As we have noted in the past, panel demand and consequently panel pricing was extremely weak in the 1st half of 2016.  This makes 2017 y/y comparisons a bit less relevant, as the gains will be greater than normal until 2H 2017.  It can be seen in Fig. 1 how the comparisons turn positive starting in September of 2016, and have remained so thus far.  The sustainability of these positive y/y comparisons is called into question as we enter the 2nd half of 2017, with September the 1st y/y comparison against more normalized results. 
Our expectations for a weaker April were correct, but panel prices for all but TV panels have begun to decline, which could trigger panel buyers to jump back into the market after working down inventory in April.  That said, this is not a foregone conclusion, as buyers tend to play the trend and could hold off another month to see if prices decline further.  We note that m/m results do follow seasonal patterns, which would lead to the assumption that May’s results should be up from April on a normalized basis using that metric, and follow a similar pattern for the remainder of the year, unless brands begin to lower unit volume expectations for the year, in which case it would take another month of inventory selling to reach new target levels.  All in, we would still expect May to be up on a m/m basis.

Veeco (VECO) has filed a patent infringement lawsuit against SGL Carbon (SGL.GR) that is focused on the graphite wafer carriers used in MOCVD tools.  Veeco alleges that its patents cover the use of certain wafer carrier designs that are being infringed by SGL and has asked that SGL immediately stop selling those products that infringe and compensate Veeco for loss suffered.  Given that there are only three suppliers of such graphite components, SGL, Toyo Sanso (4091.JP), and Xycarb (pvt), limiting a primary supplier or all three suppliers will have a significant effect on the MOCVD space.  According to sources, all three suppliers have halted sales of those products that might be affected by the 4 Veeco patents that are the basis for the suit.
The design of the wafer carrier is the issue here, and while other designs that do not infringe on Veeco’s patents are available, a number of Chinese MOCVD vendors, who continue to challenge the incumbent’s Veeco and Aixtron (AIXA.GR), use a similar design.  Should the negotiations between SGL and Veeco not produce an agreement for the use of the patents, SGL might have to halt shipments to those Chinese MOCVD tool vendors that use such a design.  Chinese MOCVD tool vendors still have a limited market share of 11%[1], but Chinese tools are being tested and validated by local LED suppliers with the state directive of making China self-sufficient in the LED space an overriding factor.  If their supply of wafer carriers is halted during an extended negotiation period or trial, it could limit their ability to deliver previously scheduled MOCVD tools and erode their market share.
Whether Veeco is interested in licensing their IP or just blocking their Chinese rivals is not known, but the impact of the litigation is certainly in Veeco’s favor.  In fact, Aixtron’s and Sanso’s tools use different wafer carrier configurations, which would allow for a Chinese LED supplier to shift away from a potential Chinese vendor if timelines are disrupted, so the benefits of the suit also accrue to the other key MOCVD vendors eventually, but certainly not to Chinese tool suppliers.  Business makes strange bedfellows…


[1] 2016 - Trendforce

Israeli-based Orbotech (ORBK) has received a $24m order for display tools from CEC Panda (pvt) for their Xianyang, China fab that is under construction.  The fab is a Gen 8.6 LCD project with a phase 1 capacity of 60,000 sheets/month, with an expected completion date of October 2017 (a bit later than our original August 2017 estimate), with high volume production in early 2018.  While no date has been set for the phase 2 build-out, we would expect at least a one year window between phase 1 and phase 2, although we include nothing in our industry model at this point.
We believe deliveries to Panda will start in 2Q and continue through 3Q, and given that the phase 2 build-out is also expected to be 60,000 sheets/month, we would expect a follow-on order for phase to be approximately the same size, when received.  What makes this order a bit more significant is that it does not fall into either of the two display growth categories that ORBK has mentioned in the past, OLED and large format LCD fabs.  As a Gen 8.6 fab, it represents a ‘plain vanilla’ order, which is important to Orbotech should timelines for OLED and large format fab be pushed out. 
Orbotech also commented that bookings for the flat panel division will reach $300m by 2Q 2017, for the trailing 12 months, and while this is obviously a sign of increasing demand, it is hard to judge how much of those bookings have been or will be delivered this year.  At the least, the optimistic company view of flat panel bookings, give us confidence in our FPD division estimate of $58.7m for 2Q, with the company having booked FPD revenue of $184.8m (average $61.6m/quarter) for the last three quarters.  We believe Orbotech is in discussions with a number of other panel producers concerning large format LCD display fab projects, and a variety of OLED projects, all of which would fall within the company’s display division segment.

