​Taiwan’s State-run power company reported that an ‘incident’ at a major power plant caused a blackout that affected ~5m of the country’s ~24m residents.  Problems at a power plant in Kaohsiung triggered the blackout but an investigation is underway with company executives ‘not ruling anything out’ until completed.  The timing of the blackout, a day after the US Air Force made a public demonstration of its military air power at an airbase in Japan as the Russian invasion of Ukraine caused increased fears of a similar move by China toward Taiwan, who China considers to be a part of the People’s Republic of China and not a sovereign country.
Much of the power grid was restored within an hour as alternative sources were brought on line and the downtime at display and semiconductor manufacturing fabs was covered by back-up power common at such facilities, however even millisecond transition times to backup power can cause steps to be missed during a number of processes, which takes a bit of time to evaluate.  With Taiwan Semi (TSM) the globe’s leading producer, having production capacity in affected regions, the company will need to evaluate production before it can accurately estimate any potential losses.  Taiwan is known to have occasional blackouts, but those are typically during the summer months when the heat causes demand spikes across a grid that was designed to meet ‘average’ power consumption.

February is an unusual month for the display panel business in that the Chinese New Year holiday disrupts panel production as workers travel home to visit family, however COVID-19 restrictions has lessened some of that travel given China’s strict local and provincial lockdowns on even the smallest outbreaks.  Typically (see Figure 7) January and February are the weakest for display shipments on a m/m basis, so we are careful not to ‘over-read’ pricing data this month, however we can give some perspective on how panel pricing stacked up against our expectations.
On a broad basis, TV panel pricing was better than expected, albeit still down 2.2% for the month.  While this is certainly a positive for those panel producers that have an orientation toward large panel production, and it sets the tone for more modest TV panel price declines or even an increase in March (typically the strongest m/m shipment month), we do not expect that it sets the tone for TV panel pricing for the entire year and would look to March TV panel pricing data to see if typical m/m seasonality holds true before estimating TV panel pricing for the year.  As an offset to the better than expected TV panel pricing was worse than expected IT panel pricing, and as we have stated a number of times previously, the industry has been moving production capacity from TV panel production to IT panel production for the last year as an offset to the rapid decline in TV panel pricing.  This creates additional industry sensitivity to IT panel price changes and as can be seen in Figure 6, the trend is down.
There is a bit of nuance toward TV panel pricing in February, with the smallest TV panel sizes (32”) seeing essentially flat pricing and large panel sizes (43” and above) seeing declines.  Figure 8 shows 32” panel; pricing since 2018, and while the recent price declines did not trace 32” panel pricing all the way back to the low set back in late 2019, it did retrace much of the 32” panel price increases seen in 2H ’20 and 1H ’21.  In the past 32” TV panels represented ‘small’ TVs and were considered commodity panels, manufactured by a large number of panel producers, however more recently some of those panels are being used for gaming monitors, which has given the size segment a bit more cache and will likely help to maintain a more stable price, but does not rule out further price reductions..
Figure 9 shows the same price data for 65” TV panels, which are more the mainstay of the TV set market.  Again much of the 2020 and 2021 panel price gains have been reversed, however the price curve has yet to bottom and is currently 9.3% above (based on March expectations) the chart period low, so we expect less aggressive declines but a near return to previous lows, as the competition among panel producers is most intense in the 55” and 65” TV panel categories.
All in, March should, if seasonality holds, be a better month for shipments, but will still see panel price declines, particularly in the IT segment (monitors, notebooks, tablets), which will affect sales performance.  That said, we expect panel producers to see volume increases high enough to offset panel price declines in March, which will help margins but do not expect that to be sustainable in April, particularly as the industry concentration toward IT products faces further panel price declines.  We note that on a long-term basis panel prices decline and while the COVID-19 pandemic has changed that over the last year or so, we expect the industry to return to that mode during 2022 as the pandemic is brought further under control.  While we still expect price cycles, as has been the case in the past, we expect the display industry, at least the LCD portion, to return to a more passive mode as long as we don’t see a major resurgence in the infection rate.

