LG Display (LPL) is expected to close its P5 LCD fab in Paju, South Korea, which has been producing LCD panels since 2003, most recently those for automotive use.  The fab has been producing panels with a-Si (amorphous silicon) backplanes, which is an older technology that now competes with LTPS (Low Temperature Polysilicon), IGZO (Indium Gallium Zinc Oxide), and most recently with LTPO (Low temperature polycrystalline oxide), which tend to have better characteristics but are more expensive to produce in most cases.  LG Display has been reducing its exposure to a-Si over the last few years and the P5 closing will further that goal.  Production at P5 will be shifted to P6, a Gen 6 fab that is housed in the same complex. 
While no timeframe was given for the closing, other than this year, we note that yesterday we indicated that Japan Display (6740.JP) was closing a small Gen 3.5 LCD fab in order to continue to bring down its breakeven, and with panel prices declining, we expect other panel producers to start reevaluating the efficiency of older LCD fabs, something most were less concerned about when demand was stronger last year.  We expect this trend to continue as panel producers evaluate the efficiency of smaller or older fabs.  LGD has not decided (or announced) what it intends to do with the P5 fab, which could be converted to OLED for small or large panel production at a later date.
LGD’s transfer of its automotive panel production from a Gen 5 to a Gen 6 fab, while seeming a small change, is indicative of how sensitive the panel space is to substrate changes.  Small panel production sees little change from a transition from Gen 5 to Gen 6 fab production, but as panel sizes reach 18” to 19” those transitions become apparent.  In the table below we show the number of units and the substrate efficiency[1] of both fab layouts, noting also that the size of a single sheet of Gen 6 glass is 1.93x the size of a Gen 5 substrate, so the number of units for smaller size panels doubles while for larger panels the number of units increases by between 125% and 150%, while the efficiency also improves.  Given that P5 was designed to produce 150,000 substrate sheets/month, those efficiencies can be a significant improvement to margins while reducing the burden on other company fabs.


[1] The percentage of the substrate that is used for actual panel production as opposed to the waste.

April was not a good month for panel producers in Taiwan, and likely similar results would be found for other panel producers.  AU Optronics (2409.TT) reported April sales of NT$20.22b ($682.37m US), down 27.9% m/m and down 31.7% y/y.  Area shipments were also down 73.9% m/m and down 23.7% y/y.  To put this data in perspective, the drop in sales seen by AUO was the largest m/m decline seen since November 2008, the y/y decline was the largest seen since January 2009, and the panel area shipments were the lowest since AUO began reporting those numbers in February 2020.  Typically these monthly numbers are reported without any commentary however this month AUO added the following to the report, which we feel sums up the month quite succinctly:
“Revenues dropped sharply in April, largely due to weaker demand amid macroeconomic uncertainties caused by war and inflation, together with higher channel inventory resulted from previous port congestions and container shortage. In addition, eastern China has introduced strict Covid-19 related lockdowns since April. Given the challenges of lack of workers, combined with supply chain disruptions under these lockdown measures, the Company has lowered utilization rates at its production sites in Kunshan and Suzhou. Meanwhile, shipments to customers were also impacted by these lockdown restrictions. Currently, lockdown measures were lifted in certain areas while the pandemic gradually eased. However, it may still take some time for market to return to normal.”
​

Innolux (3481.TT) did not fare much better in April reporting sales of NT$20.6b ($695.2m US), down 13.9% m/m and down 32.1% y/y.  Innolux shipped 10.7m large panels in April, down 10.9% m/m and down 10.6% y/y and shipped 26.55m small panels during the month, which was up 14.5% m/m but down 5.0% y/y.  Innolux made no comments about the results for the month

Hannstar Display (6116.TT), who is primarily a small panel display producer, saw  sales of NT$1.39b ($46.91m US), down 20.3% m/m and down 49.6% y/y, while large panel shipments declined to 38,000, down 39.7% m/m and down 78.3% y/y, while small panel shipments declined to 17.32m, down 41.4% m/m and down 52.6% y/y.

