​The SEC has indicated that it will enforce the “Holding Foreign Companies Accountable” act that was signed by former President Trump last December, and is adding a few amendments and opening up the additions for comments by interested parties.  The original bill amended Sarbanes-Oxley, requiring the SEC to prohibit the securities of foreign companies from being traded on US securities markets if the company uses a foreign accounting firm that cannot be inspected or investigated by the Public Company Accounting Oversight Board for 3 consecutive years (2021 is year one).
The Act itself has a number of key provisions as follows:
If the SEC determines that a public company has three consecutive “non-inspection years,” beginning in 2021, the SEC would prohibit the company’s securities from being traded on a U.S. national securities exchange or an “over-the-counter” market subject to SEC regulations.
If, following the delisting of a company for noncompliance, the company certifies to the SEC that it has retained a registered public accounting firm that the PCAOB has inspected to the satisfaction of the SEC, the Act directs the SEC to end the prohibition on trading on U.S. securities markets.
If non-inspection recurs after the SEC ends a prohibition, the SEC will prohibit the company’s securities from being traded on a U.S. national securities exchange for at least five years. Such prohibition could only be removed if, after the end of the five-year period, the company certifies that it will retain an accounting firm that the PCAOB is able to inspect.
If the PCAOB can’t inspect the foreign accounting firm’s work, then the company will be required to submit to the SEC documentation certifying that the company is not owned or controlled by a governmental entity in the foreign jurisdiction in which its registered public accounting firm is located.
In addition, for each year that the PCAOB is unable to inspect a company’s accounting firm, the company will be required to disclose in a form filed with the SEC, among other things, the percentage of the company’s shares owned by governmental entities and its relationship with the Chinese Communist Party.
The amendments add:
Consistent with the HFCA Act, the amendments require the submission of documentation to the Commission establishing that such a registrant is not owned or controlled by a governmental entity in that foreign jurisdiction and also require disclosure in a foreign issuer’s annual report regarding the audit arrangements of, and governmental influence on, such registrants.
If the Commission identifies a ‘covered issuer’ that has retained registered public accounting firm to issue an audit report where the registered public accounting firm has a branch or office that is located in a foreign jurisdiction and the PCAOB is unable to inspect or investigate completely because of a position taken by an authority in the foreign jurisdiction.  Registrants so identified (“Commission-Identified Issuers”) are required to submit documentation to the Commission that establishes that they are not owned or controlled by a governmental entity in that foreign jurisdiction. In addition, the registrant is required to submit documentation to the Commission that establishes that they are not owned or controlled by a governmental entity in that foreign jurisdiction, and includes a requirement for disclosure of state influence and the names of Chinese Communist Party members on company boards.
All in, the Act and the more detailed amendments are an indication that the tension between the Chinese government and the US government under the Biden Administration is not going to ease, as domestic Chinese regulations ban businesses from giving foreign regulators access to accounting documents without government approval.  While the new rules will likely not cause a mass move for Chinese companies already listed on US exchanges, it will encourage dual listing on other exchanges such as Hong Kong, although valuations in Hong Kong tend to be lower than in the US, and while Chinese exchanges such as STAR are able to tap into the Chinese investor base, especially for young companies, they don’t have global access like US markets.  Luckily there is time for the US and China to discuss ways to keep these issues from becoming more onerous but similar talks have been ongoing for years.  While the odds are such disclosures would not reveal much information that was not already assumed or implied…  “It isn’t the original scandal that gets people in the most trouble – It’s the attempted cover-up” – Tom Petri (R).
 
