​While Murata (6981.JP) will report its full year results for its March year tomorrow, the President of Murata, the largest producer of MLCCs (Multi-layer Ceramic Capacitors), has indicated a fear that the US-Sino trade tensions are ‘distorting’ the parts market for smartphones.  While he indicated that with Hauwei unable to take orders and produce for US customers, other brands, particularly the top three brands, have been increasing orders significantly.  “China produces a large number of 5G smartphones.  Compared with the demand of the real economy, there are signs of overheating of orders.”
This is coming from a company that is expected to see a potentially large increase in operating profit against earlier forecasts, and while analysts in Japan have mixed views as to whether current MLCC demand is real or over stated, the caution expressed seems to indicate that the company is operating with the understanding that things could change quickly if demand slows.  Japanese company managements tend to be far more conservative (at least publicly) than Taiwanese or Chinese management, so we take some of this with a grain of salt, but it is not the first time we have heard a company executive express some reservation about current component order status…

We have mentioned in previous notes that Samsung Electronics (005930.KS) has taken a different path this year in regard to its Galaxy S flagship smartphone line.  While much press focus has been on features, 5G, and the potential end to the companion Galaxy Note line, we focus more on price, with Samsung lowering the price of this year’s model, the Galaxy S21 Ultra 5G by $200, or 14.3%, making it a bit more appealing to those that might be under a bit more financial pressure as a result of the economic impact of the global COVID-19 pandemic. 
Performance increases generated by the Samsung Exynos 2100 processor, updated from the S21’s Exynos 990, and the Qualcomm (QCOM) Snapdragon 888, updated from the previous Snapdragon 865, particularly the integrates 5G modem are a substantial cost savings (we note that Samsung uses Qualcomm as the key chipset for the US and China, while it uses its own Exynos chipset for its international versions).  While we are not focusing on features in this note, we do note that both the Exynos and Qualcomm processors in the Galaxy S21 Ultra 5G have also been updated, with both having moved from 7nm nodes to 5nm nodes and both produced using EUV.  This has improves processor performance, so there is no question as to whether Samsung has made a performance sacrifice to gain a cost reduction (see table below for performance comparisons).
The Galaxy S21 Ultra 5G was designed specifically to reduce costs with the single chip orientation leading the way, while reducing the number of 5G antennae and eliminating the charger and wired earphones from the smartphone package, offset against increases coming from the inclusion of UWB (ultra-wide band) and the S-Pen.  That said, the international edition, which incorporates the in-house Samsung Exynos chipset, puts the percentage of cost borne by the Samsung ecosystem at 63%, giving the company even more ability to keep costs low.
With a BOM of ~$533 compared to last year’s $572, Samsung has reduced the BOM by ~7% and the selling price by ~14%, which has driven early sales of the Galaxy S21 line up by 22% in the line’s first month of sales on a y/y basis.  All in, it seems that Samsung has taken the right direction in terms of maintaining its flagship Galaxy S line, and with the inclusion of the S-Pen in the S line, obviates the need for the Note series.  While we expect the Samsung foldables will not replace the Note on a unit volume basis, we expect there is a good chance that by 2022 the foldable line will produce similar revenue, however with the S-pen now included in the top of the Galaxy S line a direct comparison to the Note becomes a bit more difficult. 
 

​While we tend to filter statements made by Chinese companies, given a political predilection to aggrandize data that is positive toward the plans of the government, Huawei has certainly been a company that was singled out by the Trump administration as one that represented all that the administration felt was inherently wrong with China and its supposedly one-sided relationship with the US.  Much has been said about the validity of claims made against Huawei, but other than a few ‘official denials’ much of the response to those claims from Huawei never reached US news sources.   While the US government continues to use Huawei as a rallying cry for almost anything that needs a bit of aggressive sentiment behind it, there is still little proof as to the claims made by the government.
Huawei itself posted what it calls 9 false accusations made by the US government and how it responds to each.  Rather than filter through the commentary, we post it in its entirety for review.
1.      Huawei provides the cheapest equipment because it is funded by the Chinese government
 
Huawei pointed out in the statement that any MBA student knows that cheap does not mean a competitive advantage. Telecommunications operators will not buy the cheapest telecommunications equipment without considering quality, safety and other important standards. They want value for money, and Huawei has done this.
 
