AU Optronics (AUO) purchased tools from Nikon (7731.JP) worth $86.9m for what we believe will be phase 3 development of their Gen 8 L8B line in Taichung.  The last phase in this fabs development should add 30,000 sheets/month to total fab capacity, which is currently 90,000 sheets/month, with an additional 45,000 sheets/month from the L8A line, which has been developing displays based on IGZO[i] substrates.  We believe the L8B line is based on amorphous silicon, which continues to have the largest share of the substrate market although LTPS and Oxide substrate usage is increasing.

AU Optronics has gained traction in the large panel (TV) market with an early entry into 4K panels a few years back, and has continued to focus on the premium TV display segment and has moved from a swing producer[ii] to one with a reasonably diverse customer base, supplying Sony (SNE), Vestel (VSQ.GR), and a number of OEMs including Wistron (3231.TT).  AUO has been somewhat conservative over the years as to adding capacity, with financial constraints being an issue during the last display down cycle, but has recently expanded their L8B fab as noted above, and is expanding its LTPS based Gen 6 fab in Kunshan, China.  AUO is also considering a Gen 11 fab, but plans have yet to be finalized.



[i] Indium Gallium Zinc Oxide aka ‘Oxide’


[ii] When large panel producers such as Samsung Display are unable to meet customer demand, they buy from ‘swing’ producers on a relatively short-term basis.  Good for swing producers when demand is high, but they are the 1st to be cut out when demand weakens.

US wireless carriers will be offering pre-orders for LG Electronics recently announced G6 smartphone starting today, with expected deliveries on April 7.  The phone, which is said to be a hit in Korea where it was released on March 10, is a make or break product for LG, who has seen disappointing sales for its smartphones for the last few model years.

Carriers are offering a variety of incentive to customers who pre-order, with AT&T (T) offering a 2nd G6 when you add a second line (credits needed), a Google (GOOG) Home voice activated speaker (a $129 value according to AT&T) and an LG Sport Watch (valued at $49.99) if you sign a 2 year contract.  Not to be outdone, Sprint (S) is offering a free 49” LG smart TV if you are a new customer ($349.99 value), and Verizon (VZ) offers a free 43” LG smart TV, with both offering the Google Home incentive to all.  Maybe they will tell you how to get the phone to drop all calls from the ex-wife if you ask…

Skyworth (751.HK) has lowered the price of its 55” OLED TV by over 20% to $1593 as of this week.  Current pricing is lower than typical LG 55” OLED TVs in the US, which are $1999, with a few on-line retailers offering as low as $1599.  While we cannot be sure that the model being offered is the correct one, Skyworth announced back in November ’16, that it intends to use OLED panels developed by BOE in their OLED TV offerings, and while BOE is still in very limited production, a price decrease could indicate that they have switched from LG Display (LPL) to BOE for their 55” OLED panels, or that BOE has seen improving yields at its Hefei OLED pilot line.

LG Display will still hold a very commanding lead in the OLED TV panel production market, likely over 90% this year, but should BOE gain some traction with local TV brands, it will change the pricing situation for  and put outside pressure on LGD’s OLED panel pricing.  We note that while OLED material supplier Universal Display (OLED) will reap the benefits of any capacity increases in OLED TV manufacturing, and it already counts BOE as a customer, UDC derives royalty income from LG Display based on the price sold to customers (primarily LG Electronics (066570.KS)).  This means the balance between higher unit volumes and lower unit pricing could become a significant variable in UDC’s royalty income line.  Samsung Display’s (pvt) agreement with UDC is not based on price, but reflects a previously agreed on yearly license fee that escalates each year.

According to a poorly worded exchange release, BOE (200725.CH) has received loans worth $65m from the State Development Bank of China and a smaller consortium of other financial institutions, to be used toward the construction of what we believe to be the 2nd and 3rd phase of the company’s Gen 5.5 OLED fab in Ordos, Inner Mongolia.  While we are unable to confirm the dollar amount of the 8 year loan, the amount here would be only a portion of the amounts needed for the fab’s construction although the City of Ordos, along with the State Development Bank has recently signed a $145m loan to BOE for the same production fab.  The fab itself is collateral for the loan.

