PCBs, or printed circuit boards are the mainstay of the consumer electronics industry and have been around on a commercial basis since the 1950’s, but the advent of flexible circuit boards has pushed the industry to develop new methods and concepts to make PCBs work in even smaller and more compact mobile devices.  Originally single layer boards, flexible PCBs can now have many layers connected with metal coated ‘vias’ or holes that connect various parts of multiple layers, but one characteristic that continues to drive the PCB industry is ‘smaller’. 
The vias mentioned above were copper plated holes .015” only a few years ago, and were drilled with precision drill bits that lasted a short period of time due to stress and breakage.  Vias are now as small as .001” and are created with laser drills rather than physical bits, and traces, the lines of conductive material that connect components on a PCB board have moved from .01” to .002” (that’s 1/100 of an inch to 2/1,000 of an inch!), while components with dimensions as small as .015” x .008” are machine-placed on boards at multiple parts per second rates.  Lasers are also used to image the traces and other markings directly on the PCB board prior to etching, eliminating the 10 scale drawings that were commonly used to pattern boards only a few years ago.
Flexible boards themselves, which are still made of polyimide, a polymer that has excellent thermal characteristics and chemical resistance, can now be laminated together without adhesives that can weaken when heated or flexed, but the drive to create new PCB board level products is gaining momentum as demand for mobile products increases, and PCB manufacturing processes continue to evolve.  This pushes PCB, and particularly flexible PCB suppliers to constantly upgrade tools to further automate manufacturing and remain competitive in what is a $13.5 market that will grow at a CAGR of 11% through 2025 to $33.4b[1].
During the last 15 years, the PCB industry has moved from Europe and North America to China, given its low labor cost, weak environmental regulation, and significant government support, however automation and new in-line process equipment can level the playing field as headcount requirements decrease rapidly, with traditional factory headcount of 38 and an automated in-line process headcount at 7.  When paired with wastewater and chemical recycling systems, cost advantages of 50% can be reached, even at low utilization rates, with ROI’s at less than 3 years.
More to come on the state of the PCB industry….


[1] TMR Research 3/2017

According to Digitimes, Sharp (6753.JP) is expected to expand large panel capacity at two of its fabs in Japan.  Sharp’s G10 fab in Sakai is expected to increase production from 72,000 sheets/month to 78,000 sheets/month starting in 2Q, while the company’s Gen 8 fab is expected to increase production from 50,000 sheets/month to 60,000 sheets/month by 3Q.  The Gen 10 fab is expected to be focusing on 60” and 70” TV panels for Sharp’s brand and Foxconn’s (2354.TT) ODM customers, while the Gen 8 fab will be focusing on 45” and 55” TV panels for internal Sharp use.
According to our calculations, the Sharp Gen 10 fab would see an increase of 8.3% of available capacity, which if using a ratio of 66% 60” and 33% 70” would see an increase in production from 384k 60” and 144k 70” TV panels to 416k 60” and 156k 70” panels respectively.  On a yearly basis that would be an increase from 6.335m to 6.864m panels in total.  At the Gen 8 fab, the increase would be 20% in terms of m2 and on a yearly basis would indicate an increase in unit volume from 3.6m to 4.32m combined 45” and 55” units (3:2 ratio).  Sharp is expected to use the increased capacity at their Gen 8 fab for internal brand use only.
Our display industry model has already built in the Gen 8 line increases, but the Gen 10 fab increases are equivalent to a 2.86% increase in Sharp’s overall capacity for 2017.  As noted a number of times previously, Sharp has discontinued supplying TV panels to Samsung and Chinese brand Hisense (600060.CH), in order to focus on developing the Sharp brand, as these changes would reflect.  That said, these increases will remove the necessity for Sharp to have to buy panels away from its own production to meet its internal brand goals this year, which will, to a small degree, reduce demand for those panels (mostly 45”) that are in short supply, but given the relatively small increases at Sharp, we doubt the changes will have a significant effect on the overall supply/demand balance.  What will have a significant effect, is the construction of Sharp’s Gen 10.5 fab slated for September 2019 (phase 1) and phase 2 construction finishing in 2020, as can be seen in Fig 6.  This fab will also focus on large panel TV display production, likely augmenting the increased capacity at the Sakai Gen 10 fab.  If Sharp’s new owners are able to resurrect the Sharp brand worldwide, they will have ample supply by 2019/2020, if not, and the fab is still built, there will be an increasing amount of very large TV capacity available to the overall market.