One reason is that Huawei has invested a lot of money in research and development. For example, R&D expenditure in 2018 reached 14.3 billion U.S. dollars, more than that of Apple, Intel and Cisco, and 30% higher than the combined R&D expenditure of Huawei’s two largest competitors in the field of network equipment, Ericsson and Nokia. Therefore, Huawei's network equipment is usually smaller and easier to transport and install. At the same time, it is more energy-efficient and lower maintenance costs. The price of the equipment is therefore reasonable.
 
As for Huawei's funds, it mainly comes from the company's reinvestment income. Huawei also borrows from international banks, and the government subsidy funds are relatively small. In 2018, the amount of government subsidies was only equivalent to two thousandths of Huawei's total revenue.
 
2.      Huawei's 5G equipment is not advanced, and 5G patents lack "relevance and value"
 
According to data from intellectual property consulting firm CPA Global2, Huawei has the largest number of 5G patents and 5G core patent families. Former U.S. President Trump announced the U.S. 5G plan in April 2019, but Huawei's research began ten years ago. Up to now, Huawei is the only company in the world that integrates the production of 5G mobile phones, 5G base stations, 5G optical fibers, and 5G core network hardware and software. Huawei technology is considered to be 12 to 18 months ahead of other competitors.
 
3.      Huawei steals intellectual property
 
Huawei is the world's largest manufacturer of telecommunications equipment, with 500 telecommunications operator customers and a cooperative relationship for more than ten years. Huawei does not need to steal the intellectual property rights of others, because it owns a lot of intellectual property rights. Last year, Huawei applied for more than 5,000 patents to the World Intellectual Property Organization. Huawei is the largest holder of 5G technology patents, accounting for about 20% of all 5G patents.
 
At the same time, since 2001, Huawei has paid more than US$6 billion in intellectual property licensing fees to third parties, 80% of which have been paid to US companies.
 
4.      Huawei does not agree with Western values
 
In the digital realm, the value is reflected in security and privacy. Huawei attaches importance to these factors as much as US citizens, and may even attach more importance to it than the US federal government. In terms of security, Huawei customers have never encountered major cybersecurity breaches, and there is no evidence that Huawei has been threatened by the Chinese government or other levels. On the contrary, the United States is used to changing digital product data to gather intelligence. Huawei mentioned that the United States signed the US CLOUD Act (Clarification of the Lawful Use of Foreign Data Act) to make it easier for law enforcement agencies to access server data information, regardless of the country where the data is stored.
 
5.      It will be very easy and fast to remove Huawei network equipment
 
The US State Department stated that the cost of replacing Huawei's equipment in Europe is expected to be only "3.5 billion US dollars." But according to the Daily Mail, the UK’s ban on Huawei’s supply of 5G equipment would cost more than 1 billion pounds (1.3 billion US dollars). An upcoming study by Oxford Economics estimates that the economic impact of banning 5G networks on the UK, Germany, and France may be as high as US$11.8 billion, US$13.8 billion and US$15.6 billion, respectively. These amounts have not taken into account the costs that other European countries will incur.
 
In the United States, if Huawei’s network equipment is dismantled, it will force about 40 small wireless operators in remote rural areas to scrap equipment worth millions of dollars. Some of these small operators said that the cost of dismantling Huawei’s equipment could lead to bankruptcy. At the same time, American families, schools, hospitals, farms and even small businesses may lose affordable Internet services.
Jeff Johnston, an economist at Union Bank, which provides loans to American farmers, estimates that replacing the existing Huawei equipment in the United States will cost rural operators about $1 billion and will take three to seven years.
 