BOE has been aggressive in developing OLED assets, with a Gen 8.5 pilot line in Hefei, a small Gen 5.5 line running in Ordos, and construction of Gen 6 OLED fabs in both Chengdu and Mianyang.  The Ordos fab is expected to have capabilities for 54,000 sheets when fully built out, with fully built-out capacity at the other two Gen 6 fabs reaching ~93,000 sheets/month.  If BOE completes these plans, we believe it would give them a 5.5% share of the OLED market on an m2 basis by the end of 2020.  While this is still small relative to Samsung Display and LG Display, they would be the 3rd largest OLED producer by capacity at that time.

While BOE has been extremely aggressive as to adding OLED assets, they have had a difficult time with OLED production, particularly with yield.  We believe this points to a rather large stumbling block for Chinese OLED producers, a lack of experienced OLED engineers.  BOE has poached OLED engineers from Japanese OLED developers, and sponsors a number of Chinese university programs for R&D, but the lack of experienced production engineers has slowed the growth of OLED manufacturing in China.  There are producers in China, with Everdisplay (pvt) and Visionox (pvt) both having some commercial display product, but the large capacity panel producers such as BOE and China Star (pvt) have had limited commercial success.  We do expect these issues to be overcome eventually, and as the OLED display market moves toward flexible displays, where the playing field is slightly more level, Chinese OLED manufacturers should be playing less of a ‘catch-up’ game than currently.

LG Innotek (011070.KS) has decided to spend $233.4m to increase its presence in the camera module market.  As smartphone cameras become more sophisticated and new applications for the automotive, drone, and security markets continue to evolve, the need for capacity continues to increase.  While camera modules are the end product, much of the growth is in the CMOS sensor market that supplies to module makers.  While smartphone cameras are the largest market for CMOS image sensors currently, the automotive market will likely see the most growth through 2020, as cameras replace mirrors and automotive systems such as collision avoidance and other semi-autonomous systems become more commonplace.  The CMOS sensor market is expected to grow from $9.9b in 2015 to $15.2b in 2020.

Some camera module manufacturers are semi-captive, such as Samsung Electro-mechanical (009150.KS), who saw over 61% of its business going to Samsung Electronics (005930.KS) or affiliates in 2015, or Sony (SNE), with a number of independents like Truly (732.HK) and Primax (4915.TT), while Apple (AAPL) chose to buy Israel based LinX (pvt) last year for $20m, to move forward in the multi-aperture space that is becoming a standard for high-end smartphones.  While ‘camera competition’ will see each successive model year with more sophisticated cameras ($7.3b by 2020), module producers will continue to expand their operations and likely smaller players will be unable to compete, however if the automotive market grows ($2.2b by 2020) as indicated, smaller module producers could tap into particular niches where they could maintain long-term relationships with automobile manufacturers or suppliers.

Recent articles in the Taiwan and Chinese technology press have shined light on the idea that smartphone vendors will be capacity constrained when adding OLED display-based smartphones to their brand offerings.  Recently cited was Huawei (pvt), who will have 8% of its line using OLED displays in 2017, making the assumption that they had expected to be using a greater percentage. 

Is this something new? Not really, as even when Samsung Electronics (005930.KS) was the only smartphone brand using OLED displays, they faced OLED display shortages that caused delays in shipments and apologies to potential buyers.  While Samsung Display the primary supplier of OLED displays for smartphones continues to expand its OLED production capacity, their largest customer remains Samsung Electronics, their parent, which makes them a semi captive supplier.  In fact, this issue was responsible for the delays seen in adoption by other smartphone brands, who knew that they would be in a secondary position to Samsung Electronics, holding many from adopting the technology.

That all changed over the last few months when the industry concluded that Apple (AAPL) would be adopting OLED displays for its iPhone line this year.  Suddenly brands rushed to adopt the technology to keep themselves at least in line with Apple, and the display industry responded with announcements of new OLED capacity plans and accelerated timelines, but the question remains, ‘Will the industry be able to meet demand?’  Raw capacity (see Fig.1), the measure most used to evaluate the supply/demand balance in the display space, is a poor metric for understanding what the industry can actually produce, and while the growth of capacity looks just like investors would want to see, it does not correlate to industry results.

Figure 2 shows a comparison between the ‘raw’ capacity and ‘utilized’ capacity, which includes a number of variables that are far more relevant to actual industry results that the ‘prettier’ raw metric.  Included in the utilized numbers are ramp times, experience of the producer, equipment tuning, and other variables, all of which come before the most important product level variable, yield.  Now that the industry has decided that another major player will be entering the OLED display demand side, these simple constraints become far more important to brands that might be in contention with Apple for what is already limited OLED display capacity.

While we will not detail our conclusions here, we will add a monkey wrench or two, just to let investors understand how complex these issues are.  First, OLED industry capacity is divided between small panel production fabs, such as those run by Samsung Display, and large panel OLED fabs, such as those run by LG Display for its OLED TV production.  This reduces capacity available for smartphone production on both a raw and utilized basis, as those fabs designed for large panel OLED production cannot be used for OLED smartphone production in most cases.  Second, drilling down a level further, what format will be used for each OLED smartphone brands?  Will they be built on rigid substrates or flexible substrates?  Figure 3 shows that raw OLED capacity is further defined by the type of substrate used, limiting capacity further as the industry shifts its focus from rigid to flexible OLED smartphone displays.  Fret not, as we have built our OLED industry model to reflect all of these variables, which allows us to gain significant insight into Apple’s potential path toward OLED smartphones and the opportunities available to other smartphone brands as they add OLED displays to their smartphone lineups.  Just the tip of the iceberg…more to come. 

SFA Engineering (056190.KS) has extended its contract with Chinese panel supplier BOE (200725.CH) from March to May for the delivery of tools worth $32.8m US.  SFA supplies a variety of display production tools ranging from glass handling equipment to OLED evaporators, coaters, and sputtering tools.  SFA has been under significant pressure to meet the needs of the rapidly expanding OLED production space as has its competitor Nikon Corp (7731.JP), as they both supply tools that are key to the OLED production process.  Extended backlogs have been commonplace for these tools for a number of years, with such issues increasing as new entrants try to ramp commercial production.  We would expect the contract extension is a result of delivery dates being pushed out by the vendor rather than the buyer, as they struggle to meet delivery dates, which is why we take a somewhat more conservative view of the ramp in actual OLED capacity.  Fab announcements are made and press releases show big factories, but if the equipment was a few months late, the actual production from the fab will also be delayed.  We call it ‘utilized’ or ‘available’ capacity, rather than the often cited theoretical capacity.

SFA has been involved with Samsung Display’s (pvt) A3 OLED expansion project (~$47m), China based Visionox’s (pvt) Gen 5.5 OLED fab expansion ($93.5m), and BOE’s Gen 6 OLED project, and has likely been unable to meet exact delivery dates for all concerned, similar to how Nikon has had to eliminate some customers in order to allocate resources to major projects, such as the recently announced LG Display (LPL) Gen 10.5 fab and an unspecified Sharp/Hon Hai project which will use Nikon photolithography tools.  As these tools are usually customized, they tend to have lead times of at least 9 months, and potentially as much as 18 months.  While we might assume that the BOE Gen 6 project might be delayed by 60 to 90 days, the overall OLED capacity timeline will not move much as we have built an extended ramp into our capacity model for those OLED producers who have relatively little high volume OLED production experience.

While we have noted a number of comments from the Japanese press on the potential construction of a TV plant in the US by Sharp (6753.JP), it would seem that Tai Jeng-wu, the CEO, has made the definitive statement concerning the matter.  He stated that the company has not made any decision as to whether such a plant will be built, and more to the point, if there is a policy from local governments to encourage, they will advance the plan. 

Simply, the company, and in this case the final decision will likely be made by Sharp’s new parent, Foxconn (2354.TT), must receive enough incentives to make the project financially worthwhile.  We believe this would entail significant long-term tax incentives at the least, and likely some additional infrastructure incentives relating to power, transportation access, and similar resources, not unlike the incentives given by regional Chinese governments to encourage companies to locate in such regions. 

While the US trade organizations complain about the unfair competitive nature of Chinese companies due to government support, the same type of incentives will likely be offered to Sharp/Foxconn, but in this case they will be couched in political rhetoric.   Deals are always made to ‘encourage’ growth, maybe we should stop whining and encourage US companies to innovate, retrain, and invent, something we seem to be better at than most others.  After all, we did invent the telephone, light bulb, the airplane, washing machines, television, microwave ovens, industrial robots, lasers, integrated circuits, PCs, microprocessors, and social media (ok, maybe not a great last choice).  

Following the curtailment of large panel shipments to Samsung (005930.KS) and Hisense (600060.CH), Sharp (6753.JP) now says the idea is to keep its IGZO[1] technology in-house.  Citing the continuing competition from other TV brands, Sharp’s new management and owners are trying to end all supply relationships that entail Sharp’s IGZO backplane technology.  While Samsung and Hisense were likely the biggest customers, the company still has an agreement to supply the technology to Chinese panel producer CEC Panda (pvt), who uses it at its Gen 8 fab in Nanjing, and potentially at two Gen 8.5 fabs in Xianyang and Chengdu that are under development.  Sharp says it is reviewing its JV agreement with Panda and will continue to supply the technology if contract conditions are met.

Sharp plans to use IGZO as a differentiating technology for its own brands, stating, “There is no IGZO technology at other companies, it is Sharp’s advantage, I do not intend to supply it to other companies.”  Understandably, Sharp is protecting its assets, but we believe, aside from Panda, both AU Optronics and BOE (200725.CH) have large panel IGZO capabilities, and while Sharp is known to be the leader in commercializing the technology, we would expect BOE at least, to capitalize on its ability to produce IGZO displays.  We believe that the technology represents ~8.3% of overall display capacity currently, and while Sharp and JV partner Panda are in production, BOE has the capability, if they fully utilize the technology across available production platforms, to be the IGZO share leader.



[1] Indium Gallium Zinc Oxide (aka ‘Oxide’)

Beijing based Xiaomi (pvt) was once the top smartphone brand in China, using innovative marketing and low-priced but ‘almost’ fully featured designs to lure Chinese smartphone users to its brand. Known as ‘the Apple of China’ the company was touted as the hottest smartphone vendor on the mainland, but saw its star dim as newer brands copied and then improved on the Xiaomi model.  While the company was the share leader in China as late as 1Q 2015, they were replaced in the top position by Huawei (pvt) by the end of that year, and have struggled to stay in the top 5 Chinese brand category since then.

In early 2016, the company began a program that focused on expanding its hardware ecosystem, investing in a number of startups that can produce devices that can be operated through a Xiaomi smartphone and are sold through the Xiaomi stores across China.  These devices range from rice cookers to fitness trackers and all carry the Xiaomi brand, while priced below traditional or well-known brands with the goal of serving consumers aged 17 to 35 in 2nd and 3rd tier cities.

Since then Xiaomi has made investments in 77 hardware companies, which have generated $2.17b according to the company founder Lei Jun, with 30 already marketing products, 16 generating over 100m CNY ($14.5m US) per year, and 3 companies generating over 1b CNY ($144.7m US).  The ecosystem product line has expanded to include household robots, VR/AR, air purifiers, drones, headphones, TVs, STBs, and smart automotive devices, all of which are able to be controlled by Xiaomi smartphones.  The company is planning to expand its investments to over 100 companies over the next 5 years.

While Xiaomi will likely continue to struggle against Chinese smartphone competitors such as Huawei, Oppo (pvt) and Vivo (pvt), they have a chance to grow their overall business as they continue to expand their ‘smart’ offerings.  Of course, much will depend on their ability to choose investments in potential products that appeal to Chinese consumers, and the recent loss of Hugo Barra, a former Google product designer and well-known spokesperson for Xiaomi to Facebook’s (FB) VR effort, will change the public face of the company, but wise choices and an ever expanding product line will help to establish the company as a leader in China’s consumer IOT market, which a number of well-known CE giants are trying to enter.