​The Data Protection Commission in Ireland issued a decision concerning Meta’s (FB) use and transfer of data that was collected in the EU, under its 2018 GDPR (General Data Protection Regulation), that was adopted by the European Data Protection Board.  The decision, the end result of a battle between the EU and Meta over the transfer of customer information from Meta Ireland to Meta US originally led to a judicial review in Ireland back in 2020 based on a draft decision made way back in 2016 over the same issue, which states that Meta ‘failed to guarantee a level of protection to data subjects that is essentially equivalent to that provided by EU law.”
The ruling ordered that the data transfers be suspended and processing and storage of the data in the US also be suspended, along with a €1.2b ($1.33b) fine, to own knowledge the largest fine ever levied under the GDPR.  The original complaint even cites leaks by Edward Snowden, indicating that the NSA was operating surveillance programs on systems operated by some of the largest technology companies, as a basis for the illegality of the transfers to US soil, which tacitly the US confirmed.  More to the point however was the accusation that the data was both collected without the user’s permission, and was used for purposes that user did not agree to
The ruling, which gives  Meta  6 months to halt the transfer practice, will be appealed by the EU, and will likely drag out for many months, as had the €746m GDPR that was levied against Amazon (AMZN)  in July 2021 (still unpaid), although the GDPR has certainly been a major step forward toward the protection of privacy, at least in the EU, and a similar edict should, in our opinion, be legislated in the US.  That said, with everyday partisan politics in the US, it is hard enough to keep the country from defaulting on its debt, let alone pass legislation that might cause major companies in the US to think before they mine every scrap of data and sell it to whoever might pay for it.  We can dream, right?

In order to reduce the effects of differentiated data sets, we tend to use averaged data from a number of sources and don’t draw conclusions until we have at least a few data points for each category.  That said, there are occasions when one data source is either early in their data collection or estimates, far out of line with others, or unusually different from what we might expect.  In those cases, unless we have empirical evidence that the outlier data is incorrect, we note that to our readers.  Much has been said, particularly by panel producers and CE brands about how the CE space will recover (usually with the ‘gradually’ caveat added), although there has been little, other than seasonality, to back up those predictions.  While seasonality typically does play into a better 2nd half, for any real sustainable recovery there has to be an underlying improvement in demand.
September is typically the strongest month (5 yr. avg.) for notebook panel sales, while March usually shows the biggest m/m sales gain, given that February typically records the lowest notebook panel sales month of the year.  That leaves April sales typically down 8.8% m/m, as CE ‘March Madness’ ends after the Chinese holiday in February, typically the worst m/m decline of the year.  What all of this is leading to is that one data source, Digitimes Research, is reporting that the top 5 notebook brands (excluding Apple) saw a 30% drop in m/m notebook sales in April, more than 3x the average and greater than any w/w April decline in notebook panel sales since at least 2011.  DT cited strong short-term orders in March for the large April decline, but also noted ‘a lack of confidence in end demand’ and relatively high inventories in the retail channel.
The DT data is only one source, and one that tends to be a bit on the aggressive side, but it was so far out of line with the tacit ‘gradual improvement’ scenario promoted by the industry, it is worthy of mention, although we are not surprised at the decline, only the magnitude as there has been little real change in demand after the category saw COVID pandemic demand peak in November of ‘2021.  While notebook panel sales do not always track against retail sales, the implication of a 30% decline in April’s notebook panel shipments would bring notebook panel shipments to a point slightly lower than January shipments, which were only slightly higher than panel shipments in February of 2020, the first month of the COVID pandemic, which would not be a good omen for the gradual recovery scenario, at least for notebooks., which last year represented 26.95% of panel shipment unit volume.
Before taking possible scenarios for the remainder of the year any further, we need to confirm that data with other sources, which tend to show up toward the end of the month, so we will wait to see if other sources confirm the data drop.  The top five notebook brands (excluding Apple, who would be #4 in 2022) are Lenovo (992.HK), Dell (DELL), HP (HPE), Acer (2353.TT), and Asus (2357.TT).
​

Chinese panel producers tend to be an optimistic lot, at least publicly, likely somewhat influenced by the fact that much of their funding comes from the state and local governments, who need to remain positive about their investments.  As listed Chinese companies are required to post at least a summary of conversations they have with analysts and investors, and maintain an on-line Q&A presence, there are times when one can get some insight into what the company is saying publicly, and at times a bit of nuance toward what they are really thinking.  Two weeks ago China’s largest LCD and OLED panel producer BOE (200725.CH) posted its analyst and investor meeting summary, which, if nothing else, gives some indication as to what the company’s near-term plans are based on.  Here is a summary of BOE’s most recent meeting with analysts and investors (139), and a few comments.