​Chinese OLED panel producer Tianma (000050.CH) has indicated that it has started production runs at irs T18 fab in Xiamen.  While this is likely test run production, it is ahead of our ramp schedule and while it will be a number of months before the fab will be at full (phase 1) capabilities, it will likely take considerably longer for Tianma to fill this fab’s capacity.  The fab itself consists of two 15,000 sheet/month flexible OLED production lines and additional room for a third line, for a total of (phase 1 and 2) of 48,000 sheets/month and has an estimated cost between $6.8b and $7.6b when completed.  Tianma runs a Gen 5.5 OLED fab in Shanghai that is primarily producing rigid OLED displays, another Gen 6 OLED line in Wuhan that was built for a full capacity of 37,000 sheets/month, but is not fully built out, and a Gen 4.5 pilot line used for R&D.
Last year Tianma began the year shipping ~1m OLED display on a quarterly basis, increasing that to between 2.2m and 2.8m as the year progressed, which gave them a share under 2% of the small panel OLED market, so the Xiamen fab will add to that share as it ramps production , however based on Tianma’s capacity last year, which based on a 6” display base, would have been ~154m units, so based on our estimates of ~9m units shipped last year, the  company’s OLED fabs are considerably underutilized or have extremely low yield rates.  This makes it more necessary for Tianma to build out customer relationships or improve yield rates in order for the new fab capacity noted above not to depress profitability further, which leads us to expect a later start for the second phase of the Xiamen fab. 
Tianma also runs a number of LCD fabs which provide much of its sales volume, although the newest Tianma LCD fab went on line in late 2017, and while all of its LCD assets are likely fully depreciated, they are heading into their 5th year of production and are likely in need of upgrades in order to compete with some of the newer LCD fabs in China.  This makes building expensive small panel OLED fabs that are producing at low utilization rates even more risky as capital has to be allocated to upgrade the LCD side in order to provide steady income, while the spending for OLED produces losses.  That said, in China the display industry is funded by the Chinese government at the provincial or city level primarily, and while those capital flows have shifted a bit toward semiconductor capacity, we expect Tianma will continue to be support with subsidies and financing from government entities until the Xiamen fab is completed, and while early individual stockholders might be happy with the results (initial price – 4/1/1995 - $1.82), current stockholders would be less so as over the last 2 years the stock has fallen from a peak of $17.90 in February 2020 to its current price of $12.07, not quite at its low (January 27,2022 - $11.83) but close to it.

Alienware (DELL) is the first monitor producer to use the soon-to-be-released QD-OLED panels produced by Samsung Display (pvt), in this case the 34” panels that will be produced along with 55” and 65” panels that will appear in the Sony (SNE) Bravia A95K series announced at CES, and (hopefully) parent Samsung’s (005930.KS) line of QD-OLED TVs later this year.  As Alienware has revealed pricing for the 34” QD-OLED monitor, we compared a few similar sized monitors to give some idea as the price of this new technology and how it relates to what is already available.  We note that comparing monitors is a difficult task as the number of variables can be large but we tried to show both older and newer models and how they compare along with higher and lower priced models in the table below.
We do note that the first OLED monitor, the Dell (DELL) IP3017Q was quite expensive when released, which is not unusual for an initial foray into a new technology, however the soon-to-be-released Alienware AW3423DW, while expensive as far as monitors go, is certainly not the most expensive, which is a bit surprising.  While the basic technologies, OLED and quantum dots are both used by a number of panel manufacturers, combining the two in a way that takes advantage of the positive aspects of both materials leads to new manufacturing processes that should keep costs high, especially in the early stages of production., so the fact that the Alienware monitor is not priced higher goes toward Samsung Display’s ability to maintain a reasonable cost structure from the onset of production, or at least that is what we would hope, as they also could be pricing the displays at or even below cost in order to garner interest from new customers for the technology.
Both Sony and Dell have committed to QD-OLED so there is certainly demand for the technology as a way to generate premium products, but a lack of enthusiasm from Samsung Electronics seems to have kept others in a more look-and-see mode before jumping on board.  Samsung, as the largest TV set brand, does have a vested interest in maintaining a broad line of TV technologies, with Micro-LED, Mini-LED/Quantum Dot, Quantum Dot Enhanced LCD TV, and LCD TV all part of the mix, so there are certainly pricing and tier considerations that Samsung must consider, but it will be hard for Samsung Electronics not to release at least a line of QD-OLED TVs in order to compete with rival Sony if reviews are positive, as we expect they will be, especially if prices for such sets are reasonable, as the Alienware monitor seems to indicate. 

​[1] Screen Refresh (Hz) – Time it takes for the entire screen to be repainted
[1] Display Response Time (msec) – The time it takes for pixels to shift from black to white to black, essentially how fast the display can switch colors.
[1] Typical Brightness (nits) – Average display brightness
[1] sRGB is a color space measurement that shows the percentage of that color space that the display covers.  The higher the better.

