques. We cannot measure thin films and other features hidden in the stack -- we need to do a lot of work in this area,” he said in a talk, citing stiff challenges removing as little as two microns of material using chemical mechanical planarization.

Worth repeating: "We cannot measure thin films and other features hidden in the stack -- ".

Can you say, "KLA-Tencor?"

Little said about EUV: IMEC submits that Immersion Lithography will extend to 7nm - - enough said.

Little said about EUV II: There are material challenges to be resolved . C'mon, that’s already too much!

Living in the Materials World:  I really wish there were some publicly traded, pure play, semiconductor material companies out there. Unfortunately, most of the materials business is buried inside larger companies. It’s a big business - comparable to the capital equipment industry:

Sidelight: The last time I attended ISS I had a chat with Intel's Tim Hendry. That conversation just happened to be the morning after the release of Intel's quarterly earnings. Intel was, once again, flexing its muscles but the stock was languishing. With the share price hovering around $20 even good news was perceived as a negative. ARM, TSMC and the design house minions were destined to eat Intel's lunch. The PC market was dead and Intel was following because they missed the mobile market. <sigh>

Before parting ways Tim and I agreed that the best strategy was simply a matter of exercising some patience. So far so good….

Sidelight II:  If you are interested in a funny story about Intel and a helium supplier drop me a note.

Faster, higher, cheaper: Quote from Boeing exec: "We study your industry a lot." 

Irony here, The little bit in this area fits with the conversation we've been having about chip manufacturing and how the internal knowledge can be extended to other industries. 

Comments are welcome and enabled,  Before they are published they will, to eliminate spam, be sent to me for moderation.   

Monday, January 12, 2015

Monday Food for Thought

What are companies in the semiconductor industry going to do now that they face soaring development costs, product commoditization and the end-game for a decades old manufacturing process?    

There's a dynamic I have mentioned in several, all-too-infrequent, missives and speeches over the past few years and it finally appears to be playing out in tech land.  This has taken years to build momentum and to this day I am sorely lacking a more cerebral or whizzy appellation to describe what is happening.   Boringly I have labeled this trend as one that moves companies toward the development of total systems. Because of this I believe the tech industry sits on the cusp of some serious verticalization. 

Let’s assume that soaring development costs (NREs) and weak pricing will flush a lot of the chip design houses and chip commoditization, due to lack of differentiation, grips the few players remaining at the leading edge.  Last but not least, let's go out on a limb and suggest that we are entering the final innings for Moore’s Law (did I just say that?).  If that is the case every aspect of the chip manufacturing process must, to differentiate their offerings and maintain viability, extend its purview beyond the construction of transistors.  Right now the pioneers in "system development" are marching out to the pins, the circuit board, through and to the bazillion points connected to the network - and back and forth and back and forth – ad infinitum.  Every aspect will be considered.  One could go as far to suggest that the meme of the day, the IoT, is a poster-child for this tipping point.

The signs of this happening are everywhere and a number of companies are embracing the transition. All the big names are all in play.   

If you are reading this you probably have insightful things to add so please feel free to chip in,    (To prevent spam comments will be moderated)

Call this, in addition to a new start for this blog, $0.02 for the jar.