​Characterizing April as a bad month for Taiwanese panel producers is a bit of an understatement, although some fared better than others.  As panel producers face lower sales and bring down utilization rates to compensate for weaker demand, we expect to see margins turn negative and if the recovery from China’s COVID lockdowns and the war in Ukraine take much of May, the 2nd quarter will be quite poor, even with a bit of a snap back in June.  While each panel producer will see different results in May and June, we expect Chinese panel producers will still try to maintain shipment levels by continuing to discount panel prices into June, but will soon see those prices fall below cash costs, at which point they have no choice but to lower utilization rates.  While those lower utilization rates will lead to more stable panel prices in 3Q demand does not seem to be strong enough to warrant increasing utilization to previous levels, which means 3Q panel results will be negative on a y/y basis.  We expect the only hope for results better than this scenario would be if China gets COVID under control and there is a bit of a recovery in the Chinese economy, but that is an optimistic scenario.

Samsung Electronics (005930.KS) is among the few companies with commercial Micro-LED product, one intended for what would have to be considered public displays.  To understand the size of these displays, the graphic below shows their size relative to an average height person however Samsung also considers these displays as ‘luxury’ personal viewing platforms for those who want the ‘ultimate lifestyle experience’, including using these displays for artwork when not in use, for a home office, and of course for entertainment.  That said, the cost for the larger configurations run above $500k, although free delivery is included, but the real purpose here is a platform that Samsung can use to refine Micro-LED technology until it become viable for actual retail displays.
To that end Samsung will soon be introducing the next level in its Micro-LED “The Wall” product line, which will shrink pixel spacing further and therefore pack more pixels in each ‘cabinet’ from which these massive displays are made. The cabinets, which are essentially interconnected building blocks, weigh ~27 lbs, are roughly 36” long by 20” long and about 3” deep, and can be configured anyway the customer should desire, but most important to us is the pixel pitch, which is the distance between each pixel.  In very large displays such as these, the viewer is a distance away from the screen (recommended viewing distance for an 85” 4K display is 5.25’) and cannot see individual pixels, but as the resolution of TVs moves higher, the user can get closer without the image being broken up into ‘dots’.  With each new resolution iteration come more pixels and more electronics and more difficulties in producing and placing these progressively smaller Micro-LEDs, so as Samsung and others reduce the pixel pitch, we watch to see if they can hold or lower costs.
While it might not sound like a big change, the increase from earlier models is ~7x in the same cabinet space, and while that is still 1/9 the resolution of a 4K TV, typical user distances from such displays are considerably greater than would be the case for retail TV viewing, so such high resolutions are not as necessary, although the goal is to continually shrink the pixel size and pixel pitch with each new model.  Price for the new model has yet to be announced, so we cannot yet track improvements in production methods and other costs, but while “The Wall” seems to be an expensive commercial product or a rich man’s toy, it is a proving ground for Micro-LED technology and helps to establish a timeline where Micro-LED technology will become competitive with other display modalities.  As higher resolution displays do not come without a ‘cost’, we include the power consumption of each as it increases as the number of LEDs increases.

Japan Display (6740.JP) has decided to close a small (27,000 sheets/month) Gen 3.5 LCD fab in Higashiura, Japan by the end of the company’s 2023 fiscal year, which ends in March.  The fab currently produces displays for cameras, wearables, automotive, and VR, with the production of those products to be transferred to JDI’s remaining fabs in Japan.  The 259 employees at the fab will be offered employment at other sites at JDI and the company is evaluating whether the land and buildings can be sold or used for other purposes.  The company will record expenses related to the closing in the 3/2023 fiscal year, although the amount is currently under review.
The Higashiura fab was commissioned in September 1999 and has therefore been in service for over 20 years, making it less efficient than current modern fabs.  JDI has three fabs remaining in Japan, older Gen 4.5 lines in Mobara and Ishikawa, and newer Gen 6 lines in Hakusan and Mobara.  We expect the impact of the closing will be a financial positive for JDI after closing costs, as production will move to newer fabs that will likely see higher utilization rates.  JDI has been working toward reducing its breakeven for a number of years and while the company is getting closer to its goal, breakeven will not be achieved this fiscal year (3/22) as the decline in JDI’s traditional mobile display business has been severe this year while increases in the non-mobile display business have been relatively modest.  We expect the closing of the Higashiura fab will help to lower the 3/2023 breakeven further, however JDI needs to continue to expand their non-mobile business further before they can return to profitability, which we expect could occur in the 2024 fiscal year.