​AU Optronics (2409.TT), is the 4th largest LCD panel producer globally and has taken a different path than many of its competitors over the company’s more than 25 year history.  While in the early years AUO built capacity in Taiwan, purchased capacity in Singapore, and eventually expanded into China, the company has done relatively little in terms of expanding production capacity over the last 5 or 6 years, a period when Chinese panel producers have expanded LCD capacity rapidly while South Korean panel producers have gone in another direction, reducing LCD capacity and focusing on OLED production. 
We have noted recently that AUO has decided to expand capacity at their Kunshan, China Gen 6 fab by 28.6% this year, but has remained more focused on upgrading its older fabs than constructing new ones.  That is going to change over the next few years as the company has announced it will be acquiring land and developing power and engineering capabilities for a new fab in Taiwan.  The capex budget for this project will be NT$5b ($180.3m US) this year, and will increase to NT$28b ($1.01b US) as the project develops, with the equipment purchases likely to begin in 2023, pushing the overall project cost to between $3.6b and $5.4b.  Based on those expectations, we expect the new fab will be built in at least three phases, with the first completed in 2025, and the additional phases being built out over an additional 2 – 3 years.
AUO did not give additional details about the fab other than it is expected to be a Gen 8.5 or Gen 8.6 fab and that it will be focused on IT panel production, but owing to AUO’s conservative nature as to capacity expansion, we would expect the full build out might take many years and potentially change a bit over time.  The land and infrastructure construction is just a start for what will become another major fab for AUO, and will eventually add to their automotive and premium IT panel production capabilities.  AUO however has shied away from OLED panel production and has become more focused on LEDs as an advanced display source, so there is the possibility that phase 2 or 3 of the new fab might have a different production orientation when it is built out, although for now, the orientation would still be LCD..  

All three Taiwan based panel producers saw revenue decline on a m/m basis, with both Innolux (3481.TT) and Hannstar (6116.TT) seeing y/y declines, while AU Optronics (2409.TT) saw a 4.5% increase.  For reference, AUO’s typical[1] January sales decline is 10.8% m/m.  That said, AUO saw both a m/m and y/y decline in area shipped, likely a reflection of the lower utilization rates seen by many LCD panel producers as demand slows and panel prices decline.  Innolux saw sales declines on both a m/m and y/y basis and while large panel shipments were up 2.4% m/m, they were down 1.1% y/y along with a steeper m/m and y/y decline in small panel shipments.  While increased large panel shipments might have offset some of the impact of declines in large panel prices, Innolux also saw a substantial drop in small panel shipments, both on a m/m and y/y basis.  Small panel production is less meaningful in terms of fab usage (m2), however small panels are usually more profitable on an m2 basis, so we expect the lower small panel shipment volume will be a contributor to lower margins for 1Q if such declines continue.  Hannstar is primarily a small panel producer, so we put less weight on the large drop seen in their large panel business, which is primarily IT products, but it does give some indication as to the demand level for such products, which has been weakening as COVID-19 restrictions have been reduced.
All in January results for Taiwan based panel producers seem to be in line with our expectations, which are based on our outlook that panel prices for both TV and IT panels will continue to decline in 1Q.  While 1Q guidance from panel producers is a bit more optimistic than we might expect, with the concept that panel pricing will stabilize during 1Q, that has not been the case thus far, and we would expect that if February results continue to be weak, which we expect will be the case, the hope that panel prices will stabilize will likely be pushed out to 2Q and panel producers will likely have to lower utilization rates again in order to slow the decline in panel prices. 


[1] 5 year average (2017 – 2021)
 