Industry Overview

• In the 1st quarter the market supply and demand gradually stabilized.
• LCD TV panel prices started to rise, driven by increased demand and promotional activities.  We would say ‘seasonal demand’ and an emphasis on promotional activities.
• Panel manufacturers adjusted production rates based on demand and focused on healthy development and dynamic control.  Not sure that ‘dynamic control’ is anything more than ‘build to order only.’
• Overall inventory levels were normal, and the industry is expected to gradually recover.  At least back to where it was before the pandemic.

Company Performance

• The company maintained its leading position in the semiconductor display industry across various applications.  BOE holds the largest share (31.5%) of global large panel LCD sales (March) and represents 57.7% of China’s large panel LCD sales.  With LG Display (LPL) #2 at 18.5% global share, BOE is certainly the leader in the large panel LCD space.
• The flexible OLED business showed steady growth, with a target of over 120m shipments for the year. Here is where BOE shines.  The company saw ~4% q/q flexible OLED unit growth in 1Q against an 18.4% decline for the industry and saw an 81.4% y/y unit increase in flexible OLED volume, against industry y/y flexible OLED growth of 27.2% (units).  For perspective, Samsung Display (pvt), the industry leader saw a 20.6% decline q/q and a 2.9% decline y/y.
• Short-term performance was affected by industry and macroeconomic conditions, but the company maintained resilient operations.  On a relative basis, yes, as BOE large panel LCD sales were down 2.7% q/q and down 30.0% y/y while the industry was down 15.9% q/q and down 36.1% y/y in 1Q.
• The company implemented lean management, resulting in improved capital structure and reduced expenses.  Companyspeak

Q&A

• Regarding industry production rates, the industry has been adjusting production rates to achieve a heathier supply-demand balance.  LCD TV panel prices have rebounded, and the production rate is expected to gradually recover.  Based on current estimates for demand and LCD panel capacity, and using an overall 85% utilization rate, there remains considerable overcapacity in the LCD space, so it is hard to call such a situation ‘a healthy supply-demand balance’.  We understand that demand estimates are likely at their lowest at this time of year and in current economic circumstances, but even with some demand improvements, LCD overcapacity will likely remain an issue this year.
• The company’s flexible OLED business has shown significant growth, and the shipment volume is expected to continue to increase.  The company aims to improve the proportion of high-end products and expand into new areas.  We give credit to BOE here as they have done a good job increasing OLED display volumes, particularly with Chinese brands.  They will have to push hard to challenge Samsung Display’s Gen 8 OLED push and a similar one from LG Display, but if nothing else, they are persistent.
• The MLED business is positioned as an important platform for the company’s next generation display development, focusing on various applications in backlighting and direct display.  BOE produces a number of LCD modules with it own Mini-LED backlighting, which puts those panels in the premium category and at a higher-than-average ASP, so we expect this segment will grow throughout 2023, but will also be somewhat dependent on the Chinese economic picture, which is a bit less vibrant than expected late last year.  It’s good business for BOE but higher-end products are in the sights of almost every panel producer and that leaves a concern over the necessity to compete on price to gain a foothold against SDC, which could impact MLED margins.
• The company’s financial performance in the first quarter improved due to cost reduction measures, lower asset impairment losses, and increased product prices.
• The company’s future capital expenditures will focus on strategic planning including enhancing OLED competitiveness, IoT solutions and upgrading manufacturing capabilities.  While BOE has been a bit more careful about capex, despite continued funding from the government, they will have to start spending on Gen 8.6 OLED capacity before the end of the year if they want to get a place in Apple’s OLED IT conversions in 2024.  The company has been working toward the development of Micro-LED displays for wearables, which could become a selling point for the Apple watch down the road, but in the interim they will have to continue to spend on LTPO small panel capacity and IT OLED.