​The US government has made it extremely difficult for Chinese semiconductor companies to develop advanced process nodes by banning the sale of such equipment if it contains any US software, components, or hardware, but it has allowed equipment used in more mature nodes to bypass those restrictions, with the idea that Chinese semiconductor fabs will only be able to produce the more ‘common’ parts that are needed around the globe, including in the US.  Without the equipment for advanced nodes, China’s semiconductor producers are unable to compete with semiconductor technology leaders such as Samsung and Taiwan Semiconductor (TSM), with ASML (ASML) the only source for the EUV tools needed for such advanced nodes, honoring the US ban.
It seems that the US Department of Commerce is evaluating the impact of even more stringent rules that would ban the sale of equipment used for the mature nodes that have previously been excepted, as a way to push China toward acknowledging sanctions against Russia for its invasion of Ukraine.  While the Chinese government has been making a concerted effort to aid the development of a homegrown semiconductor equipment industry that would not rely on the US for any parts or equipment, no matter how much money is thrown at the development of such tools, much depends on the experience and expertise of the engineers who design those tools, and it has proven to be a slow process for Chinese engineers to develop such expertise.
While we expect there would be the usual hue and cry from the Chinese government should the US expand the existing semiconductor trade restrictions, it will have a devastating effect on the Chinese semiconductor industry and stagnate China’s advancement as a global semiconductor supplier, and even a domestic one.  It would be difficult for the Chinese government to publicly censure Russia or reverse its existing stance, but if the US tightens the semiconductor noose further, China will have to find a way to satisfy the US while not seeming to acquiesce to it demands or face even slower growth from it semiconductor fabs during a period when the profitability of same should be at its peak.
If the US decides to go further with trade sanctions against China, it will also have to contend with US equipment suppliers, particularly Applied Materials (AMAT), LAM Research (LRCX), and KLA (KLAC), all of whom have significant sales in China.  While political pressure continues to push the administration toward tighter rules, US companies will bear the brunt of those restrictions, along with Japan’s Tokyo Electron (8035.JP).  Finding a path that will not penalize the US and Japanese semiconductor equipment business yet will exert additional pressure on China toward the US agenda with Russia is an almost impossible task and even if a solution is found, that does not guarantee that all countries will honor it  despite the political pressure the US might exert, and if it is successful politically will the outcome, both financially and technologically worth the price?

CE brands use trade-ins as a way to encourage consumers to upgrade their devices and trade-in values can become a significant selling point for smartphones, where programs like Verizon’s (VZ) recent ‘…even broken phones…’ trade-in promotion can embolden those whose phones have case or screen cracks or barely work to trade-in those phones rather than junk them or pay for expensive repairs.  That said, as inflation lowers the buying power of those dollars you have earned, Apple (AAPL) has taken the concept to heart and devalued the trade-in values for many of its products, with the iPhone being the exception. 
While we make light of Apple’s move as inflation related, we expect Apple’s cost to refurbish, transport, or recycle those devices have risen along with food and pretty much everything else, so we look at the change as more of ‘what else is new’ rather than blaming Apple for the global economic situation, but while everyone struggles to meet such rising costs, one would have to assume that reducing the trade-in value of its products will have at least some effect on sales during a period where the external factors are enough to slow sales on their own.  Hopefully other brands do not follow (from a consumer’s perspective), but more likely others will join the trend after a quarter or so of eating the higher costs, but we can hope…
Here are the changes:

​We have mentioned the Chinese company Royole (pvt) a number of times over the past few years, in particular relating to folding displays, where the company was the first to release a development kit for a foldable device, beating out CE giants like Samsung Electronics (005930.KS) and LG Electronics (066570.KS).  That said, the company has not done as well as early expectations might have predicted and in our 4/422 note, we described some of the problems Royole was facing, including rumors that the company had been failing to pay worker wages starting last December.  Some of the wage arrears were paid in January, with the promise the rest would be paid after the New Year festival, which did not happen, however we have now heard that after some employees were told to ‘take a long vacation’ (three months), it seems that some workers have been paid missing wages.
The company’s only comment was that ‘funds would be received in batches’, with those employees that resigned without arbitration getting the first batch and those with arbitration having to wait for the 2nd batch.  The source of the funds however was not made clear, other than the remittance itself came from Qingdou Rouyu Technology,  a company established in August of last year (wholly owned by Royole), rather than the typical Hua Xia Bank (600015,CH).  It was noted that in March of this year Royole pledged a number of patents to Qingdao Urban Investment Engineering Construction Development (Group) Co., Ltd. and Qingdao Urban Investment Industrial Investment (Group) Co., both of which are controlled by Qingdao SASAC (State-owned Asset Supervision & Administration Commission of the State Council), whose obligation is to ‘perform the responsibilities mandated by the Central Committee of the Chinese Communist Party’, so one might imply that Royole is being bailed out by the government.
What seems to be a difficult time for Royole is made worse by the late 2020 announcement by the company that it was planning to build a $2.4b flexible OLED fab in Qingdao, which remained on the “List of Key Projects” by the Qingdao Municipal Government earlier this year, which is likely a massive financial burden, if it is still under construction.  Details on the fab’s status are sparse, although the original theory was that Royole’s ULT-NSSP (Ultra-low Temperature Non-Silicon Semiconductor Process) would allow the company (and potentially other Chinese OLED producers) to produce foldable OLED panels without paying royalties to foreign entities.  This was likely the impetus for the original project, but Royole’s lack of traction relative to their earlier products (they produced only 48,600 units in 2020 at their existing Gen 6 OLED fab) has made it difficult for the company to exist independently, as the company’s failed attempts to list in the US and China seem to indicate.
How much capital the local governments are willing to fund going forward is an open question and while Royole has raised ~$1.1b in 7 rounds over the last nine years (the last was in 2018 with a $5b valuation), we expect they will have to give up considerably more IP and ownership if they are to remain in business going forward.  As a private company we expect to see little financial or ownership information forthcoming, with much of what we know about the company coming from previous IPO filings, but we suspect local government organizations will wind up being the majority owners of the company if it survives.

Ennostar (3714.TT), the parent company of Epistar (pvt), the top LED chip producer in Taiwan and sister packaging company Lextar (pvt), reported 1Q results NT$ 8.45b ($283.67m US), down 13.3% q/q but up 24.2% y/y.   The consolidated company generated $13.3m in cash in 1Q, after CAPEX of $31.3m and $33.68m in cash in 4Q after $34.1m in CAPEX.  While we don’t often show results for LED manufacturers, Epistar indicated that May results would be affected by Chinese COVID lockdowns, although they expected to see results return to normal in June.  As Epistar is a major Mini-LED supplier to Apple (AAPL), this represents one more datapoint toward understanding how the Chinese lockdowns are affecting major CE companies.   Figure 1 shows the combined sales of both Epistar and Lextar, which have been on a positive trajectory since last year, but look to be leveling off so far this year. 

Exposure to the IT (Monitors and Notebooks) and smartphone backlight category has grown almost 3x since last year in terms of sales contribution, which makes the company quite sensitive to demand weakness and pricing pressure for smartphones and IT products, hence the warning about said impact in May.  However the company continues to expand its capacity, most recently at Lextar where new lines dedicated to Mini-LED backlight production have been installed, but more telling is Ennostar’s decision to increase the company’s overall 2022 CAPEX budget by 45.5% from $184.5m to $268.4m this year, a hint that despite the short-term issues they currently face in 2Q, they remain positive about their Mini-LED backlight business overall, which has to reflect at least a bit on the company’s relationship with Apple.  