 

​Opto-Tech, (2340.TT) formerly one of Taiwan’s leading LED producers, has changed a bit over the last few years with a shift from its traditional LED production to a focus on silicon based sensors, now accounting for between 60% and 70% of revenue, while LED products have been reduced to between 20% and 30%, although last year the company was the largest provider of pulse oximetry (green-light LEDs) sensors for the Apple Watch.  Opto-tech is also becoming certified as a VCSEL sensor supplier and will begin shipments in 2H, which is a plus, but that is not why we note the company.. 
The reason we mention the company is that they have just put in place a price increase of between 10% and 30% across all product lines, a result of rising material costs, which will affect all orders received in 2021, going back to January.  In an attempt to soften the blow, the company states that they will not raise prices further in the 2nd quarter, but went no further than that about additional price increases later in the year.
So far this year a number of lighting companies have raised prices for all sorts of reasons.  One company cited increasing steel prices as a motivation for raising lighting pole prices by 5%, a lighting fixture company citing raw material shortages for a 7.5% across-the-board increase, another with a 3% increase on LED lamps and drivers along with an 8% increase on fixtures, poles, and controls, and the list goes on with increases from between 3% and 10% from a wide variety of commercial lighting companies, most of which came in mid-February.  We believe few of these increases reflect LED prices and are primarily based on raw materials, so we would expect another round of increases as the silicon side of lighting sees higher prices.

​Samsung Display (pvt) is a primary OLED display provider to Apple (AAPL) and has dedicated much of their A3 fab to the production of displays particular to Apple’s needs.  While there are few CE brands that have enough demand to create such deals, Apple is certainly one and has made similar deals with others, particularly Japan Display (6740.JP), with whom they ‘pre-paid’ over $1b to help finance an LCD fab that was to be exclusively for production of Apple displays.  That said, the deal with Samsung Display is more of a contractual obligation, which requires Apple to purchase a certain number (or value) of panels over the course of a contractual year.  As Samsung Display has committed production resources to Apple under the yearly contract, there seems to be a ‘take-or-pay’ clause that requires Apple to meet its estimated demand goals or make up the difference as a payment.  While Samsung Display is not the only possible small panel OLED vendor Apple could use (Apple also uses LG Display (LPL) and BOE (200725.CH) to a lesser degree), they are the industry leader in flexible OLED displays and tend to be the highest volume supplier to all smartphone brands.
Given this leadership role in flexible OLED displays, and the resource commitment SDC makes to Apple each year, there is logic behind the take-or-pay situation, as SDC could be using those resources for other customers, so the onus falls on Apple to make accurate estimates of its anticipated production targets, usually made in 4Q of the previous year.  This is a tricky situation for Apple, who at times has faced display shortages when a particular iPhone model turns out to be more popular than expected, however the opposite has also been the case in 2019 and 2020 when Apple has had to pay almost $700m and over $900m, according to estimates, to SDC for ordering less volume than the company estimated, and it seems that Apple is already off to a similar start this year.
As we have previously noted, the iPhone Mini has been a disappointment in terms of unit volumes and has been in part responsible for a 20% reduction in OLED display orders for the iPhone 12 series in 1H.  As the sole supplier of displays for the Mini, Samsung Display saw a 9% decrease in OLED panel shipments in January, while combined OLED/LCD shipments were only down 4% during the month for the industry.  While we don’t know how granular the contract requirements are between SDC and Apple, without a very popular iPhone 13 series offering, it would seem that Apple will again have to pony up this year.  As we all can attest to, demand estimates are moving targets and it doesn’t take much to make them look like fever dreams or wishful thinking.  Lost revenue from being unable to provide customers with product is both a financial and psychological issue for a brand, but at the same time smartphone brands have a tendency to overestimate future demand and Apple does not seem to be immune from the problem.

​LTPO (Low-temperature polycrystalline oxide) is the new kid in town.  The backplane technology combines more common LTPS (Low-temperature Poly-Silicon) and IGZO (Indium Gallium Zinc Oxide or ‘Oxide’) to create a backplane that drives the display using IGZO and switches using LTPS.  This allows for a power savings of ~15% over LTPS alone.  While we say new kid in town, Apple (AAPL) introduced the technology in 2014 in the Apple Watch Series 4, driving a larger OLED display than in previous models.  Samsung Display (pvt) has been developing LTPO for Apple, who is expected to use it in the high-end model of the iPhone 13 out later this year, and has been converting part of its A3 fab (aka ‘the Apple fab’) to the technology since last year.
That said, SDC has recently announced that it is supplying LTPO OLED displays to Chinese smartphone brands, Oppo (pvt) and OnePlus (pvt), calling the LTPO driven displays ‘adaptive’, meaning they can change the refresh rate of the display from 120 Hz for high quality video or 30Hz for text, on the fly.  This is key to the power saving that the technology is able to offer, which Oppo says will be reduced by 46%, although that would be quite dependent on the content be displayed.  Samsung (005930.KS) itself used the technology in its Galaxy Note 20 Ultra, but not on other Note flagship models.  SDC is said to have already built out 30,000 sheets/month of LTPO production capacity at A3 and is expected to to have a total of 70,000 sheets/month by the end of 2Q.  We note that LTPO reduces line capacity relative to LTPS by between 25% and 33% due to additional process steps.
 

The CE space and more particularly the display space are cyclical, and after we noted the data from panel price increases yesterday, we checked to see how previous display price cycles compared to the one currently underway.  The one problem associated with tracking display panels over an extended time is they change.  They have increased resolution, obviously get bigger, and have other lesser known characteristics that make them different over time.  As these variations and new models are priced according to the market, they each carry a different price and are therefore not comparable.  That said, when looking at TV panels, there is one that has remained consistent since the beginning of 2012, which gives us just a bit over nine years of price data with no inconsistencies.
The panel we are speaking about is a 32” LCD TV panel, one we highlighted in yesterday’s data.  We broke down the price trends shown in Fig. 1 into ‘cycles’, some of which are long, and some, primarily newer ones, are ‘mini-cycles’ of short duration.  The table below delineates those cycles, showing a peak and base price for the panel for each cycle, when and how long the cycle was, and the price difference between the peak and the base of each cycle.  We also added a final column that shows how much of a price gain was achieved between the base of the previous cycle and the next peak, indicating how much of a ‘price recovery’ took place for each cycle.
While looking at the chart (Fig. 1) gives an impression of the cyclical nature of TV panel pricing, the table is more specific.  The average duration from each cyclical peak to its base is 304 days, although the cycles have been getting progressively shorter (Fig. 2), and the average duration from the previous cycle base to the next peak is 208 days, with little change over time, although we note that the base to peak for cycle 7, which we are in currently, if we assume today is the peak, took a long time to develop, which we believe accounts for some of its intensity.  That said, the base-to-peak of cycle 1, which had the same base-to-next-peak duration, saw a price change of 24.7% whereas the price change in cycle 7 (current) is already 113.9%, already 4 times the price increase of its closest looking cycle.
While this is really a statistical look at the pricing cycle of 32” LCD TV panels, it does give some indication as to how different this cycle (cycle 7) is from others.  It will likely be longer than any other cycle when measuring time from previous base to peak (when that occurs), and certainly has been far more intense from base to peak, than any other cycle.  We note that there are a number of exogenous factors at play in cycle 7, the largest being the effects of COVID-19 on demand, and to a lesser degree capacity, but we also have to note that much of LCD TV panel capacity is being produced by Chinese fabs who have a different mindset and profitability criteria than more traditional South Korean, Taiwanese, and Japanese producers.
The subsidies offered by local and provincial Chinese governments can allow Chinese panel producers to operate at lower margins and take higher inventory risk than might have been the case in the past.  This mindset also color’s Chinese TV set brand aggressive sales goals and panel buying plans, which are less colored by the devastating and long-lasting effects of early cycles, which has made more experienced panel producers a bit more conservative.  China’s share heading into the first cycle was under 10% and now exerts far more influence on panel capacity and panel prices, which we expect to expand further in 2022, as Korean producers reduce LCD large panel capacity.
Looking at the cyclicality of the panel space, recognizing that 32” TV panels are but one part of overall panel pricing, we can see that current panel pricing presents a high risk situation to panel producers, while also providing considerable profit incentive.  This makes the temptation to keep increasing panel prices all the more attractive, despite its eventual effect on the elasticity of demand.  Panel producers are not known for their restraint when it comes to panel prices and we can’t fault them for taking advantage of the unbalanced market currently existing, however component shortages can limit production even if panel prices continue to rise, which could both cap out panel producer profitability and push consumer level pricing higher, a combination that would destabilize the industry.
Taking it one step further, the profitability of panel producers currently has also given them the incentive to expand capacity, particularly Chinese panel producers, and while this has some effect on slowing panel price upward momentum, it will have a bigger effect once the industry passes its cycle 7 peak and prices head down.  Panel producers that have built new capacity will be incentivized to offer lower panel prices to fill fabs that are now being depreciated, and the effect on the slope of the back end of cycle 7 could be more intense, just as the front side of the cycle has been.  We expect little change in panel producer behavior however, until there is obvious evidence that cycle 7 has reached its peak, and a month or two of ‘disbelief’, but the trend line in Fig. 1 will eventually point the way.

​With daily trade press commentary on shortages, particularly silicon based shortages, it is refreshing to hear a company president state that they are seeing an easing on what had been component shortages and a reduction in lead times.  In this case, and it is certainly isolated, the company is Alpha Networks (3380.TT), based in Taiwan, a producer of switches and routers for both data centers and smaller networks.  The company was not specific as to which components had been in short supply, but did mention that production lead times for ICs specific to their products had seen lead times down to between 20 to 30 weeks from a peak of 46 to 52 peaks, and expects further lead time reductions in 2H.
While this is certainly a one-off, Alpha says it has now been able to fill some customer orders that had been deferred earlier and has begun a more robust deployment of its high capacity switch products which will start in 2H.  Again the components were not specified and could be very specific to the company’s products, but in light of the constant trade press mention of component shortages, we felt obligated to mention it.

​A fire at a Renenas (6723.JP) semiconductor fab yesterday has set in motion another round of concerns over the availability of automotive related silicon products.  The Renenas line that was affected was in Naka, Japan, and while it is early for estimates, we have already seen expectations for anywhere from a month to 6 months to fully restart the 300mm line that operates at the 28nm node.  Renesas also has a 200mm line that was not affected by the fire.  ~2/3 of the 300mm line’s production was for automotive products which are already in short supply, causing some manufacturing slowdowns.  Inventory levels will begin to be affected in about 30 days as existing parts and worked down.
Given that smoke from the fire filled the clean room, we expect much of the production that was on the line will have to be scrapped and while the Japanese government has offered to help Renesas to ‘quickly acquire alternative manufacturing equipment’, a nice idea if the equipment was available and not pledged to other suppliers who are trying to expand capacity.  We note that the Renesas plant was shut down for three days last month after an earthquake, which reduced available inventory to make up for lost production, increasing the impact of the fire, and further raising concerns that automotive suppliers will face additional shortages.

According to a report from China, BOE (200725.CH) is expected to be supplying Samsung Electronics with flexible OLED displays.  The Samsung M series is a very low-priced line intended for the Indian market, with prices ranging from $460 (M62) down to $96 (M02s) and was recently announced and released this month, however the only one that we believe is based on a OLED display is the M62, the highest priced model, with the others based on LCD.  The report indicates that BOE will be providing flexible OLED displays for the series, which would seem to imply the M62 model, the only one able to absorb the higher cost of an OLED display.
That said, it would be an unusual situation for Samsung, the parent of Samsung Display (pvt), to bypass its own subsidiary in favor of a foreign competitor, unless BOE is offering a highly discounted price to garner favor with Samsung, which the report intimates.  Logic holds that if BOE were to undercut SDC’s potential price for the flexible 6.7” OLED display, it would be in Samsung’s best interests to take the offer.  Of course the article applauds BO’s quality as the basis for the implied contract, but also notes that BOE gave Samsung a ‘shocking’ offer.  If true, this would be the first time a supplier other than SDC has provided flexible OLED displays to Samsung and would certainly help BOE fill the gap left by the reduced orders from Huawei )pvt).
​