All in and corporate speak aside, BOE’s investor meetings gave only an indication of what management is thinking about the prospects for the current year, particularly for the LCD display business.  Seemingly more confident about the growth of their OLED business, likely as Chinese brands see the company as an alternative to more expensive Samsung displays, and a growing relationship with Apple, albeit with the drama associated with that relationship.  As BOE learns more what it can and cannot do as to its Apple relationship, the company should see its small panel OLED business expand, but the OLED IT space is more of an open field, giving BOE the opportunity to gain an early foothold if it both has the expertise to build out a Gen 8.6 OLED fab that has a reasonable yield and that they can meet Apple’s typically stringent specifications, which have proved challenging for BOE in the past.  That said, they are certainly the most successful Chinese panel producer and have a dominant share of the global large panel LCD space.  Whether that proves ultimately the correct positioning over the next few years, however, is still an open question.

Quick, what is the password for your wi-fi router?  If you know the answer, you are either working for your carrier or you are using the same password for the router as you are for other applications.  If the later is the case, don’t feel bad, as a recent survey by Specops Software (pvt) indicated that 29.03% of respondents use the same password for every application and 34.4% only change their password when they forget the old one, while the average person in 2020 was managing ~100 passwords.  Passwords are obviously a pain, but remain, at least, the frontline of defense against hackers who are trolling for information or looking to steal assets outright.
Simple password entry has given way to 2-step authentication, where when you try to log into an application or account, you are required to enter an ID and a security code that is sent to you, usually on a mobile device.  Once that code is entered on the application you must still enter your password to gain access.  This prevents hackers who have stolen your password from access your data without a secondary device to confirm identity, but you still have to remember the password for each application  There are password managers that help with managing a multitude of passwords by using a master password that gives access to the password manager, or you can do what many others do, jot them down on pieces of paper or in a ‘secret list’ that no one should be able to find.  Of course, pieces of paper are easily lost and a paper list of passwords is an accident waiting to happen, so many keep such lists on devices, in the hope that no one will recognize what they are.
The good news is that steps are being taken to simplify the authentication process in ways that will free you from the burden of remembering, changing, and keeping track of passwords for all of your devices with what is called a passkey.  Google (GOOG) has implemented passkey use when logging into your Google account, also letting you still use passwords, and as of May 3, you can use passkeys to log into Google websites. Apple, Microsoft (MSFT), PayPal (PYPL), eBay (EBAY), and a long list of others are developing or implementing passkey authentication for their applications, with Apple having already built it into the latest version of IoS, so .  By using passkeys you no longer have to remember strings of numbers and characters (Is it a capital V or a small v?) or your 4th child’s birthday, but you do need to have your phone, tablet, or computer nearby, as the passkey system needs to communicate with your device in order to verify your identity.
Passkeys use public-key cryptography to authenticate your access to websites when you register on a site.  At that time the system generates a public and private key, with the public key being stored on the site’s web server as it has no value to a hacker by itself.  The private key remains on your device and when you try to log in to an application it sends a ‘challenge’ to your device.  As the public and private keys have a mathematical relationship, the private key completes the challenge and ‘signs’ a response to the server, identifying you, and the server retains only the public key and still does not know your private key, even though you have been identified.    Your device, however, also checks, via master password, fingerprint, or biometrics, that you are the correct person that the private key will identify, as a safeguard in case your device is lost or stolen.  At no time is any sensitive data exchanged between your device and the server, as does when using standard passwords, making passkeys more secure.
Hackers cannot guess a passkey, or can you accidentally reuse a passkey on another site, and because they are unique to each site, tricks that send you to look-alike sites to collect your password will not work.  That said, because they are unique to each site, you need to set them up each time you open a new account or join a new site, and as the private key still resides on your personal device, care must be taken not to lose the device, although if that were to happen, a hacker would need to know your master password, or find some way to beat your biometrics, which certainly lowers your risk against a hacker figuring out that you only change two numbers in your password across all of your applications.
It will be a while before passkeys become ubiquitous, but with a number of the largest CE companies taking steps to implement the concept across multiple applications, the momentum will build, and standards organizations, such as the FIDO (Fast Identity Online) alliance, are working to set standards that will hopefully make it unnecessary for iOS users and Android users to have different passkeys for the same application.  At that point you would be able to throw away your secret lists and scraps of paper and no longer worry that you haven’t changed your WEP key password in 5 years, letting your next-doors entire family piggyback on your Wi-Fi. 