​Normally, pictures of display fab construction sites are few and far between, surreptitiously taken by employees or from satellite images, but China Star’s (pvt) Gen 11 fab, which we noted back in April 2016 was under development, and updated at the beginning of the month, has once again appeared in the local Chinese press.  Given that this project, which will use the largest glass format to date (just under 10 m2 or ~107 ft2, and could actually be larger, given that no real “Gen 11” format size has been established.) is also the largest industrial project in the city’s history, with an area of 600,000 m2 (a football field is 5,351 m2) and will cost $6.75b, it seems to be getting an unusual amount of attention, which allows us to more closely monitor the construction progress.
As previously noted, he fab will be based on IGZO backplane technology, China Star has yet to decide if some of the 45,000 sheet/month capacity will be allocated to producing large panel OLED displays, which are currently only produced by LG Display .  As we believe the fab will be built out in three phases of 15,000 sheets/month each, we would expect the phase 1 build out to be LCD based, with the potential for some OLED capacity in phase 2 or 3.  China Star is running a small Gen 4.5 OLED pilot line, which could lay the groundwork for a large panel OLED line in 2020 or 2021.  China Star is owned by TCL Corporation (000100.CH) and Samsung , who has a 10% share and received ~14% of its panel supply from China Star last year.

You bet they are.  According to the Korean press, Samsung Electronics showed a number of new display technologies to a number of selected customers at this year’s Mobile World Congress in a private room setting.  Among those technologies privately shown are an OLED display that does not use a polarizer[1] (which would increase the brightness of the display), a very high resolution VR headset, and a ‘blue-less’ OLED display, all of which are expected to be commercialized within one to two years.  The company also showed a force touch system, a technology used by Apple in a number of their mobile products.  The Samsung force touch system is expected to be incorporated in the upcoming Galaxy S8, but the article cited viewer’s comments that the system did not look ready for production.
While private room viewing is commonplace at most CE shows, Samsung did not show its foldable phone to potential buyers, although it was shown to a small number of potential telecom customers, but the high resolution VR display seemed to be the closest to release.  The device shown had a resolution of 1,200 ppi[2], against the current version that has a ppi under 1,000, with a 1,500 ppi device promised.  Samsung has since announced the product (1,200 ppi) but without a release date.
Samsung has many display projects under development and is careful as to what is shown, so the idea that they show some of these potential products to selected customers is not a new one.  The company periodically updates customers to keep them interested in such potential products, even if they are years away from commercialization.  Samsung has selectively shown a holographic display despite the fact that it is many years away from production, if at all, but it tends to be the most popular display whenever it appears.  Not being able to reach that carrot makes you want it more…


[1] Polarizer – aka anti-reflective film

[2] PPI – Pixels per inch

The JD Power “2017 Full Service Smartphone Satisfaction Survey” ponders whether your smarthome is driving your smartphone satisfaction (the answer is yes if you care) but we were more curious as to whether issues with the recall of the Samsung (005930.KS) Note 7 last year had changed users satisfaction with Samsung products, particularly smartphones.  According to the 2017 survey, Apple (AAPL) edged out Samsung, by 0.12% (or one point on a 1,000 point scale), the same one point that Apple beat Samsung with last year.
That said, the differences between the two years were very small, and the satisfaction rating for both Samsung and Apple increased marginally, while the overall satisfaction rating for smartphones increased by 1.2%.  With both Apple and Samsung’s y/y ratings increasing, it would seem the Note 7 issue had little effect on user satisfaction, although the survey is taken from those who have already made a purchase, so it does not reflect ‘avoidance’ or those that switched to a competitive product.
So we like our smartphones a little more this year than last, but why?  According to the survey, the category that received the biggest increase y/y was ‘Satisfaction with the phone’s features’, which was up 9.4%, indicating that the quality of features (and we assume pre-packaged applications) has improved, but what about the original question concerning whether smarthome products were making you more satisfied with your smartphone?  The survey says, “Customers using their smartphones daily to control their connected devices have higher satisfaction than those who use it less frequently.”  So does that mean in order to be more satisfied with our smartphones we need to add smart thermostats or smart refrigerators to our houses?  We would rather go with applications that don’t freeze, drop calls, drain the battery, gather personal information, or just stop working, than having to buy a new refrigerator so we can see if the milk has expired from a remote location, but that’s just us. 

We have already commented on February panel pricing, but we have filled in our data for the display space as to broad unit volumes and individual company results as seen below.  We note that y/y results reflect the severe downturn in panel pricing seen early last year.