​In early December we noted that there seemed to be problems with the deal between Samsung (005930.KS) Electronics and Wonik (240810.KS), who last August had signed an MOU for $69.4m to purchase SEMES (pvt), the largest semiconductor and display equipment supplier in South Korea.  This was after a deal with KCTech (281820.KS), another Korean semiconductor and display equipment supplier fell through earlier last year.  At the time, the price did not seem to be finalized and questions about compensation for the 300 SEMES employees, which would have become Wonik’s burden after the sale.    At the time SEMES employees were earning ~$52,600 on average, while Wonik employees earned $39,600.  Also in contention were receivables from Chinese display companies, which Wonik was willing to accept as part of the updated price ($75.5m) price it seems to be willing to pay.  That price increased to $82m as negotiations progressed.
Wonik issued a statement on Friday indicating that SEMES called off the talks as the potential agreement did not meet the demands of the existing employees who would have been transferred to the new owner.  Given that SEMES’s main customer is Samsung Display, who has stated their intent to end LCD large panel production, the bulk of SEMES’s business, which consists of coating, etching, and stripping tools, would have to shift to other customers, other than the inkjet segment, which Samsung was to retain.  Not only has Samsung (SEMES) failed to agree on terms with the two mentioned above, but another company headed by an ex-SEMES employee also hosted discussions about a purchase but to no avail.
This leaves Samsung to deal with the money-losing operations at SEMES, which would have been primarily OLED related if the deal(s) had been completed, now the costs associated with SEMES and the higher employee salary rates will continue to be absorbed by Samsung Electronics.  At some point if another buyer is not found, we expect Samsung will take action on its own and rationalize the business as originally planned, which could make the SEMES employee’s decision over compensation look a bit like short-term thinking rather than taking the entire picture into consideration.  For the time being, SDC is still producing large panel LCDs, but at a reduced rate, so we expect SEMES has a bit of time to find another buyer, but once SDC pulls the plug on its remaining large panel LCD production, SEMES will look even less attractive to potential buyers than it does now.

In early January we noted the rapid increases in prices of some materials used in the CE space, in particular copper, which is used in hundreds of CE applications especially copper laminated PCB boards which are of primary importance to almost every CE device.  At the time we noted that the price of copper had been rising, especially since the end of last year, and had reached a price of $3.60, the highest point seen in 5 years.  Since then, the price of copper has continued to rise, now reaching $4.11, up 14.16% since 1/15/21 and closing in on the 10 year high of $4.40, adding to the BOM of electronics across the board. That said, of the metals noted in the 1/15/21 note, only one has seen a 90 day price decline, and only two have seen a 30 day price decline, while all of the others have seen price increases.
While most of the price increases seen in these CE oriented metals have been under 5% over the last 30 days, there has been one that has been unusually strong to say the least, and that is iridium, the metal typically used as an organic bonding material in phosphorescent OLED emitters that are used in many CE devices.  The price of iridium has increased by $36.4% over the last 30 days and 135.3% over the last 90 days and has just reached $6,000/ounce due to supply disruptions in South Africa, which produces 80% to 85% of the less than 10 tins that is mined each year.  Because the amounts are so small, mining does not track demand and remains relatively static, which gives the metal its sensitivity to production issues and rapid changes in demand.
Increases in these and other basic materials for semiconductors and display materials are going to filter through the supply chain and force price increases above what might be the result of current semiconductor foundry shortages and display capacity.  We expect demand to soften later in the year if COVID-19 is brought further under control, but it will take some time for high-cost stockpiles and contracts to be worked down.