IHS has indicated that they believe LG Display will focus its OLED TV business on a transition from 55” to 65” OLED panels, not dissimilar to what LGD has already indicated when speaking about the popularity of its 65” OLED TVs.  LGD had originally focused on 55” OLED panels as these were moving more quickly down the cost curve, while 65” and 77” OLED TV panels were far more costly to produce.  With yields relatively low on the 65” and 77” panels, the decision to focus on 55” was obvious.  Consumers however, became more interested in 65” OLED TVs than the company had originally expected, and production shifted somewhat to 65” last year, with the company now expecting the combined 65”/77” category to be greater than 55” this year.  IHS expects LGD’s capacity to be equal to 372,000 55” (equivalent) panels/month by the end of 2018, while we expect ~6.5% more, or 396,000 55” equivalents.
The problem, according to IHS, is that the Gen 8 fabs where LG Display produces it OLED TV panels are limited physically to 2 or 3 65”/77” panels from a sheet of Gen 8 substrate with an efficiency of between 42% and 64% for the former and 59% for the latter, while 55” panels can be cut at a 91% efficiency ratio.  This is an important characteristic to fab owners, as that ratio is key to the overall profitability of the fab.  So this leads to the question of what LGD needs to do to increase its production of the 65”/77” OLED TV panel category, without reducing the fab’s efficiency and profitability.
The implication is that a solution would be for LG Display to build a Gen 10.5 fab for its OLED TV production (see note above), where 65” OLED TV panels can be produced at a 94% efficiency ratio, even better than those for 55” OLED TV panels, and while the 77” size would still be at a relatively low efficiency, the demand for that particular size has yet to be determined.  As noted above, the risks involved in a Gen 10.5 fab are considerably higher than those for a Gen 8 fab, and OLED adds another layer of complexity to the equation, given that no OLED Gen 10.5 equipment exists in a production setting.  Initial yields would be very low, as they were when LGD started up their 1st OLED TV fab, and expectations for bringing those yields up to current levels are measured in years, not months, so the decision for LGD is not a light one, and in fact, could be the most critical in LGD’s history, but if OLED TV demand continues to increase, the necessity to bring the cost of such panels down to more competitive levels might be the final push that LGD needs to make the commitment to such a large format OLED fab despite the significant risk.

Last September we noted that LG Display (LPL) had yet to make a decision concerning the format of the new P10 fab complex it was building in Paju, South Korea.  The original $1.67b construction project was to house a new fab, although there was a great deal of speculation as to what substrate size the new fab would use.  We heard Gen 9, Gen 10.5 and even Gen 11 at various times, and further speculation on whether it would be exclusively OLED or both OLED and LCD, and whether it would be for TV panels only.  By February of this year, the Korean press had indicated that all the final decisions had been made and the new fab would be a Gen 10 format for OLED production, with oxide backplanes, although we reserved endorsing final fab configuration until we received final confirmation from LGD.  Later in February, Taiwanese press indicated that the fab was to be a Gen 10.5 LCD line originally and will add an OLED line at a later date.  We indicated that we believed that information was incorrect and while we had built a number of potential fab configurations into our OLED industry model to see the effect on both LG and the industry, we indicated that we did not believe LG had made final decisions concerning the fab.
Fast forward to today, where Korean press now indicates that the new fab, which now has a $3.56b price tag, is going to be a Gen 6 flexible OLED line, with the original thinking having moved from the OLED TV perspective to small panel OLED production, perhaps to focus on retaining some of Apple’s (AAPL) potential OLED smartphone business.  This speculation is based on LGD sharing plans with equipment suppliers, who will be receiving orders for the new fab’s tools in the near-term.  Now the fab is expected to be able to produce an initial 30,000 sheets/month at Gen 6 in 2018, rather than the 7,500 sheets at Gen 9 that was last estimated.  Should this be the case, and again we have not confirmed either the ‘old’ or ‘new’ plans, it would represent an initial increase in m2 of 29,202 m2/month or 54% for the initial build-out, although that capacity is likely going to be for both rigid and flexible OLED lines.
Timing of initial production still remains at September 2018, which we believe should be the case unless LGD has difficulties with deposition tool scheduling, which has been an issue for producers in the past.  Primary OLED deposition tool supplier Canon-Tokki (7751.JP) has been expanding capacity to meet significant demand, but has not been able to commit to all customer requests, with Sunic (pvt), SNU (080000.KS), SFA (056190.KS), and Applied Materials (AMAT), looking to pick up the slack.  Equipment deliveries should begin in March 2018 for the phase 1 build-out and phase 2 equipment deliveries are expected to begin in late 2018 or early 2019 for phase 2, which we expect will be completed in the summer of 2019.  Along with investments in two additional OLED fabs that LGD is building, this change, should it prove true, would give LGD a 30% share of total OLED capacity by the end of 2019, vs. Samsung Display’s (pvt) 57.1% in the same time period.
We do believe that the risk involved in a Gen 6 OLED fab is far less than that of a Gen 9 or 10.5 OLED fab, given the more developed state of the small panel OLED market relative to the OLED TV market, and while even without the specter of Apple’s entry into the small panel OLED space, LG Display would likely be better served as a secondary small panel rigid OLED supplier and potentially a primary small panel flexible OLED supplier, with the ability to expand their Gen 8 OLED TV capabilities, either at P10 or their existing E3 or E4 OLED TV fabs, based on the growth of that market.  Again however, LG Display’s company representative’s comment was, “Although we are looking at many ways regarding investments for P10, we have not finalized on anything yet.”