The EU can be a tough market for CE companies, with all sorts of regulations, such as those that have forced Apple to change its charging connector, or the power consumption regulations that are plaguing the 8K TV market, but the EU is getting ready to make another proposal that will potentially affect a slew of technology companies that network providers say have been riding for free on their networks.  Over the last few years, particularly during the COVD-19 pandemic, network traffic has grown considerably, and while mobile network traffic growth gets much of the press, the growth of all network traffic continues to grow at a rate over 20%/year[1], and while a number of applications that drive network traffic saw peak levels in 2021, at the height of the pandemic, there are more high traffic applications driving current network growth.
The EU has instituted a program that sets connectivity targets that must be reached by carriers by 20230, but EU telco’s, such as Orange (ORAN), Telefonica (TEF), and Deutsche Telekom (DTE.XE) have been writing letters and speaking to officials concerning the cost of building out network capacity to meet those goals.  It is not that they believe such goals are unrealistic, but more about the cost, which they believe should be shared by those that generate internet traffic.  This sounds like a typical cost shift from corporations to consumers, but that is not what the telcos want, they want the companies whose applications are generating the traffic, not consumers, to share the cost of the expansion.
As the data shows, in 2021, almost 57% of internet traffic was generated by applications relating to six very large technology/CE companies (Google (GOOG), Netflix (NFLX), Meta (FB), Microsoft (MSFT), Apple, Amazon (AMZN)) and while that declined to ~48% last year, most of that decline came from Google and Facebook, while all others, particularly Netflix, saw increases.  The EU telecoms are asking that these companies, who pay nothing to ‘piggyback’ their services across carrier hardware, pay a fee to help finance the mandated coverage expansion that the EU has mandated, under the theory that at least a portion of the traffic expansion that will be seen over the next few years, is a result of applications that benefit these companies, while they pay nothing for the use of the network, with carriers and consumers (price increases) baring the cost of the expansion.
We expect that the beginnings of such a proposal will be presented at Mobile World Congress, which begins on February 27th, with a ‘consultation’ between the EU, the US, and the UK, and the collection of opinions from stakeholders that will help to form the legal wording of the proposal, a process that takes about 3 months.  Once the bill is formalized, it is reviewed by all EU members and the EU parliament, with the hope that it can be finalized by the end of this year.  The bill is expected to offer a potential alternative to the original telecom proposal of passing the full cost of 5G and additional broadband coverage equipment on to those who benefit directly or indirectly.  With CE company contingent relying on the concept of net neutrality to quash the bill, the proposal will also have to gather support from those outside of the EU, as global support is essential for the proposal to work, a difficult task at best, but one that will allow carriers to expand broadband coverage, financed by those who benefit, without the public having to bear the cost.  


