​2/3rds of the Way through the 2nd quarter panel prices continue their rapid increases with only a few exceptions.  Monitor panel prices increased by 5.8%, almost identical to the previous month and the average monthly increase (5.7%) seen so far this year.  Monitor demand continues to be strong enough to support price increases on a general basis, but shortages of a number of components, particularly controllers, allow panel producers to remain in a favorable bargaining position which has allowed them to offset lower unit volume with higher prices.  The aggregated current price for monitors is currently 24.4% above the high point last year and 51.5% above the low (see Fig. 5 for earlier years).
Notebook panel prices increased by 3.9% in May, a bit below the previous month’s gain, but again, almost on the average gain for this year of 3.7%.  Demand also continues to be strong, led by a number of global educational purchase programs that have continued this year, and the same component issues as mentioned above.  While we expect some of those high volume purchase programs to end by the end of 2Q, the component shortages have pushed out delivery for panels and extended lead times for notebook delivery, which could extend demand strength into the summer a bit.
Aggregate TV panel pricing was up 3.7% in May, showing a much lower increase than last month (7.3%) and lower than the average increase so far this year (5.6%).  While calling a 3.7% monthly increase ‘weak’ might be overstating the facts, given the last 12 months, especially the last 6 or 7 months, the increase was unusually low, and while TVs face the same or greater semiconductor related shortages as other panels, the aggregate price of TV panels is now up 28.7% from last year’s highest point and up 107.3% above last year’s low.  TV set producers have increased set prices at least once this year, with some twice, and are worried that another price increase, which seems inevitable will begin to have a more significant impact on demand.
North America has been the growth leader in the TV market, with stimulus checks and continued stay-at-home attitudes keeping demand high, especially relative to China and Europe, where demand is weaker, but as COVID-19 restrictions are lifted, and TV set prices continue to rise, we expect demand to slow, which it did in April.  There are still TV stock outs at TV retailers in the US, as panel supply is constrained by component shortages, but we see the continued upward pressure on set prices as a potential factor that could change the positive outlook that set makes still seem to have. 
While we have not seen any major changes to TV set yearly targets, the somewhat bleak outlook for any significant change in semiconductor supply leads us to expect that component shortages will continue into 3Q, which will limit panel volume and keep upward pressure on TV panel prices, which will have an increasing effect on demand.  Whether May’s ‘weaker’ panel price increase reflects a shift in TV set demand is moot as one month certainly does not make a trend, but should we see a lower rate of change continue, it could signal a change in demand over the next few months.  A bit too early to tell, but certainly a possibility.
Mobile phone panel prices did not change in May, and while that would be below this year’s average increase of 1.2%, mobile phone panel prices have been much more stable than other categories.  While demand spikes do occur when new release inventory building arises, smartphone demand has not seen the increases that other CE product categories have this year, and Chinese demand is beginning to look like it is resuming its more typical weaker trend, with April seeing negative y/y growth after positive y/y growth in the 1st quarter.  The COVID-19 issues in India are also a drag on smartphone demand, all of which give us the view that mobile panel prices will see relatively mild increases in 2Q overall and potentially for the remainder of the year, other than the typical seasonal spike in September.

Large panel LCD sales saw a decline of 2.1% in April after a large increase in the previous month.  Large panel shipments also declined (-6.6%) but large panel ASP increased by 4.8% offsetting much of the unit volume decline, while on a y/y basis the large panel segment was up 65.6%.  On a regional basis the industry saw growth in large panel sales in only one region, China, where sales were up 4.0% m/m and up 82.1% y/y.  On a weighted average regional basis, the industry saw a decline of 1.5% in sales in April, but on a weighted average basis saw y/y growth of 50.6% this year.
Among those panel producers that are focused on large panel production[1] those with positive m/m April results are in the 1st column in order of gain and those with negative m/m sales are in the 2nd column from largest to smallest.  We note that while Panda (600775.CH) shows up as a big gainer in April on a m/m basis, the sale of two of their fabs at the end of last year, makes their y/y April comparison -57.9%.  Both Samsung Display (pvt) and LG Display (LPL) have also rationalized their large panel LCD production, but LG Display still remained up 18.3% y/y in April, while Samsung Display was down 58.3% during the same period.
Panel producers remain in an advantageous position, with demand outstripping supply, but the ASP increases seen each month must be maintained going forward as shipments decline, a result of component shortages, with mix shift limited by those same shortages.  Large panel sales, on a m/m basis, have declined each month this year other than March, which saw a large (13.4%) increase as the New Year holiday in China ended and inventory needed to be replenished.  Thus far large panel demand has remained strong enough to offset component shortage based delivery push-outs, but at the first sign of weaker demand, panel producers will begin to see their absolute control over panel pricing erode.  That said, thus far this year, even with relatively flat unit volumes, panel producers have been able to show very strong y/y results, but will see those comparisons getting more difficult for the remainder of the year.  Without the aggressive panel price increases we have seen over the last year, the game will become more difficult for panel producers going forward.


[1] We exclude Hannstar, IVN, Japan Display, Panasonic, and Tianma

​RealMe (pvt) is a smartphone brand that was spun off of Oppo (pvt), one of the largest smartphone brands in China, and while they are relatively new (only 3 years old), it has commanded considerable attention as it has been growing rapidly, particularly in emerging markets, and has grown from a miniscule share in 2018 to over 4% of the global smartphone market currently, commanding a 12% share in India and is in the top 5 in a number of Southeast Asian countries.
While RealMe is growing rapidly, they are still facing the same issues that other smartphone brands are facing on the supply side, and do not have the same bargaining power as volume giants such as Huawei (pvt), Samsung (005930.KS), Apple (AAPL), or Xiaomi (1810.HK) to gain leverage.  That said, they have been particularly up front about the rising cost of producing smartphones, particularly as their phones tend to be mid to lower-tier priced models (Their latest, the GT Neo, sells for ~$315 and they have phones priced down to $50) and recent comments by senior executives indicated that they expect prices to increase significantly in the 2nd half of the year, citing storage, charging components, and batteries as some that are seeing rapid price increases.
The company expects that these shortages will continue through the 2nd half and will push smartphone prices up by ~10% in 2H, and indicated that some local (China) smartphone manufacturers have lowered orders and shipment targets.  April, as we have noted previously, was a weak month for smartphone shipments in China, down 23.8% m/m and more importantly down 34.1% y/y, after three months of positive y/y growth, although we note that those y/y comparisons were relatively easy given the timing of the initial outbreak of COVID-19 in China.
Unfortunately such a price increase would begin to erode the smartphone price declines we have seen among major smartphone brands, who are trying to appeal to a more price conscious customer, rather than feature oriented buyers, which has helped to at least stabilize the smartphone business during the pandemic.  While a 10% increase might not be enough keep a replacement phone from being purchased, we expect the actual production costs in 2H could be greater than 10%, which means that the price increase will not cover the added cost.  Brands will expect the retailer to share some of the reduced margin, but that just means the retailer will be incentivized to sell higher margin products, a catch- 22.  

Last week (5/20/21) we noted the pricing on Samsung’s new line of QLED and QLED/Mini-LED TVs, ranging in price from $600 to $9,000.  Surprisingly we noted today that Samsung is offering discounts on a number of these TV sets, including selected models of the mini-LED line, which is surprising this early in the release cycle.  Of the 32 models and sizes that were released 10 are being discounted, 53% of which are mini-LED/QD models, and the discounts range from2.9% to 12.5% as shown in the table below.
It is hard to draw any conclusion as to why Samsung is offering these ‘limited time’ discounts on such new models, other than as a way to stimulate sales that might have been less than expected in South Korea, where they were released, however it is too early to gather real data on sales for these models.  Samsung is likely seeing improving pricing on mini-LED backlight units as volumes increase and could be passing such on to customers, but we assume that the units might have been initially overpriced, and competition from others, particularly TCL (000100.CH) has put some pressure on Samsung’s pricing.

​As sort of a final touch to his master plan, then President Trump, just a week before he left office, decided to add Chinese smartphone brand Xiaomi (1810.HK) to the list of companies his administration alleged were owned or controlled by the People’s Liberation Army and was therefore a “Communist Chinese Military Company”.  The list (See our note of 5/19/21 for list), maintained by the US Department of Defense and the Department of the Treasury, broadly includes a number of Chinese companies that are directly or indirectly tied to the Chinese military and a number of companies that are not, but the list Xiaomi was placed on was one that singled out certain Chinese companies that were publicly traded either in the US, China or Hong Kong.  The Executive Order surrounding the list instituted a ban on investment by US citizens and entities in these companies.
In some cases these companies protested, particularly Huawei (pvt) and ZTE (000063.CH), who were also banned from selling to or buying from US companies, or any company that used US goods to manufacture their products.  Unfortunately for Huawei, their constitutional challenge to the law that was the basis for the EO was rejected by a US District Judge in a summary judgement in February of last year, and the company has been struggling under the ban ever since.  Xiaomi however took hope in the fact that the Trump administration’s ban on Chinese video sharing app TikTok (pvt) had been blocked by the District Court and in February of this year filed a suit against the Trump administrations investment ban.
Xiaomi announced yesterday that it had received a final judgement ruling from the District Court of the District of Columbia that lifted the Department of Defense’s recognition of Xiaomi being a ‘Chinese military company’ and formally revoked all US investor restrictions on buying or holding the company’s securities.  In March the Xiaomi ban was temporarily blocked by a Federal judge who called the decision to blacklist the company ‘deeply flawed’ and the Biden administration has indicated that it would not challenge the ruling stating that ‘the Trump administration failed to develop a legally sufficient basis for imposing restrictions on the company and compelled this action.”  While we expect others, particularly Chinese semiconductor foundry SMIC (688981.CH) to file similar suits against the investment ban, we also expect the Biden administration to draw a line at how much leeway they will give some of the companies on the list as not all are as clear in their relationship with the Chinese military.

​BOE (200725.CH) is now expected to begin production on the first phas4e of its Gen 6 OLED fab at the beginning of 4Q this year.  This fab, known as B12, will have a total capacity of 48,000 sheets/month, although the initial phase will be 16,000 sheets/month.  We had expected the fab to begin operation in July, so this lowers our expectations for this fab this year by ~33,300 m2.  As this also pushes back our expected phase 2 and 3 start dates to March and October of 2022, total capacity for this fab in 2022 is also reduced, although we believe the timing for phase 2 and 3 (March and October ’22) is still a variable.  BOE has announced plans for a 4th Gen 6 OLED fab in Fuzhou, but that project has yet to be started.
We note that BOE is providing relatively small quantities of flexible OLED displays to Apple  for the current iPhone line but has yet to be named as a primary OLED supplier, so it is essential that they continue to increase their OLED capacity to show Apple that they have the capacity to be a reliable primary supplier, particularly in regard to LTPO displays, which are now produced at the B11 fab in Mianyang.  The B12 fab will also be capable of LTPO backplanes, which will, in theory, give BOE the opportunity to compete with Samsung Display (pvt) and LG Display (LPL) for that business. 
By mid-2023 BOE will have matched Samsung Display’s Gen 6 OLED capacity, in its challenge to unseat SDC as the leading supplier to the iPhone line, with LG Display currently running two 15,000 sheet/month lines in Paju and is expected to start production on a 3rd 15k line in May (tentative), but even with the new line at capacity LGD will have capacity for 45,000 sheets/month dedicated to Apple, along with another30,000 sheet/month G6 line in Gumi that is not dedicated.  Given that LG Electronics (066570.KS) has phased out their mobile phone business, LG Display could upgrade the Gumi line to LTPO, but even with that upgrade, they would have a maximum capacity of 75,000 sheets/month, while both Samsung Display and BOE would have 140,000 sheet/month capacity.
One important point that usually gets left out of fab capacity information is yield, which is the basis for actual production volumes and the profitability of the fabs, and BOE has been suffering from lower flexible OLED yields at it fabs for some time.  Given SDC’s expansive experience with flexible OLED production, they have a distinct advantage in terms of their ability to meet timeline production targets, something Apple is particularly sensitive about.  While some customers might be attracted to BOE’s flexible OLED pricing, which we assume is a bit lower than SDC’s, SDC’s ability to produce at high yields gives them the advantage of reliability and higher margins.

1Q was a good quarter for semiconductor companies with the top 15 sales leaders (foundries and fabless) all showing y/y sales gains except for Intel (INTC) and the same group, even with Intel, showing a 21.1% sales increase y/y, according to data provided by IC Insights.  Ranking changed a bit, although the top 5 remained the same as last year as indicated by the colors in column 1.  We also look at Worldwide Semiconductor Billings, which are calculated on a 3 Month Moving Average basis..  That data shows that the industry billings were up 15.2% y/y in 1Q, which is higher than the 5 year average of 11.3% and the 10 year average which is 9.5%.  We note that 2020 was a bit of an aberrational year, particularly 1Q for much of the world although the semiconductor industry was affected less than many, so we look to the trend line in Fig. 1 as a better reference to where this year’s 1Q industry results actually fit, and that shows that 1Q this year is above the line and the highest on record.

​As we start to see data on TV shipments for 1Q, a very regional picture is developing (sorry).  North America was strong, in fact the strongest 1Q on record, while China and Europe were weak, with the strength attributed to the recent economic stimulus.  Total TV set shipments are falling into a range of between 50m and 51.5m sets in 1Q, up from ~45m units in 1Q last year., or up between 11% and 14% y/y, and while TV set prices continue to rise against increasing panel and material costs, some TV set shipment estimates for 2Q are being increased, but not for the right reasons.  Those increases are based on the inability of set manufacturers to source both panels and electronic components, which have moved lead times out enough that shipments that should have been made in 1Q have now fallen into 2Q, while at the same time some of the more aggressive yearly set forecasts are coming down slightly. 
While most other CE product y/y comparisons are of little value given last year’s unusual 1Q circumstances, TV set sales have been more akin to the stay-at-home mandates seen over the last few quarters, particularly in North America, reflecting a strong desire to be entertained.  We do expect rising prices and an inability for brands to aggressively discount to have some effect on 2H sales, although with recent COVID-19 outbreaks and potential lockdowns, it is hard to gauge how much control over the pandemic will be seen in 3Q and 4Q and therefore global demand. 
As TV brands make little margin on smaller sets, the tendency is to push toward the ‘premium’ market, defined as over $2,500, but Samsung remains the leader in that space, as it does in the overall TV set market.  Samsung (005930.KS) saw roughly 17m units shipped in 1Q, giving it a ~33% share of the global market and a ~47% share of the premium market, much of which was driven by Samsung’s QLED TVs (>2m units in 1Q) where it holds a ~75% share of total QLED TV set shipments, and their domination of the 75” and 80” markets, where they hold a 46.5%[1] and 52.4% share respectively.
While LG Electronics (066570.KS) holds 2nd place in global shipments with ~10m units in 1Q, or roughly 1 19% share, far below Samsung, however LG grew its OLED TV shipments in 1Q over 110% y/y as it ramps production capacity in China and is expected to sell almost 6m OLED TV sets this year, roughly 2/3 of the overall OLED TV market.  While Sony (SNE), TCL (000100.CH) and HiSense (600060.CH) were in the top 5, their combined share is just a bit over 22%, while the share of LG and Samsung was over 52%, despite the fact that both  company’s panel supplier affiliates have reduced large panel LCD TV production. 


[1] OMDIA – 5/25/21

Micro-LEDs are a numbers game, with 24.9m very small LEDs needing to be transferred from a wafer to a display substrate in a cost effective manner.  As we have noted previously there are many ways in which the display industry has been approaching this problem,  a major stumbling block to the commercialization of micro-LED displays.  They range from sophisticated pick and place tools, fluidic self-assembly, where micro-LED die ‘float’ into place on the substrate in a liquid, stamps that stick to the die, to laser based tools.  Each of these processes has its pluses and minuses, with tradeoffs usually being speed, accuracy, and pre and post steps.  While some of the transfer processes we have seen can move millions of micro-LEDs per hour, many involve pre-processing before the transfer takes place or a bonding step as or after the transfer is made.  While these are in most cases separate from the transfer process itself, they have to be considered part of the transfer and included in the total transfer time, which can then become radically different from the stated transfer rate of the tool.
While this seems nitpicking, it has a significant effect on the overall cost of producing micro-LED displays, and as the feature size of micro-LEDs continues to be reduced toward 5um or less, the cost increases.  In order to solve this problem, and the problem of repairing those micro-LEDs that were damaged during the transfer (even at five 9’s there would still be at least 249 LEDs that would need to be replaced), the Electronics & Telecommunications Research Institute (ETRI) in South Korea has come up with a solution called SITRAB, which stands for Simultaneous Transferring and Bonding, or simply a way to transfer and stick the die to the display substrate in one step.  By using a laser (aka Laser Assisted Bonding or LAB) to heat this material (known as an anisotropic solder paste or ASP) the die can be picked up and bonded as they are placed, but instead of a very precise laser that steps through a bond for each die, a wider beam laser can be used  to bond a large number of die at once.
According to ETRI this will reduce the time it takes to transfer an bond I reduced to 1/10th of normal, and since the ASP is easily applied and bonded, repair time can be reduced to 1/100th of normal.  But there is a catch, although the researchers say the equipment investment cost is low and domestic equipment could be used, but so far ETRI has been able to produce a 100 mm2 display that contains 1,225 micro-LEDs, which would be the equivalent of just over 76 pixels/inch for a mono-chrome display and 25 PPI for a full color RGB display, which is far below commercial pixel densities.  Taking it one step further, ETRI says the process could be adapted to related products within two years, which seems quite optimistic in our view.
As noted this is one of many potential ‘solutions’ to the issues facing micro-LED commercialization, which we believe will see viable product in 2025 at the earliest, with new materials and processes being developed almost daily.  The good news is that considerable sums are being spent on the development of micro-LEDs under the expectation that it will eventually be, in some form, a replacement for LCD technology, and those with key IP and process technology will have an edge over rank and file display producers.  That said, success is certainly not guaranteed, and what eventually becomes a high volume micro-LED product could look completely different from what we have seen thus far.

​Monitors are the Rodney Dangerfield of the CE space, typically a simple rectangular display plugged into a PC, doing little other than converting information, video, or similar data into something visual.  Until a few years ago, they were merely those same simple boxes that have been on desks for over 70 years, although LCD monitors replaced CRT (Cathode Ray Tube) monitors around the turn of the century.  Gaming has helped the monitor gain a little respect, as requirements such as high refresh rates, high resolution, and wide formats, have pushed gaming monitors forward a bit, but CE brands, looking for ways to capitalize on the multitude of digital devices that surround us, are presenting us with another ‘smart’ device, the smart monitor.
So, what does a ‘smart’ monitor do?  Does it talk to your ‘smart’ refrigerator? (“I saw you talking with that new toaster last night.  She’s hot!”) Does it speak to your ‘smart’ doorbell? (“Get those kids off the front lawn!”).  None of the above, but here’s what the new Samsung M7 ‘Smart Monitor” can do, as an example.

1. Its 43”, so it’s the size of a medium size TV (4K).
2. It is directly connected to your Wi-Fi, which means you can view content from OTT or streaming services without it being connected to a PC.
3. It can be connected to a PC (Windows or Mac) and supports Microsoft 365 apps through the cloud.
4. Amazon (AMZN) Alexa, Google (GOOG) Assistant, and Samsung’s Bixby are all supported.
5. You can mirror your smartphone just by tapping it on the screen
6. It has USB Ports and Bluetooth and is rechargeable and the remote is battery and solar powered.

No longer do CE companies try to attract consumers to such devices by calling them ‘hubs’ (actually some still do), but such ‘smart’ devices are still vying for your attention and are getting closer to being the kind of device that serves a multitude of purposes, replacing your smartphone when you are home.  While you can buy a 43” 4K LCD TV for under $300, this ‘smart’ monitor can do a few things most TVs cannot, and cost a mere $578, but give it another year and you won’t be able to tell the difference between a TV and a ‘smart’ monitor.