OLED materials are a growing business as the adoption of OLED displays continues to increase along with unit volumes, and the key ‘ingredient’ in OLED displays, the emitter, is the most expensive in the OLED material stack.  While the goal of OLED display producers is to produce the best display possible, they also face the issue of cost, and must focus on reducing the overall cost of materials in any way possible.  That said, the emitter materials used by OLED display producers, are covered by IP owned by Universal Display (OLED), and must be licensed, either for R&D or commercial usage.  These emitter materials are based on heavy-metal phosphorescent technology, of which UDC is the sole provider, which gives UDC an upper hand in controlling emitter material pricing.  Red, green and yellow/green emitter materials are therefore subject to no outside competitive pricing pressure, while blue emitter material, which remains a fluorescent compound, is produced by a number of potential suppliers, as no phosphorescent equivalent has been developed.
So there are two gaps in OLED emitter materials, blue phosphorescent emitters, and a way to reduce the cost or increase the competitive nature of phosphorescent emitters, and with panel producers working toward reducing OLED costs while improving products, these topics are of great importance to OLED panel producers.  Blue phosphorescent materials are difficult to produce as the energy associated with such molecules is high enough to make the molecules unstable.  There are some light blue phosphorescent emitters being tested, but a dark blue phosphorescent emitter material with a lifetime similar to other OLED materials has yet to be commercialized.  While UDC has purchased significant ‘blue related’ IP and other material suppliers have focused research on the same concept, that gap has yet to be filled.
Currently, as noted above, blue fluorescent emitter material is currently being used in RGB OLED devices as a substitute for blue phosphorescent material, but fluorescent materials do not give off the same amount of light as phosphorescents, and therefore force panel producers to make concessions to the lower output fluorescent blue that reduce the overall display potential of the device.  Two private companies, and a wide variety of OLED material suppliers however, are working toward a solution that does not entail the development of a phosphorescent blue emitter.  This developmental technology, known as Thermally Activated Delayed Fluorescence (TADF) is a process by which a blue fluorescent material can be made to produce the same amount of light as a phosphorescent equivalent would produce.  Again, the main issue, even for TADF blue emitters, is lifetime, and as of yet, no commercial blue TADF material is able to meet current lifetime demands from OLED panel producers, but material improvements are being made and lifetimes are extending.  That said, we would not expect to see a commercial blue TADF product that could meet specs and be in volume production before 2019, but it does represent at least a challenge to those developing a phosphorescent deep blue emitter.
Germany-based Cynora (pvt) is the private company most directly involved in the development of blue TADF emitter materials, and is currently venture funded, although we would expect the current round of financing to include some ‘strategic’ investors.  Cynora is focused on developing a blue TADF emitter and while attention is given to other potential TADF emitters, Cynora is working toward filling the ‘blue’ gap mentioned above.  Kuylux (pvt) is a Japan-based company[1], with substantial industry investors[2], that is also working toward developing TADF emitter materials, but with more of a focus on being a potential alternative supplier of red and green emitter materials.  As these red and green TADF emitters are not based on heavy metal phosphorescent technology, they would not fall under the UDC license umbrella, and would offer an alternative material to OLED producers if they can meet current specs.  Both companies have specific material technology platforms that define them, but both are working toward the same goal of producing commercial TADFs for OLED panel producers, as are other OLED material suppliers, but we note that most large specialty chemical companies are interested in high volume materials and current profitability, which gives them less flexibility to allocate resources to such a specific goal.  Of course, should either company develop a successful TADF material, there will be no lack of large material suppliers who will look to license the IP.
We will go into more detail on the development of TADF materials in notes over the next few weeks.