[1] Based on average monthly usage

​A Taiwanese technology daily, known for its somewhat misleading headlines and extreme forecasts (our opinion), has published the headline “Tablet shipments to slip below 30 million units in 1Q 23”, an ominous prediction that looks to strike fear into the hearts of those that participate in the CE space and set a negative tone for the 2023 year.  We find this a bit headline a bit misleading in that it gives little or no perspective as to what such a prediction actually means.
If we look at quarterly tablet shipments over the last six years, the trend is a bit difficult to see as tablet shipments tend to be volatile, and not always seasonal, but when viewed on a yearly basis, it seems more obvious that tablet shipments had flattened during the two years before the COVID-19 pandemic.  During the two years before the pandemic (2017 – 2018) the quarterly rate of tablet shipments was 35.4m units, while during the pandemic (2Q ’20 – 2Q ’22) the quarterly average was 41.1m units, 18.9% higher.  As the pandemic winds down, we would expect tablet shipments to return to pre-pandemic levels, which would imply a rate close to that 35.4m units/quarter that we mentioned above.  That said, it can be seen in Figure 1 that there have been a number of quarters when tablet shipments fell below 30m units, both before and during the pandemic, so while a prediction of ‘less than 30m units’ for 1Q is a bit troubling, it is not out of line with past history and is likely reflective of both the winding down of COVID, and the global inflationary macro environment.
Our bigger concern however is less about the level of tablet shipments in 1Q and more about whether tablet brands have reduced expectations with suppliers, particularly panel producers, and whether panel producers will maintain the low panel production levels needed to balance supply and demand.  After a poor 2022, panel producers will have a strong desire to push prices or production to generate some momentum early in the year.  Figure 3 points to the fact that in pre-pandemic years, tablet panel production was, for the most part, behind tablet shipments, which should drive panel prices higher, however this was not the case as tablet panel prices declined (Figure 4), more in keeping with the slow deterioration of demand through the end of 2019.
In 2020 tablet panel producers stepped up production, matching demand closely and tablet panel prices increased rapidly.  That said, we believe this was more of a function of strong demand, rather than a closer balance between tablet demand and tablet panel supply, and as the tablet demand peaked at the end of 2020, panel producers continued to oversupply, causing tablet panel prices to decline in 2Q ’21, along with other IT product prices.  During 2022, tablet panel production was closely matched to demand, yet panel prices still declined, more likely picking up the long-term declining trend that had been interrupted by the pandemic.  If panel producers take time to respond to the lower tablet shipment levels expected for 1Q (Figure 5 & Figure 6), tablet panel prices will decline further, rather than level off as they seem to be doing in recent months.

Ai is a bit far afield from our usual hunting ground, consumer electronics, but recent stories in the press about ChatGPT have sparked interest from some of our readers toward the topic, and without even stretching a little, we can easily draw the increasing connection AI has to consumer electronics.   Most consumers’ perception of AI is a large bank of servers, lights flashing, churning over billions of data points in order to arrive at a conclusion, but that is only one part of the AI story, and usually the part that has little to do with consumer electronics.  But some form of AI is built into many consumer electronics products and performs numerous functions, some of which we never see or hear, while others are commonplace.
Among the most common “AI” tools is the spell checker usually included in word processors, an outgrowth of Ralph Gorin’s 1971 SPELL program that checked letters in a word against a list of correctly spelled words by comparing single or adjacent letters, and while early spell checkers were limited to Mainframes, they migrated quickly to PCs in the early 1980’s.  These relatively simple systems were based on a comparisons, which would imply that the more words to compare against the more accurate the checker, but it was found that as the word lists grew over 100,000 the number of misspellings increased. 
The problem facing early spell checkers is that they were not able to discern context, so they faced difficulties with words that sounded the same but had different spellings and different use cases, such as ‘there’ and ‘their’.  As local computing power and storage increased, systems were designed that could evaluate context by comparing phrases or sentences against grammatical rules that were embedded in the system and could determine the correct spelling of a word based on both the word list and the rules, but data scientists found that as processing power and storage increased further, they could enter millions of data points that AI systems could ‘look at’ to better understand context, not only for language, but for almost anything, leading to the development of AI algorithms that fall into three classes.
Supervised – The most common category, where labeled data is feed to a system and is used to train the system to predict outcomes as it is given new data.  Typical systems uses decision trees that test the data using sophisticated ‘if…then’ nodes to move through the decision process until a conclusion is drawn that is based on all of the tagged data that the system has seen.  Systems like these are the basis for systems that have been in the press in recent months, such as those that can ‘paint’ a picture based on the style of a particular painter, having been fed the digital images of all of that artists work.  The system can create an image based on the user’s request using the characteristics it has recognized for that artist.  In its simplest form we asked an AI art generator to paint a picture of daisies in a vase using a generic style and Figure 1 is the result.
Unsupervised – This category does not use tagged data, so the algorithm must evaluate the relationships between the data points without specific definitions, which they tend to do by creating clusters of similar data around central data points, and using a number of methodologies to evaluated the clustered data.
Reinforcement – This category is a bit like training an animal, as the system uses an ‘agent’ to perform an action within a certain ‘environment’, and then rewards the agent when the action is completed.  The agent uses the reward (positive or negative) to help it with the next iteration of the action until the environment ends the process, just as rewarding an animal for a correct response helps the animal to understand what the correct response is.