The US Department of Commerce’s Bureau of Industry & Security (BIS) has added 33 Chinese companies and university labs to the ‘unverified’ list it maintains.  This list is different than the ‘entities’ list also maintained by BIS, however it does not preclude US exporters from ‘engaging’ with listed parties or that there are specific foreign policy or national security concerns with those on this particular list, only that the BIS is unable to establish the legitimacy and reliability relating to the end use or end user of the items that fall under the US Export Administration Regulations (EAR) by such parties.  The reasons for this inability could be an inability to contact or locate the parties involved, failure by those parties to demonstrate the disposition of items subject to EAR, or lack of cooperation by a host government with the BIS when making end-use checks.
While this listing leaves trade open, something the ‘entities’ list does not, transactions with those on the ‘unverified’ list require additional documentation, including a statement from authorizing officials of ‘listed’ parties, and cannot receive Ear regulation exceptions that might normally be available to those not on the list.  What all of this legalize means is that the US government was not able to verify if the products that are being exported from the US to produce products by these companies in China are going to programs supporting the Chinese military or any program that “…are contrary to the national security or foreign policy interests of the US.”  Since the US government cannot verify the end-use case, these companies and labs go on this list.  If it is later found that they violate the EAR act, they will likely wind up on the ‘entities’ list, but still have the option of verifying that their products are for non-military use, which would remove them from the ‘unverified’ list.
The listing seems to be the result of the company that owns HKC’s Chongqing Gen 8.6 LCD fab (Chuzhou HKC Optoelectronics), which is owned by HKC (0248.HK) and two Chongqing government entities, is submitting proposals to US suppliers for the purchase of analytical instruments used in the display process.  Given the ongoing trade friction between the US and China, such proposals likely triggered closer examination by the BIS, who remains unsatisfied as to verification that the end use of HKC’s products will not violate BIS rules.  HKC has three producing fabs, although not all are completely built out.  We believe the company has a 3.6% share (2022) of the global LCD panel market on an m2 basis, which would increase to ~5.3% in 2023 if they continue with current expansion plans, with ~80% of panel production oriented toward TV panels according to Trendforce. 
While HKC certainly has the opportunity to verify the end-users of products produced with such equipment, they would have to open a path to any potential display projects that would benefit the Chinese military and would likely need the approval of the Chinese government to do so, if such were the case.  That said, it seems a bit of a stretch to assume that HKC’s reason for purchasing display analytical equipment is rooted in some nefarious government activity (although it cannot be ruled out), and the risk that HKC takes in the possibility of having the company moved to the ‘enties’ list, would restrict its ability to be supplied by any US company without a special license, which would then become hard to procure.  Corning (GLW) is a glass substrate supplier to HKC, as are a multitude of other companies either in the US or allied with the US, so we expect the addition of HKC to the UVL is more of a ‘polite’ warning that it either lets the US look behind the curtain or it could face the kind of restrictions that have cut deeply into larger companies, such as Huawei (pvt) or ZTE (000063.CH).
​

​AU Optronics (2409.TT) held a ceremony to officially begin its capacity expansion project at its Kunshan, China Gen 6 LCD fab.  While the expansion itself is relatively small, adding an estimated 9,000 to 10,000 sheets/month to the fab’s current 27,000 sheet/month capacity, we expect AUO has a further plan to add an additional line of the same capacity at a later date, for a total of 45,000 sheets/month.  We expect there will be other upgrades to the fab, which began production in 3Q 2016.  Based on equipment timelines, we expect the initial expansion to be completed in September of this year. 
AU Optronics has been relatively conservative toward capacity expansion over the last few years, looking more toward developing leading edge products than garnering market share, the opposite tract of Chinese LCD panel producers who have been expanding capacity despite the risk of cyclical overcapacity that has plagued the display industry over much of the industry’s recent history.  While AUO has spent ~$367m for equipment purchases to upgrade LCD production facilities in Taiwan since the end of August last year, much is going toward upgrades rather than capacity (other than Kunshan), which has allowed AUO to maintain a key position in a number of premium panel markets.  While this philosophy might not have looked good during the rapid panel price increases seen in late 2020 and early 2021, the true test comes when the effects of weaker panel pricing drives down profitability for those in the commodity panel space, while premium product pricing is affected to a lesser degree, which we believe is the case currently.
.
 

Display panel pricing is a touchstone for the CE space as displays tend to be the largest cost single component in almost all mobile devices and many static CE products.  Predicting such prices on a monthly basis takes a combination of data from a number of sources, and while conversations with suppliers, assemblers, and buyers add to the calculations, much of such prognostication is based on trends and a gut sense of the balance of tension between buyers and sellers.  In recent months, after a long spate of price increases, panel prices have been declining, and in some cases at a rapid case, so the art of predicting panel prices has moved from being one where actual prices and expectations are only off by 0.25% or 0.5% to one where the difference can be 3% or 4%.
That said, as we have noted previously, TV panel prices began the downward pricing cycle in July of 2021 and have been declining precipitously since, while IT panel (Notebook, Monitor, Tablet) panel prices continued to rise.  This pushed panel producers to accelerate a move from TV panel production to IT panel production to maintain margins, but that ‘lemming-like’ move also has consequences, that of oversupply, and IT panel prices have also been on the decline since last July, albeit at a much slower pace than TV panels.  There are however resistance points when it comes to panel pricing, at that point is cash cost, at which point the producer has to decide whether to take an order that will generate little or no gross margin.  Although cash cost varies with each producer, on a general basis the last time TV panels reach cash cost across the industry was in September of 2019, and since then there has been the usual cost reductions that panel producers work toward each year, along with both general inflation and the effects of component shortages, which we expect nets out as a panel cash cost a bit above the 2019 level.
The industry usually looks at the approach of cash costs for a particular panel type as the point at which panel prices stop falling, and rightly so in many cases as taking on money losing production tends to be done only when it is necessary to fulfill a commitment to a large customer, and with TV panel prices nearing cash costs, expectations are that prices will stop falling.  This hope is fueled by the fact that there is less TV panel production than a year ago (the shift to IT), which tightens the market a bit, but there is also that balance between buyers (TV set producers) and sellers (panel producers), which was heavily in favor of panel producers until last July, forcing set manufacturers to eat the higher cost or raise prices and lose consumers.  Historically, such a shift takes time to reset and has a tendency to encourage panel buyers to keep pushing for lower prices, even when they know they ‘sense’ that panel producers are hurting, and panel producers are also both competing with others and are want to keep as much of the fab utilized for as long as possible, so while cash cost levels are an indicator that TV panel prices ‘should’ stop falling, that can be a simplistic look at a complex equation.
But, more important currently, given the transition mentioned above, are IT panel prices, as the IT production concentration of many panel producers is weighted toward such panel products.  This makes the effect of IT panel pricing more sensitive to the display business than might normally be the case, and the recent slowdown in demand, as COVID-19 vaccines allow for some semblance of normality, has pressured IT panel prices.  We have seen very recent trade press headlines indicating that IT panel prices in February are falling faster than TV panel prices, and while that might be attention grabbing, we went back to our data to see if such was really the case.
Aside from the headlines, we note that panel price estimates vary across a number of sources, and we use averages whenever possible.  We also deal with variations in the type of panel, which can affect price estimates, and we caution that in most cases we do not have sample sizes from sources, but here’s what we found.  Expectations for February TV panel pricing, sort of an ‘average of averages’ across sizes, indicated a decline in February of 3.0% while that same ‘average of averages’ for IT panels estimates a decline of 3.2% for February, so technically the headlines are correct in that IT panel prices in February are declining faster than TV panel prices.  That said, the question is which is worse, TV panel price declines that could put some panel producers’ TV panel business in loss mode, or the increased effects of IT panel price declines in an industry concentrated toward IT panel production? 
Neither is good, but if we had to pick one, we would say the continuing decline in TV panel prices (see Figure 7), even if they dip below cash costs, is the better of the two, as it gives TV set producers the ability to bring down set prices over the next few months, while IT panel prices are still a distance away from cash costs, giving them room to fall further, and considering the shift toward IT panel production over the last 7 or 8 months, the competition in that space could become brutal.  February is an unusual month for the display industry as the Chinese New Year tends to slow production across the industry, so the trend in March will be a bit more telling, however we expect at least the 1st half of 2022 to be a difficult one for the display space, at least from the perspective of panel producers, so we look toward any data that might help us to understand how the CE space will look in 2022.

While we already have a production schedule for Chinastar’s (pvt) T9 LCD fab in our model, sources in South Korea confirm that the company has placed orders for the equipment necessary for the new Gen 8.6 fab, which was announced las June.  The construction part of the project, which will have a capacity of 180,000 sheets/month when fully built-out, is expected to cost $565m US, with a total fab project cost of ~$7.2b, which we believe also includes a 60,000 sheet/month Gen 8.5 ink-jet printing OLED fab that will be built next to this new fab at a later date.  Charm Engineering (009310.KS), a South Korean company that produces inspection equipment (AOI) for LCD, OLED, and LED displays, and laser repair and die transfer tools (LIFT), has acknowledged that it has received a $11.2m order from Chinastar for the new fab, along with Top Engineering (065130.KS) and DMS (068790.KS). 
Based on the timing of the construction and the equipment orders and lead time, we pull in our Chinastar phase 1 production expectations for 2024 and push out expectations for production in 2025, although the amount of the change is relatively small.  Given the plans to begin construction of the T8 IJP OLED fab after the T9 project has begun commercial operation, and the fact that T8 will be using IJP instead of typical deposition methods, we keep a production date for T8 open for now.