While no promise had been made as to when Samsung’s management would decide to create a holding company structure and potentially list the stock outside of South Korea, it would seem that it will not be happening soon. At the Samsung Electronics annual meeting, the company’s CEO indicated that the review started last November has revealed some ‘adverse implications’, and that the review would continue, dashing investor hopes for a quick resolution (if you can call 6 months quick).  While the difficulties cited at the meeting are what the board is using to justify the delay, it seems that the arrest and trial of family heir apparent and vice-chairman Lee Jae-yong, has put a hold on much of the restructuring ideas that he was championing, although the board promised to set up a new governance committee made up of independent outside directors by the end of April.  Thus far the board has been unable to find such foreign directors given the ‘uncertainties’ surrounding the company.  Any takers?

AU Optronics (AUO) has purchased $61.8m worth of tools from Hitachi High-Tech Fine Systems (8036.JP), continuing their aggressive upgrading and expansion of their Taiwan and Chinese display production facilities.  Hitachi Fine Systems produces both glass inspection systems for small and medium LCD and OLED production, and exposure tools for color filter production.  In this case it is difficult to figure what fab these tools are destined for, but regardless of the destination, they continue to represent AUO’s push toward increasing capacity at their L8B line in Taichung, and their Kunshan, China Gen 6 fab, where we expect additional capacity to come on line later this year.

On 2/3/17 we noted that Plainview based Veeco (VECO) has signed a definitive agreement to purchase Ultratech (UTEK) for $815m.  San Jose based Ultratech is a semiconductor tool supplier with significant history in the photolithography business (~85% share) and laser annealing with the largest exposure to advanced semi packaging and front-end semi, while Veeco has exposure to advanced semi packaging, lighting, and data storage through its MOCVD and MBE tools.  The deal is expected to close in 2Q and will be financed by both stock ($200m) and cash, with Veeco entering the 4th quarter with $336m in cash and ST investments and has recently completed a $335m convert deal, while Ultratech ended 2016 with $267m in cash.
Sounds reasonable, but last week a group of Ultratech investors filed a class action suit in the US District Court of Northern California to block the transaction.  The complaint, which is focused on a violation of the Securities Exchange Act of 1934, alleges that the preliminary proxy statement filed by Ultratech omitted material with respect to the transaction, violating Sections 14(a) and 20(a) of the act.  Simply, the complaint states that the participants agreed to a ‘no solicitation’ provision that prohibits Ultratech shareholders from soliciting alternative proposals and constrains their ability to communicate and negotiate with potential buyers, and, of course, the $26.5m that would have to be paid to Veeco by Ultratech, if Ultratech shareholders were to cause the deal to terminate.
Additionally, the suit states that the financial analysis made by Merrill Lynch, Ultratech’s financial advisor, showed a value as high as $35.90, and a DCF of $31.65, a bit higher than the deal’s $28.64/share valuation, and questioned why such information was omitted from the proxy.  The suit also brings into question whether the officers of the company, who would receive $17.6m as a result of the proposed deal, have a conflict of interest in the matter, and under what circumstances Ultratech’s management discontinued discussions with other suitors or strategic buyers. 
Statistics concerning merger deal litigation dating back over 15 years indicates that ~ 12% of merger deals involve class action litigation (although ~90% of major deals involve litigation), with the eventual outcome being a 5.8% decrease in deal completion on average, and a 9% increase in takeover premium.  However in recent years the courts have taken a less favorable view of such suits by legal firms that matter-of-factly litigate mergers, commonly known as ‘litigation kennels’.  By not routinely approving settlements, the courts, particularly those in Delaware, have shown law firms that it might not be in their best interests to litigate every major deal, with the lawyers hope of quick (and less expensive) settlements being eroded by the courts.  How the Veeco/Ultratech suit proceeds might help to set the tone for other merger suits in the California courts.

MLS Co. (002745.CH), China’s largest LED packager has agreed to form a joint venture with Rhode Island based Lighting Science Group (LSCG), with MLS providing lower cost LEDs to Lighting Science’s US LED bulb product customer base. MLS will invest $4.9m for a 49% stake in the JV, with LSCG owning 51%.  Lighting Science has been a private label supplier of LED bulbs to Home Depot (HD), although in 2Q ’16, HD elected to purchase bulbs directly from Asian suppliers, reducing LSCG’s revenue stream, but allowing them to pursue other big box retailers, which was prohibited under the old agreement.  The agreement with MLS could give LSCG a bit more cost control over its products, making them more viable to new potential customers.