[1] Spinoff from Kyushu University – US operations in Boston

[2] Japan Display, JOLED, LG Display, Samsung Display,

May display panel pricing continued to indicate a divergence between display panel types, with weakness in IT panel pricing (NB and Monitors) and mobile devices (tablets and phones), while TV panel pricing remained stable.  As a reference point, TV revenue has been between 43% and 53% of panel producer totals in recent quarters, so it carries significant weight in the overall sales picture for most large panel producers.  As has been the case for much of this year, TV panel prices have remained stable, after seeing large increases last year.  Panel producers have held TV panel pricing steady despite tepid demand, as capacity issues for certain TV panel sizes have given them the leverage they need to hold back pressure from set makers.  That said, indications are that Chinese TV inventory has been rising, as holiday sales have been a bit less than expected, and we expect lower TV set targets to be announced from Chinese brands within the next 60 – 90 days.  While this still will not force panel producers to lower prices, we would expect selective deals to be made for those set makers looking to offer discounts to consumers to move inventory in upcoming months.
On the IT side, panel producers have been a bit more flexible on price as high brand inventories reduced demand and panel producers have responded with lower prices for the most part, although notebook sales have improved a bit and helped to hold panel prices somewhat steady for the month.  Tablets and phones however are facing a transition, as the new ‘bezel-less’ smartphones, particularly the Samsung Galaxy S8 line, have an 18:9 aspect ratio, different from the typical smartphone 16:9.  The popularity of this new size has caught suppliers off guard and has left demand for the older panels less than expected, leading to the lower prices for both smartphone and tablet panels.  While excitement for the newer panels will continue, there remains significant demand for the older aspect ratio panels from white box producers, and eventually the older panels will find a home, while demand for the newer panels (which are not yet included in pricing data) will increase along with pricing.  We would expect published smartphone and tablet panel pricing to continue to decline this year, but if 18:9 panel pricing data becomes available, the aggregate will look far more normal.

If one were able to look at the show from high above the floor, a more general perception of the focus of the products would be clarified.  Of course the general focus was on LCD displays, which remain the bread and butter of the industry, but OLED was clearly a big topic, as it has been for the last few years.  But flexible OLED, which in previous years was usually in a special section of the booth, surrounded with ‘Do Not Touch’ signs, was now in a more prominent location.  That said, much of what was shown as ‘flexible’ was, as has been the case with consumer products that are called ‘flexible’, conformed, meaning it is shaped at the factory and the shape cannot be changed by the user.  Samsung Display (pvt) is obviously the leader in the ‘conformed’ space with a number of models that use rounded edges for better feel and extra features, but we saw a number of similar conformed demos from other display manufacturers.  The real question however was when will other producers be able to move such demos into production, and here the story was quite different.
There were no display producers who indicated that they had moved from typically rectangular OLED display production to being able to produce conformed product in quantity except LG Display (LPL), who has experience in flexible smartphones, but not in conformed OLED displays.  We have mentioned previously that Samsung Display is said to be working on a 4 sided conformed display, which would set them even further apart from other OLED display manufacturers, but we would not expect to see such a device until 2018 at the earliest.  But the real question was not whether you can make even a conformed OLED display in an R&D setting, but can you produce them on a high volume production line.  We asked the question to as many panel producers as possible and the answer was a resounding ‘not yet’, and this was not about a flexible OLED device, but a conformed one.  Many cited the capacity issue, with fabs in various stages of planning or construction, and a few even admitted that they do not have a timeline yet, but none were willing to give anything more than ‘next year’, and those were followed by various forms of ‘we hope’ gestures and smiles.
But what about Samsung Display, the leader in ‘flexible’ displays?  We have mentioned the rumors that they are expected to have a flexible smartphone display out before the end of the year, and various renderings have painted pictures of phones that bend, fold, and even roll, with little said form Samsung.  There were ‘stretchable’ display demos, where the entire display was produced on a material that could be stretched a bit, but again these were demos, not production models.  So what about the expected foldable/flexible/bendable phone that is expected?  There were many demos of foldable OLED displays, with mechanical bending apparatus folding and opening mobile sized displays like smartphone fish, but even less was said about when those demos, which were better this year (higher resolution) but pretty much the same as last year’s demos, would be available for production.  Since we got little in response to the flexible production question, we went to a number of equipment manufacturers to see if they had some idea as to when such devices would be available, since without their equipment, nothing can be made.
The equipment manufacturers, who have the onerous task of translating customer requests and specifications into a working device, looked at the mass production of flexible displays as a bit more of a challenge than most of the panel producers, and when speaking with the engineers that have to design and build such tools the prospects were even less than from the tool vendor marketing side.  According to the tool vendors, there remain a significant number of problems that must be solved before a truly flexible OLED smartphone could be produced, and they don’t expect those to be solved this year.  While we had already taken that as a given, it calls into question whether that means 2018, 2019, or 2020, but at least there was optimism that eventually the problems could be solved. 
Of course the next step was to try to understand the specific issues and potential solutions, but more important was the overriding conflict between the engineering and marketing departments, not at the equipment companies, but at the panel producers.  One side, and you can guess which it is, says “What if we spend all this time and money on developing a flexible display and consumers don’t care?”, with the other side saying, “We can do this eventually, and we can do it better than anyone else”.  While this sounds pretty basic, it was mentioned a number of times as part of the discussions equipment vendors have with their display counterparts, and seems to be part and parcel of what is slowing the development of flexible displays, essentially, “What if we throw the party and no one comes?”  We make light of the conflict, but it is one that is key in determining timelines, spending, and producer focus, all key topics for display producers. So the bottom line, after speaking with a variety of participants in the flexible display segment is,  “Probably soon, but not exactly sure when”, which loosely translated means, “As soon as someone else produces a high-volume flexible product, and we can see if anyone buys it, we will set real goals as to our own production.” 
Most are secretly hoping that Samsung Display will do just that, but Samsung has expressed the same consumer-based questions and while they would desire the absolute cache that a successful flexible OLED device would garner, they do not have unlimited resources and have to balance those desires against the reality of quarterly performance.  We would expect to see some sort of flexible display in 2018 on a limited basis, most likely from Samsung Display or possibly LG Display, but most likely not at the flagship level, and more as a trial balloon to gather information about consumer acceptance and how much of a premium the display could generate, with real competition, assuming consumers like the idea and are willing to pay for it, starting in 2020.  All in, our conversations with supply chain participants lead us to push out our ‘flexible’ timeline a bit, although with the number of participants involved in the process of developing the technology in a practical way, we still feel confident that the potential manufacturing and other issues will be resolved. Tomorrow…TADFs.

As we mentioned previously, we have been at the Society for Information Day “Display Week” conference for the last few days.  This yearly seminal event is the place where members of the display space come together, tell each other what they have discovered, and show off products.  While not the size of CES, this show is far more specific to advances in the display space as opposed to the CES philosophy of considering anything that a consumer might buy as potential show fodder.  This gives a much narrower focus to the companies involved and the orientation of the conversations.  No washing machine talk but lots of OLED and bezel-less conversation.  Of course, much of what is said is rhetoric, particularly at the more public venues, but on the exhibition floor, far more specific conversations can be had with key operational personnel, marketing types, and a variety of others.  It is far more educational, both from an industry standpoint and a specific company standpoint than any other display related gathering and we make the trek each year to drain the heads of as many smart folk in the display space as we can.