​CE products vary widely in their use of AI systems, from virtual assistants such as Amazon’s (AMZN) Alexa  or Apple’s (AAPL) Siri to embedded Ai systems such as those in your TVs.  Virtual assistants ‘learn’ about your behavior based on the tasks they are asked to perform, so, in theory, they should be able to respond to requests with increasing accuracy as they learn an individual’s patterns.  TV AI is more technical in that it can sample incoming images for patterns and make adjustments on-the-fly to emphasize or de-emphasize a particular aspects of the image based on the user’s settings, or remove ‘noise’ or scale an image from one resolution to another.  These systems can process images on a pixel-by-pixel basis and with TVs generating ~30 frames each second, that comes to 14.9b pixels that can be processed each minute, but the AI component also ‘learns’ about each image, rather than comparing to a built-in list as in older sets, and can make better adjustments based on what it learns.
Both of the applications mentioned are common to CE products and in many cases are integral to their function, but AI continues to develop and will become more integral in CE products and daily life.  Having a background in the music industry, we were intrigued by Google’s (GOOG) MusicLM system, a bit less sophisticated than the artistic systems mentioned above, but nonetheless amazing in its ability to create music from text.  In its simplest application the words “acoustic guitar” or “flute” generate the following sound clips[1]:
 https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/instruments/acoustic-guitar.wav,
 https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/instruments/flute.wav
but more interesting is how it generates clips from the text “blues”, “West coast Hip-hop”, “East coast Hip-hop”, or “Reggae”
 https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/genres/blues.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/genres/west-coast-hip-hop.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/genres/east-coast-hip-hop.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/genres/reggae.wav
However things get even better when the system gets more complex text that does not specify a genre but a caption, such as “beach in the Caribbeans” or “escaping prison”,
https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/places/beach-in-the-caribbeans.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/10s_samples/places/escaping-prison.wav
with applications in TV, gaming and advertising quite obvious.  Not only is the system able to create genre related music or even contextual music, but it has also learned about diversity across all forms of music and with the same text token “progressive rock guitar solo” came up with a number of different guitar solos based on its learned data.
https://google-research.github.io/seanet/musiclm/examples/audio_samples/diversity-samples/progressive-rock-guitar-solo/same-text/0.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/diversity-samples/progressive-rock-guitar-solo/same-text/1.wav
https://google-research.github.io/seanet/musiclm/examples/audio_samples/diversity-samples/progressive-rock-guitar-solo/same-text/2.wav
 
 
While this is only a single AI application, there are many, including the famous/infamous ChatGPT and its many clones, that will find their way into CE applications regardless of whether we believe they are useful, helpful, legal, moral, or justified as they represent ways in which content can be created without the expenses associated with human involvement.  We have given up moralizing about the potential ‘disintegration of society’ that AI has the potential to cause, since almost all generations have heard the same thing about some technological change (TV, video games, CDs, etc.), and society has found a way to adapt to the change.  However, as only a small portion of humankind has the desire to challenge the validity of textual or visual media, it will take some time before we understand how AI will affect the fabric of human creativity.  Will it turn us into a society of eloi[2], allowing AI to keep us docile with a constant supply of curated entertainment, or will it free us from the burdens of everyday life, allowing our creativity to blossom?  Honestly, we don’t have a clue but to paraphrase Bette Davis “Fasten your seatbelts, it’s going to be a bumpy ride.”


[1] These sound clips reference the Google MusicLM site links for each file.  Ctl+Click should play each individual file.

[2] “The Time Machine” – H.G. Wells 

​Until recently Netflix (NFLX) has been inured to the idea of account sharing as a way to encourage the building of a massive subscriber base, but over the last year, while the company has made no overarching policy change, there is considerable evidence that there is a move to end the account sharing practice that so many subscribers partake in.  Netflix is quite open about its view of the practice, stating “A Netflix account is for people who live together in a single household” at the top of the “Sharing your Netflix Account” help page (https://help.netflix.com/en/node/123277), and goes further, stating “People who do not live in your household will need to use their own account to watch Netflix.”
 