​We have noted the ongoing negotiations between OLED TV panel producer LG Display (LPL) and Samsung Electronics that have been the fodder for headlines citing “Signed, Sealed, & Delivered” to “Still Very Far Apart” over the last few months.  As the days pass, it gets more difficult for Samsung to gain enough momentum to establish an OLED TV presence, with a Chinese firm indicating that it believes Samsung will not offer an OLED TV line this year, other than its own QD/OLED sets produced by affiliate Samsung Display (pvt). 
Citing a number of factors the Chinese firm RUNTO has lowered its forecast for global OLED TV shipments  from 10m units this year to between 7.9m and 8.1m, reducing its growth rate expectations from 53.8% to 23.0% based on 6.5m units shipped last year and also lowered its forecast for TV shipments overall from 219m to less than 210m units.  With the negotiations still not final, and TV panel prices moving further in Samsung’s favor, it gets progressively harder for Samsung to avail itself of the ~4m total OLED sets it needs to represent 10% of its TV sales according to RUNTO.  With expectations that Samsung Display will be able to produce between 750,000 and 1.4m QD/OLED units this year (We expect between 700,000 and 725,000 units – as per our 04-18-22 note) that would leave Samsung with the need for purchasing between 3m and 3.6m WOLED panels from LG Display this year to meet RUNTO’s 10% of total TV sales target. 
With RUNTO’s 11m unit expectations for LG Display’s WOLED panel production and using the same percentages purchased by LG Electronics, Sony (SNE) and other customers last year, leaving ~1.5m units for Samsung to purchase.  When added to the QD/OLED sets purchased from Samsung Display, the total of 2.25m to 3m units falls short of the 10% of TV sales RUNTO feels is necessary for Samsung to initiate an OLED line this year.  Based on that conclusion, the war in Ukraine, the COVID outbreaks and lockdowns in China, and inflation, the company makes the assumption that Samsung will not purchase WOLED panels from LG Display this year.
Again, these are not our assumptions or numbers but we present the concept as a possible outcome (one of many) that are possible given the current circumstances surrounding the display business.  While we do not agree with some of the estimates, not would we assume that unit volume share allocations would be identical to those made last year, the above is certainly possible, although such a decision by Samsung would put the introduction of the expected QD/OLED TV product in a different light.  As part of an overall OLED TV strategy QD/OLED can be offered as the ‘top tier’ of OLED TVs, while on its own it is more of a curiosity than a full product line. 
While we value the scenario noted above, we expect there are many more variables that play into the Samsung LG negotiations and delays have been working in Samsung’s favor.  If Samsung can negotiate a long-term unit volume based contract with LGD at a lower price than late last year, it would serve them well over the next few years, and if that means postponing the full OLED line this year, we expect Samsung’s TV marketing department will find another hook to attract customers during the holiday season.  Regardless, it is interesting to hear what is in our mind a more radical view of the situation, rather than the daily “We know the answer” headlines that tend to appear after an influencer speaks with a low level supplier...

8K TVs are a difficult sell to consumers and rightly so as there is little 8K content that would allow consumers to take advantage of the increase in resolution that 8K TV sets provide.  However each year TV brands offer the public 8K sets, promising that they will be able to watch their favorite 4K (or less) programs using the upscalers built into 8K sets, while waiting for the world to catch up to 8K broadcast and content standards.  The problem is that by the time there is enough 8K content to make a difference to TV purists, the 8K set you just purchased will be either outdated or unable to display the then current content, which would cause considerable consternation from past purchasers.
There are lots of reasons why 8K broadcast content is not available (other than in Japan in limited quantities), but the bottom line is cost for broadcasters and bandwidth for streaming services, so unless you are satisfied with watching a few travelogues and animals in the wild, 8K TV set purchases should be put on the back burner.  We have watched a continual decline in 8K set unit volume estimates over the last year, although brands themselves don’t seem willing to make an estimates of where the 8K TV set market will be over the next few years, so we have put together an aggregation of those estimates we have found, excluding those that we know have been updated during recent reality checks.  We note that assuming a steady-state 220m TV sets sold each year shows that even the most aggressive estimates in 2025 show a penetration rate below 2%, so we would not be expecting TV brands to be adding significant 8K capacity, other than as demo units or for sales to China where 8K seems to be growing the fastest, but with Samsun owning some 65% of the 8K TV market recently, there js little left for others and producing such 8K sets in small quantities is a costly process.  Here are the aggregated estimates: