While it might not have seemed so during much of last year, 2021 was a good year for smartphones on a global basis, and while there are certainly disagreements as to the growth rate of shipments last year (range among sources is 1.0% to 8.9% y/y) it seems everyone has agreed that for the first time since 2017 smartphone shipments grew on a y/y basis.  As we are data source ‘agnostic’ and are wary of outliers, we gather as much source data as possible and average the data to come up with what we believe is a consensus for yearly smartphone shipments, as shown in Figure 1, an increase of 4.7%. 
That said, there has been much comment on what drove smartphone shipment growth in 2021, with much speculation on the ‘premium’ price category.  Since ‘premium’ is a subjective term and is defined differently by data sources, we do not give as much weight to subjective data as we might to objective data, however before we go further, we note that 2020 was an aberrational year due to COVID-19, and based on the trend in the two previous years (2018 – 2019) we estimate that 2020 came up 119m units short.  If we consider 2021 shipments inclusive of that pent-up demand, 2021’s ‘normalized’ smartphone shipments would have been 1.238m units, or down 4.5% y/y rather than up 4.7%, and while this is an exercise rather than an actual estimate, it shows that headlines pointing to the growth of smartphones in 2021 need to be examined further, and even on face value are still down 5.4% from 2019, the year before the COVID-19 pandemic, which might change the word ‘recovery’ in tech headlines to ‘bounce back’.
There were other factors that played into unit shipment growth in 2021, primarily 5G phones, which began shipping in 2019 in small quantities but saw rapid share increases in subsequent years, particularly in China where 5G base station implementations were strongest.  While 5G smartphones currently represent just under 80% of all smartphones shipped in China (less globally), they are just over 50% of new models, which led us to look more closely at a different metric than units shipped.  Taking ‘models’ a bit further, a number of data sources have focused smartphone growth in 2021 on the ‘premium’ category, again a subjective term, so while we did not disaggregate smartphone shipments by price tier for 2021we looked at what we call ‘model availability’ or the number of models available to consumers in each price category over the last 4 years. 
Rather than break down price tiers into subjective categories, we used $100 increments to eliminate all bias and we include data from 1Q ’22, which we annualize to show what this year might look like, although seasonality is not considered.  Again, these are not shipments but the number of smartphone models available to consumers across all brands.  Based on that data it can be seen that the number of smartphone models in the $100 - $200 and the $200 - $300 price tiers has continued to grow from 49.9% of total models in 2019 to 54.5% of total models last year and the top two price categories, $1,000 - $1,200 and $1,200+ saw model availability growth last year of 4.4%, as Samsung’s success with foldables give more high-end choices to consumers.
Rather than make our own subjective categories, we then took the 12 price tiers and created low, mid, and high price tier categories by taking 4 price points each from the data.  Looking at this data (see Figure 4) the lowest price tier has seen the number of models available to consumers rise over the last three years, albeit marginally, the mid price tier has seen the number of available models decrease and the highest price tier has been less consistent, although it increased in 2021.  This data, while it does not relate directly to shipments, it does give some hints as to brand positioning and consumer demand, and could be applied more readily to sales value data, but while headlines read that ‘premium’ smartphones were leading the recovery last year, our data shows a bit less enthusiasm for high end smartphones and a steady demand for low end smartphones.   All in it is another few data points in the battle to lessen the impact of headlines on investors in the CE space and make data more reliable.

5G is really two things, the 5G that is commonly used by carriers, which is based on sub6 spectrum bands, and mmWave, which is based on higher frequency bands.  As we have noted in the past, while 5G overall allows for faster connection speeds and lower latency than 4G systems, that performance is directly based on the spectrum used, with the speed increasing as the frequency increases.  The thought would be “why use lower frequencies when higher ones are faster?” and that is logical except for the fact that as the frequencies increase the signal loses strength, with the lower 5G frequencies being able to broadcast over miles, while mmWave frequencies, despite their very high speed, can travel less than a mile and are easily blocked by buildings and even trees. 
This creates a significant problem for carriers that want to provide mmWave service to commercial customers who require the highest possible speed and lowest latency as bringing the mmWave 5G signal into an office or factory can prove challenging with the building itself blocking the signal.  There are antenna receivers for mmWave 5G that can be mounted outside a window to bring in such service but they block the view and reduces light.  One solution has been to install the antenna and the CPE equipment directly to the window glass, but was found to generate enough heat to crack the glass, especially in cold climates.  It seems that Asahi Glass (5201.JP), aka AGC, has developed a 5G mmWave phased array transparent antenna that uses liquid crystal[1] and therefore does not block the window or generate enough heat to break the glass.  The liquid crystal is used to change the direction of the incoming RF, in the same way it is used to block or allow light to pass through an LCD display.
That’s the good news however AGC does not expect to have a mmWave product out until 2024, although they have a similar product for sub^ 5G, which is unfortunate as mmWave is a technology that actually fulfills the promise of high speed 5G for businesses.  There are less attractive ways of bringing in mmWave signals to an office, but most involve bulky antennae or similar configurations mounted to buildings or nearby structures, but they will have to do until AGC is finished with their development.  Then we can say, ”What light through yonder window breaks?”


[1] Actual LC phased array antenna design comes from ALCAN Systems GmbH
 
 
 
 
 

​Ennostar (3714.TT) has indicated that it will issue 70m shares in a private placement to finance the construction of a new 6” Micro-LED wafer fab.  67.25m shares of the placement will be purchased by AU Optronics, who holds a board seat at Ennostar and 2.75m shares will be purchased by Innolux (3481.TT) (no board seat).  The price will be determined based on either the average closing price 30 days before the issuance or the average of the 1st, 3rd, or 5th day closing price prior to the offering, but not less than 90% of those values.  AUO will hold a 17.31% stake in Ennostar when the deal is closed
We note that AUO and Ennostar, through their relationship with Apple (AAPL) have been working toward the development of Mini-LED and Micro-LED processes to further commercialize those technologies.  We believe Ennostar sees further commercialization of the technology and is building out capacity to develop a Micro-LED product line that can be sold to CE companies.  There are still a number of significant bottlenecks and costs associated with the production of Micro-LEDs that make such products too expensive for mass market devices, however as those bottlenecks are removed, the technology has the ability to change the face of the display space down the road. 
One of the more significant challenges to Micro-LED production is that red, green, and blue Micro-LEDs must be produced on separate wafers and then transferred to a new substrate.  Given their small size and vast numbers (there are 8,294,400 LEDs of each color in a 4K display), this can be a time-consuming and therefore expensive process and can lead to a number of damaged LEDs that must be repaired.  One solution that has been proposed is to produce Micro-LEDs of only one color (blue) and use red and green quantum dots to convert 2/3 of the Micro-LEDs to the proper colors.  If this can be done on a usable substrate, it would eliminate or reduce the need to transfer millions of small structures and reduce production time, cost, and damage.  That said, much of this is still in the developmental stage and we expect Ennostar to be working more toward a less mass production type product, perhaps one for an AR device where the size of the display or projector would be small, but in any case they seem to have AUO’s confidence that the project will have a positive long-term result.

​Following moves by a number of Taiwanese companies, AU Optronics (2409.TT) indicated that it will be reducing capital by NT$28.82b ($1.001b US) through a cash dividend of NT$1 ($0.03 US) and the cancellation of ~20% of the company’s 9.6b shares this year.  The company also indicated that it expected to return to shareholders a total of NT$55.77b (~$1.937b US) over the years 2022 through 2024, of which NT$26.95b (~$936.28m US) would remain in 2023 & 2024.  AUO reported its best year in 14 years in 2021 and has already announced plans to build a new Gen 8.5 LCD fab in Taichung which is expected to begin production in 2025, (we expect it might be a bit earlier) and which will see its 2022 capex increase from NT$17b (~$590.6m US) last year to NT$45b (~$1.56b US) this year, with a total cost (including other fab upgrades) of between NT$100b (~$3.47b US) and NT$150b (~$5.21b US).  The capital reduction will be voted on at the June shareholder meeting.
AUO’s Taiwan shares declined by 1.96% last night with investors seemingly disappointed at the dividend value, although the total payout represents 46.5% of the 2021 full year EPS.  It has been a tough 12 months for AUO’s stock price, peaking at NT$33.35[1] on April 28, 2021 and falling to a low of NT$16.80 on October 4, 2021, despite strong results, as investors have remained concerned about the volatility seen in LCD panel prices.  AUO has and continues to push a transition away from generic LCD panel production toward premium and specialty panels, which represented ~ 50% of revenue in 4Q ’21 and should lower the company’s sensitivity to typical display industry cyclicality.


[1] Closing Price

​As far back as 6/24/20 we noted that strained relations between Japan and South Korea has pushed the South Korean Ministry of Trade to sponsor a Fine Metal Mask ‘bakeoff’ pitting a number of South Korean companies against each other to find the best production methodology to develop an in-country supply chain, allowing South Korea to become less dependent on Japan for this critical part of OLED production technology.  The South Korean Ministry of Trade chose APS (054620.KS) (laser patterning) and Poongwon Precision (pvt) (etch) as the winners in the project, leaving Philoptics (161580.KS) and Olum Materials (pvt) to work on their own. 
Much of the momentum behind the South Korean government’s interest in Fine Metal Masks is due to the fact that the world’s largest producer of OLED displays, Samsung Display (pvt) buys almost all of its masks from Japanese suppliers Dai Nippon Printing (7912.JP) and Toppan Printing (7911.JP) who dominate this market.  The masks, which must meet extremely fine tolerances in order to correctly pattern OLED materials, must be cleaned and replaced regularly, and any disruption in mask supply could slow or halt OLED display production, regardless of whether the reason is political or production related. As the material used to produce such masks is made from Invar, an iron/nickel alloy (36% - 42% nickel) and the price of nickel has increased from $20,912/ton at the end of last year to $35,500/ton now (with a peak of $48,196 in the interim), as Russia is a top 5 producer, the sensitivity toward ensuring mask supply and proper pricing has returned.
To make things even more difficult, there are a number of ways in which the masks can be made, with Dai Nippon and Toppan using an etch method while APS uses laser patterning and Philoptics uses electro-forming, a method that deposits the metal on a mandrel.  When the South Korean government created the ‘bake-off’, much was to see which technologies would be most practical, with laser patterning and the more standard etch winning out.  That said, Philoptics has continued to refine it electro-forming process, despite its bakeoff loss, and has been supplying evaluation product to a Chinese OLED panel producer.  The company expects that the evaluation process, which began late last year will take roughly a year and will begin to generate sales (if accepted) in 2023.  The Chinese OLED producer is looking to add a mask supplier as it depends entirely on Japanese mask producers and Philoptics also hopes to sell to Samsung Display as well, although a similar qualification process would have to be completed.

​We have noted a number of times that Apple’s Air Tags and similar tagging devices have been used surreptitiously to track ex-wives, girlfriends, expensive cars and other items by stalkers and thieves without the knowledge of those involved, and Apple and others have taken this seriously by adding features that alert potential ‘trackees’ when they are near a device that they do not own or is moving with them.  There are other even less expensive devices that can provide such illegal functions, but Mr. Lawrence Welch, formerly of Nashville and now a resident of the local jail, decided that he was going to up the ante a bit when trying to track his ex-girlfriend.
Mr. Welch, who is already facing two domestic assault charges from almost a year ago, found that his ex-girlfriend had turned off the Life360 app that they had previously been using to keep track of each other and began texting demands that she call him and reveal where she was.  He did manage to track his ex down to the local Family Safety Center where she was filing an order of protection against him but was spotted leaning down beside the front passenger side tire of her vehicle and the police were called.  When they examined the car they found that he had placed an Apple watch on the wheel, a bit more expensive (~$300) than an Air Tag (~$30).  Upon questioning Mr. Welch admitted that the watch was his and was arrested.
While there is no anti-stalking feature on the Apple Watch, it is tied directly to an Apple account along with a cellular plan which makes it quite easy to trace back to an owner, and Apple does cooperate with law enforcement to identify owners based on the devices’ serial numbers, leading to where and when the watch was purchased, so the owners identity can be found.  But there is no anti-stalking protection on the Apple Watch as it is likely that Apple would not expect someone to use a $300 watch as a tag, so while Mr. Welch inadvertently found a way to track his ex that would not have alerted her, it was both accidental and a ridiculous use of an Apple Watch.  Perhaps Apple will add some protection to a Watch OS upgrade when this incident gets back to them but for now be on the lookout for an Apple Watch attached to your tire if you are having trouble with an ex-boyfriend…

Samsung and Apple have stopped shipping smartphones to Russia.  Given that Russia is among the largest smartphone markets (6th globally as measured by number of users) and has the 2nd highest adoption rate based on population, this is a significant blow to both brands.  In term of vendor share currently Apple holds a slight lead over Xiaomi (1810.HK) and Samsung, with Huawei (pvt) and RealMe (pvt) with lower shares, and Android is the obviously dominant smartphone OS.  Google has stopped Russian users from making purchases from the Google Play Store or YouTube although free apps will still be available, and Russian state-owned media outlets have also been banned. 
This has set the Russian government toward the potential that Google will cut off all Android support to Russian smartphone users or will limit licenses for new phones and has pushed the government to examine other operating systems going forward, including Huawei Harmony OS, which is the alternative OS used by Huawei in China, after its ban from trade with the US.  Huawei has indicated that while it has sold 220m devices with their OS, they have no plans to sell any phones with the Harmony OS overseas.  Unfortunately, that does not rule out the licensing of the OS to others, such as Russian smartphone producers, who can use it in their phones to replace Android going forward, giving Huawei an income stream without selling phones into the country.
That said, the Russian government is urging citizens to not worry about Google, Samsung, or Apple and purchase the domestically produced smartphone, the AYYA T1, which was developed by Smartecosystem (pvt) a division of Rostec, (State Corporation for Assistance to Development, Production, and Export of Advanced Technology Industrial Product Rostec), a state-owned organization created by V. Putin in 2007.  The phone uses a Mediatek (2454.TT) Helio processor and has a 6.5” screen, with the retail version running on Android while other models are available with the Aurora OS, developed by Jolla Ltd (pvt), a Finnish company, although the code is open-source.  The AYYA’s claim to fame is that the phone has a button that disables both the cameras and the microphone to allow users to be free of potential surveillance.  The phone sells for what was 15,000 to 19,000 rubles, which at the beginning of this year was equivalent to $170 - $255 US, although now is between $156 and $198.
 

Back in late January we noted that battle between Epic Games (pvt) and Apple (AAPL) over 3rd party payment restrictions for applications in the Apple store has blossomed into a global ‘review’ of Apple’s store practices by a number of governments.  The ruckus was over Apple’s draconian policies concerning the use of ‘alternative’ payment methods on the app store where Apple App Store application developers forbidden from selling subscriptions directly or outside of the AAS in-app purchasing system, but had to give 30% of that subscription revenue to Apple, unless the user signs up for longer than 1 year, when the fee drops to 15%.
Apple made some minor concessions to what is a mounting tirade against such limiting policies however Japan’s Fair Trade Commission forced Apple to allow some applications in the store to direct users to their own websites to sign up for subscriptions, while South Korea passed legislation that banned all application store operators from forcing developers to use in-store payment processing however just before Christmas last year the Netherlands Authority for Consumers and Markets published an order requiring Apple to “adjust the unreasonable conditions in its App Store that apply to dating-app providers”, as Dutch dating app providers are currently unable to choose a payment system.  The ACM went further in that the required Apple to adjust the system within 60 days or it would be fined €5m per week, up to a maximum of €50m.
Apple has made some attempts to comply with the new order although they were not enough to satisfy the ACM which continued to levy weekly fines against the company and last week the European Union published its Digital Markets Act, which details requirements that Apple and other firms (Google (GOOG) tends to be a 2nd focus) to offer alternatives to the App Store and its payment system restrictions., with those rules to be implemented by October.  It seems that the momentum against what many consider anti-competitive practices by Apple are not going away easily.  To make matters worse, the fines in the Netherlands, which we originally believed would max out at €50m, do not, and the Dutch authorities have the option to  increase the weekly rate once €50m is hit, which should be in early April.