Netflix tracks sign-in data for each account, so if a device signs in that is not associated with that account or is used often, Netflix may ask the account holder to verify that the device is legitimate before it allows access.  In order to verify a device, Netflix will send an e-mail link to the account holder with a short-term PIN.  That PIN has to be entered into the device requesting access to the Netflix account within 15 minutes or the device will not be allowed access.  Netflix indicates that ‘device verification may be required periodically’, although it does not specify whether that is for all devices, or just ones that do not have the same IP address as the account holder, but there are instances where Netflix will allow the user to watch from different locations.
 
If you have two houses, in theory, you need two Netflix accounts, as the IP will differ at each location, however Netflix will grant you the ability to watch from both locations, albeit with an occasional verification request.  Netflix uses the IP, account activity patterns, and device IDs, which are embedded in all mobile devices and many others and remain constant as long as the device does not undergo a factory reset, so there are instances when Netflix might question the validity of a particular device, but while the rules are specific as to ‘no sharing’ the policing is far more difficult to enforce and seems to spark considerable feedback from subscribers. 
 
Recent promotion of the ‘no sharing’ concept in South Korea led to a survey by the Korea Information & Communication Policy Institute late last year, which indicated that more than 40% of domestic users said they would cancel their paid subscriptions if forced to pay an additional fee for a 3rd party user, with only 2% indicating their willingness to pay the extra fee.  According to the data, only 42.8% of Netflix users in South Korea use accounts listed under their own names, with ~5m paid account subscribers listed at the end of  2021, which means another 7.9m users are using shared Netflix accounts, and if the statistics are correct, 2m paid and therefore an additional 3.14m subscribers would depart, bringing actual viewership (paid and unpaid) from 12.9m to 7.76m, a ~40% reduction.
 
While Netflix has not officially made any statements other than, “The time and fee for account sharing in Korea have not been formalized”, most industry observers believe the change will occur in March of this year and will likely be applied within 30 days of the announcement.  If it goes well and much of the survey is consumer bluster, Netflix will see incremental revenue.  If not, it could start a consumer revolt that would have imprecations across the entire streaming video industry, especially after a number of streaming services have recently increased monthly fees.

We had made a forecast for Chinese smartphone shipments for the 4th quarter and while the December numbers are not yet in, the official Chinese smartphone shipments for the last two months are 47.16m units, surprisingly 15.6% higher than our estimate, which pushes us to raise our December shipment number from 19.4m units to 22m units.  If this is correct (or close), that would put the full year at 265,65m units, down 24.3% y/y, and the 4th quarter up 16.4% q/q but still down 32.0% y/y, and above our most recent 252m unit estimate.  Most recently we have seen a full year tally of Chinese smartphone shipments totaling 287m units, which would imply a 44m unit shipment month in December, making it the best month this year, by 33%, so we are wanting to understand what the basis for that estimate was, given that our 11 month data is based on information provided by the Chinese government, which, if anything, leans a bit to the high side.
Over the last year 5G shipments in China have tracked fairly closely to overall shipments, representing between 85.1% and 72.2% of total shipments, although the spread narrowed considerably during 2H of 2022.  For the full year, using a proportional 16.97m 5G units in December, we estimate 5G shipments for the full 2022 year to be 207.6m units, or 78.1% of total shipments, down 22.0% y/y, the first year 5G shipments have declined since 2019 when China began releasing such data.  Given the current share, which we expect to expand a bit further this year, we expect 5G shipments in China to be a function of the overall health of the Chinese smartphone market, with far less dependency on the growth of 5G installations across the country than in previous years and more on China’s own macro environment. 
Strict COVID restrictions slowed 5G deployment progress across China last year, and with many of those restrictions being lifted, deployments will likely pick-up, but we are quite skeptical as to the coverage data that the Chinese government has been promoting, and if the data in the US is any example, it bears little practicality to consumers.  Typically many cities that are claimed as ‘covered’ have very spotty 5G service and while the coverage maps look well filled, the detail is far less so, with statistics used more for promotion and politics than actual consumer results.
Chinese brands continue to hold the major share of the Chinese domestic market (84.9%) and while it has deteriorated a bit over the last 4 years (86.5% last year), incursions, particularly by Apple (AAPL), have made little impact on overall domestic brand share in the Chinese smartphone market.  With Samsung Electronics (005930.KS) the only major smartphone brand with a very minor share in China, we expect relatively little change unless Samsung continues to expand and market its foldable line on the Mainland, which, while facing domestic competition, seems to be considered among the most prestigious smartphone devices with Chinese consumers, along with Apple.  However price sensitive buyers do have domestic foldable alternatives and there are still many pockets of nationalistic pride that would push marginal buyers to domestic brands, so for the 2023 year, we expect only a minor reduction in domestic share.
All in, it was a difficult year for CE in China, and smartphones in particular, but while we are careful to temper any enthusiasm against both a poor macro-economic situation on the Mainland, and a festering anti-China manufacturing posture for many global companies, we do take heart in that in recent months, while poor in terms of y/y performance, shipments have been, at the least, relatively stable, which we expect to continue into 1Q.  The Chinese smartphone market is maturing, and with that comes slower growth and more macro affectations, which certainly seems to be the case in recent months.  As that continues we expect the Chinese smartphone market to have basically the same competitive and economic characteristics of the global smartphone market and track more closely to the global markets each year.  That said, Chinese brands have the advantage of lower cost, but also face some political issues, such as those in India, that can offset those cost advantages.  It is up to both the Chinese government and Chinese smartphone brands as to how those issues will play out this year, and over the last year both did little to smooth those relationships that can help Chinese brands gain share outside of the Mainland.  Maybe 2023 will be different, although it is still hard to be overly optimistic.

​As we have noted, the EU will implement strict power consumption rules in March that would disqualify 8K Micro-LED and OED/LCD TVs from being sold in EU countries.  Under the new regulations the power consumption of 8K TV sets, as they currently stand, would not qualify under the regulations and could not be sold in EU countries.  While the power consumption of control and driving circuitry  for each sub-pixel in both 4K and 8K TVs is small, there are 24.88m sub-pixels in a 4K TV and 99.53m in an 8K TV, which makes even this basic power consumption math simple to understand, although TV set brands do little to help consumers understand TV set power consumption, with many stating only a ‘predicted amount of electricity needed to operate the product under average conditions for one year’, without specifying the number of hours.
Simple comparisons between 4K and 8K sets (same series, same size, same brand) indicate that total energy consumption for some sets can be as high as 3.3x for 8K, while others show a more reasonable 2.0 to 2.5x ratio.  Regardless of the numbers, the EU seems to be holding steadfast to the new rules and thus far has not even  granted a hearing where TV brands could plead their case.  That said, LG Display (LPL) has taken matters into their own hands and announced its 3rd generation OLED display panel, which combines Micro-Lens technology and a new ‘brightness enhancing algorithm’ with last year’s ‘Ex Technology’ which swaps the strong atomic bonds of Deuterium (aka Heavy Hydrogen) for those of Hydrogen in a number of materials in the OLED stack.  According to LGD, the combination will improve panel energy efficiency by 22% at the same brightness level.
If those specs prove out, LG Display OLED panels would be able to operate under the new EU requirements, allowing products made with such panels to be sold in EU countries.  While LGD’s 55”, 65” and 77” 4K panels will use the new technology, the most important application will be for the company’s 77” and 88” 8K panels and is expected to push the technology toward its 97” OLED panel in 2024.  The company indicated that mass production of the updated panels will begin some time this year, which we expect will begin in 1Q, at least on a limited basis, as each of the company’s OLED TV panel fabs will have to make production process changes and requalify designs.  That said, unless similar power efficiencies are found for LCD and Micro-LED 8K sets, OLED will rule the European 8K market for at least a portion of 2023, although with less than 400,000 8K units expected to have been sold in both eastern and western Europe last year, it is still a very small market that is struggling to grow..