AU Optronics (AUO) has been in and out of the OLED space for a number of years, both investing and then reducing their OLED development efforts as financial constraints varied.  Originally focused on smartphone sized displays, the company was not able to generate acceptable yields and refocused their OLED effort toward specialized OLED displays for wearable and VR last year.  The company continues to produce samples and small commercial product runs at their Gen 4 Singapore fab, which they share with LTPS backplane production lines and have a small R&D line in Kunshan, China that is also shared with an LTPS production line.

Recent reports that AUO has improved product yields implies that they are refocusing some OLED production assets on smartphone sized displays, particularly 5” and 5.5” models which have become available on a number of 3rd party vendor sites with capacitive on-cell touch.  AUO was an early producer of circular OLED displays for wearables (see below) but has seen mixed results in the smartphone space, announcing a 5.7” 2560x1440 pixel OLED display in 2014 (very high resolution for the time) but seemed to be unable to reach manufacturing yields above 50%, keeping the firm from entering mass production.

Hopefully AUO has reached production yields that will allow them to be profitable in their OLED efforts in the near-term, but we believe they need to expand capacity if they are to compete in the OLED space in anything other than niche products.  The company is currently expanding its Gen 8 LCD lines in Taichung, Taiwan and its Kunshan fab in China, which has an OLED R&D line as noted above.  The company has also mentioned the potential for a Gen 10+ fab, but has no specific plans for such a large format fab currently.

Israel based Orbotech (ORBK) received a $5m order for multiple direct imaging tools from an unnamed Japanese electronics manufacturer.  As the company has averaged $29m in PCB product sales for the last 4 quarters, the sale represents ~17% of an average quarter’s revenue and was customized to meet the specific requirements requested by the customer after a number of months of product testing. 

The sale consisted of both the Paragon™ Ultra and the Nuvogo™ 1000R tools.  The Paragon Ultra can image lines down to 8μm and depending on the application, can run up to 110 boards/hour and the Nuvogo 1000 tool series is designed for use in flexible solder mask imaging using a multi-wave laser that allows for highly accurate line structures with minimal non-uniformity.  Both tools are necessary for fine line width boards used in typical mobile devices where space is at a minimum and registration accuracy requirements are as high as +/-10μm.

We believe Orbotech continues to lead the market in the PCB space, which is one of three product categories where it provides tools.  The company has been quite aggressive in designing new tools to meet the requirements of advanced board designs that are necessary for products, particularly mobile products that have become more complex and now flexible as part of current high-end smartphone designs.  Orbotech now provides the only PCB inspection and repair tool that can provide both additive and subtractive repairs, which allows board manufacturers to automate processes that would have been done manually in the past, saving dollars and improving yields.

TV panel prices rose significantly in 2016, and while panel producers benefitted, TV brands suffered as the higher panel price eroded their gross margins and gave them less flexibility to discount during holiday periods.  China, which has been the driver for TV growth during the last few years, saw TV unit volumes grow only 7.9% to ~51m units, despite relatively strong sales early in the year, but the impact of higher panel prices on TV set pricing was felt during the Chinese holiday season, which saw an 11.4% decline in unit volumes y/y during the Chinese New Year and Spring Festival in late January and early February.  The cause of the decline is considered to be the lack of discounting and overall higher TV pricing.

Expectations for the Chinese TV market this year (2017) are for a decline of ~2.8% in units (1.7m units) making the sensitivity toward TV panel prices extremely high.  A continuation of price increases could further damage the growth of the Chinese TV market, and while brands and panel producers all focus on large screen sizes and 4K/HDR premium oriented sets to maintain margins and dollar volume, the industry, particularly China is facing negative growth as a result of the TV panel price increases.

Would we expect panel producers to altruistically lower prices to help brands generate profitability?  Not really, especially as they were on the other side of the ‘seller’s market’ for quite some time, but as TV set producers struggle to burn through inventory and lower expectations for quarterly and yearly shipments, a lack of buyer enthusiasm might spark a little price competition among panel producers, who need to keep fabs at near 100% utilization or see rapid declines in margins.  Should that occur, and we believe it is a possibility for those TV panel sizes that have not seen increases from artificial supply issues, TV brands might have a chance to see profitability improvements later this year, with the tradeoff being lower margins at panel producers.  There is potential for a balance, with panel producers still seeing profits, albeit not peak level, and brands seeing margin improvement, but it is a fine line that is rarely walked in the display space. 

Back in early February we noted that the Korean press was indicating that Samsung Electronics (005930.KS) and Samsung Display would be showing a prototype of a foldable OLED smartphone at Mobile World Congress to selected customers, and not the general public.  We noted that we did not expect that such demos (and we still do not know if or to whom they were shown due to NDAs) would lead to an imminent product, adding that we expected another 9 to 12 months before being able to qualify and scale manufacturing for such a product.

The Korean press is back with the notion that the two partners will be producing a flexible smartphone prototype, but this time by 3rd quarter of this year, with the delay coming from further development toward improving the appearance of the device, making it ‘one-step higher’ than its regular flagship smartphone.  No mention of the difficulties involved in manufacturing flexible displays, the amount of available flexible capacity, the risk of some catastrophic issue that might further tarnish Samsung’s reputation, or the idea that such a device might not have appeal to consumers, but that the idea was to make it more attractive.

While the press goes on to cite competition from other potential flexible OLED device producers, many of whom have shown flexible OLED displays over the last 12 to 18 months, it does acknowledge that Samsung has considerably more experience in both high volume production of OLED displays, and ‘flexible’ (actually conformed) displays generally.  We don’t doubt the prowess shown by other potential OLED producers in developing flexible display prototypes, but moving from building a few demos in an R&D lab to even producing on a small pilot line is a major step which few have both the resources and the expertise to accomplish.  This gives Samsung et al a distinct advantage, should they decide to make such a move, which they will continue to use as a carrot to keep consumers focused on ‘potential’ Samsung products as the new Apple iPhone is released later this year.

While Apple (AAPL) has yet to decide what iPhone models it will release later this year, speculation continues on every aspect of the potential devices. According to Taiwan based Digitimes, the iPhone variant that is expected to be based on an OLED display will use on-cell (‘Out-cell’ – their words) thin film touch sensors, with the sensors themselves being supplied by Japan based Nissha Printing (7915.JP), and TPK Holding (3673.TT) and General Interface Solution Holdings (6456.TT) and TPK continuing to supply more traditional touch modules for the other variants.  According to ‘sources’ TPK will supply 10% - 20% of needed TF sensors while Nissha will supply 80% to 90% at the onset.

While specifics as to the type of sensors and modules was not given Nissha supplies both capacitive and resistive sensors and force sensors that detect touch pressure.  Resistive sensors are relatively simple structures (see Fig.1) and capacitive sensors are slightly more complex (see Fig.2) but thin-film comparisons against glass based sensors and modules have somewhat more viable metrics for use in mobile devices, with a thickness 61% of glass, and weight of 34% of glass[1], but the general thinking here is that the touch will be an ‘add-on’ module, not the far more complex ‘in-cell’ touch that is built into the OLED display module itself.  While we would expect Samsung Display (pvt), the supplier of the OLED display module, to want to capture the higher value ‘in-cell’ process, rather than have it added by a 3rd party supplier, in-cell touch technology is still relatively new to high volume manufacturing, and as such adds to the potential issues that might occur during the OLED display production.  Apple is already putting itself at risk with a new technology and a new supplier, so it makes sense that they would opt for the less risky technology.

Almost as important however, is if the use of thin-film sensors is correct, it could lead one to think that the necessity for such materials would indicate that the device will not be shaped in a typical smartphone shape, with the film indicating either a rounded or flexible configuration.  While it would be encouraging to imagine such a breakthrough device, it ain’t necessarily so.  The film sensor modules could be used toward a conformed edge, as is the case in the Samsung Galaxy line, which would make the application of the touch module far simpler than depositing sensors on a conformed surface, so while we are encouraged by the use of TFS, we don’t assume it means something other than a potentially conformed device.



[1] Metrics supplied by Nissha

Tianma (000050.CH) announced that it will restructure a portion of its display holdings through a non-public offering where companies of China Aviation International (pvt) (aka AVIC), a state owned aerospace and defense company, will their portion of Xiamen Tianma, the display producing assets owned by a JV between China Aviation and other investment and industry funding organizations, which are managed by Tianma Microelectronics.  Entities of AVIC will retain an ownership of 31% of Tianma itself. Tianma will also issue shares to 10 or less non-public investors to raise capital for the transaction and its G6 LTPS fab, which is expected to begin production within the next month or so and will own through other entities, a 100% stake in the Gen 6 OLED fab under construction in Shanghai.

According to the company, the restructuring will reduce the layers of management in their display assets and will increase the company’s competitiveness.  Tianma’s current share of the display market by capacity is ~1%, with a focus on small panel display production.  Tianma has restructured a number of times over the last few years under the same basic mantra of ‘increased efficiency, faster time to market, and flexible customer support’ but has been unable to grow significantly over the last three years.

Korea based Charm Engineering (009310.KS) will sell $8.8m worth of equipment to Samsung Display through mid-April.  Charm sells a number of LCD and OLED related tools, particularly display array testing and repair tools, and similar repair tools for PCBs.  Given the capacity expansion that Samsung Display has underway in the OLED space, with both the A3 and the L7 lines being expanded or rebuilt for OLED, we assume the tools will be used for OLED production. 

We believe Charm has a significant share of Samsung Display’s array test and repair business, as market share leader Orbotech (ORBK), who supplies similar tools to almost all other OLED and LCD producers has only a modest relationship with Samsung Display after a 2012 incident where a number of Korean based ORBK employees were accused of stealing OLED technology secrets from Samsung when on premises to test company tools.  Orbotech bore no responsibility for the incident and was fully cleared, but the relationship remains minimal.   

Last November we noted that Chunghwa Picture Tubes (2475.TT) had indicated that it will sell 152.982m shares of its holdings in Giantplus (8105.TT) to Japan based Ortus Technology (pvt) and will sell the remaining 84m shares that it owns on the Taiwan Stock Exchange through March 15, 2017, with the total holdings equal to ~54% of Giantplus. According to Taiwan Stock Exchange filings, CPT sold 42m shares of Giantplus at NT$16.57 ($0.54 US) today. The price of Giantplus stock when we noted the potential transaction was NT$16.20.

CPT acquired its original stake in Giantplus when it sold one of its Gen 3 fabs to the company in 2007 for $200.1m, which included a 30% stake in GP, followed by CPT selling a color filter plant to Giantplus  for $40.94m in 2010.  CPT acquired an additional stake in Giantplus in April of 2013 in a tender offer that gave shareholders .72 shares of Tatung Co. (2371.TT) plus NT$4.03 in cash, which was equivalent to ~NT$9.65, and just three months later CPT sold one of their module assembly plants to Shenzhen based China Star (pvt) for $71.8m.

Coming as no surprise to anyone in the display industry, the Photonics Industry & Technology Development Association (PIDA), a Taiwan based industry trade group, has indicated that Chinese OLED panel producers are expanding their capacity through 2020.  PIDA indicated that eight Chinese OLED panel producers will have a capacity of 335,000  glass substrates/month or 4.02m/year, and noted that this was far below the 100m glass substrates produced by Samsung Display (pvt) by 2020, and therefore Samsung Display is expected to dominate OLED supply for the next 3 to 5 years. 

While it is nice to see statistics from trade organizations, this information is not new, and the data presented does little to paint a real picture of the industry.  As glass substrates can be vary in size from Gen 1 (.12 m2/1.3 ft2) to Gen 11 (11.52 m2/124 ft2), and each fab chooses what glass substrate size it will use, the sheet metric tells investors little.  That said, when looking at each producer and each fab actual relevant information can be derived.  As we track 15+ OLED producers and over 80 individual OLED fabs and lines, we look at substrate size on a more specific basis.  Figure 1 indicates raw OLED capacity by generation (glass substrate) size, Figure 2 shows raw OLED capacity by region, and Figure 3 shows raw OLED capacity by producer share.

But to the point, the relevant data, rather than the number of sheets produced by China, would be the square meters of capacity, given that this will determine the maximum number of units that can be produced.  Looking at the underlying data for Figure 2, we can see how each region stacks up, and while China will see very significant growth in share, particularly in 2019 and 2020, the share maintained by South Korean producers Samsung Display and LG Display (LPL) remains dominant, and breaking down share further, as in Table 2, Samsung Display will remain the dominant supplier of OLED capacity through 2020 with over 50% in 2020. 

We note also that raw capacity, or the stated maximum capacity that a fab is able to produce, is a relatively poor metric for actual display production, as fabs take considerable time to get to such metrics, with build-outs usually done in two or three phases, equipment utilization, overall factory efficiency, and a variety of other factors reducing raw capacity to what we call ‘utilized’ capacity, or the actual amount of capacity available to the producer at specific moments in time.  This metric and another key variable, product yield, are what drives unit volumes across the industry and are far more important than the number of glass sheets.