6.      U.S. activities against Huawei have nothing to do with the Sino-U.S. trade war
 
Former US President Trump stated in May 2019 that if China and the United States reach an agreement, Huawei may be part of it.
 
7.      It is easy for Huawei to install backdoors in its products
 
Huawei is the most censored telecom equipment manufacturer in the world. Independent institutions in the UK, Germany and Belgium continue to test Huawei products. Since its establishment in 1987, Huawei's customers have never encountered major cybersecurity breaches. Because of this, customers trust us, and they trust our products. Telecom operators scan their networks 24 hours a day to find anomalies. Any behavior that deviates from the normal pattern will cause alertness sooner or later.
 
8.      Huawei has an unusually close relationship with the Chinese government, and the identity of its owner is "strictly confidential"
 
Huawei is a private company that is 100% owned by its employees. Our largest individual shareholder, CEO and company founder Ren Zhengfei owns a little more than 1% of the company's shares. Another 99% of the shares are held by the Huawei labor union, which is entirely owned by Huawei employees.
 
9.      Blocking Huawei will not slow down the pace of global 5G deployment
 
Huawei pointed out that it will be telecom operators that will launch 5G networks, not the US State Department. Operators in the UK, Germany and other countries and regions have made it clear that the blockade of Huawei will delay the launch of 5G networks.
 
Although there are other 5G equipment suppliers, Huawei's technology is more advanced. In 2019, T-Mobile postponed 5G network business. The American Telephone and Telegraph Company (AT&T) announced that its so-called 5G network speed is lower than 200Mbps, while South Korea’s LG U+ deploys 5G with Huawei technology to provide download speeds exceeding 1.3Gbps.
 
According to Oxford Economics, preventing Huawei from building 5G infrastructure may cause a country to increase its investment costs in 5G by 8% to 29% in the next 10 years. This may delay millions of people from enjoying 5G services, leading to a slowdown in technological innovation and slowing economic growth.
 

​Corning (GLW) reported 1Q results of $3.3b and EPS of $0.67, with core (non-GAAP) EPS of $0.45, ahead of the $3.1b and $0.43 (non-GAAP) consensus estimates.  While the company saw strength in sales and net income (y/y) from all divisions, the company saw an increase in freight and logistics costs that reduced profitability by $50m which they believe will be reduced in 2Q and beyond.  The display segment saw a 3% increase in sales (q/q) for 1Q, which has seen negative 1Q q/q results since 2015, so increased utilization at large panel producers has certainly had a positive effect on that segment.  However, references to y/y growth for display (up 15%), are against the onset of COVID-19 last year and have less relevance.  Specialty Materials, the segment that produces Corning’s cover glass products, saw a 17% q/q decline in sales (up 28% y/y), which was a bit better than typical 1Q (q/q) performance over the last few years.
We believe that the tightness of the glass market, in part due to capacity events at competitors, and Corning’s previous comments concerning a glass price increase, might have driven perceptions above reality, and while Corning is in a very advantageous position in terms of demand and pricing, they do face some of the same issues relating to cost and shortages that others inI the industry also face.  Some are direct and some are indirect, but we expect that some investors have assumed that Corning has stayed outside of such issues and will remain so.  While we reiterate that much of Corning’s display related business is seeing positive conditions, that enthusiasm should be tempered by the possibility of component shortages putting caps on display space performance.  Corning has the edge in large panel TV substrate supply, but if panel producers cannot produce due to driver or other component shortages in 2Q, it could mitigate some of that advantage.  It’s hard not to be enthusiastic when the metrics are positive, but “The enlightened ruler is headful, and the good general is full of caution.” -  Sun Tzu  

​In a letter to employees, Huawei (pvt) founder Ren Zhengfei, made references to the potential for a public offering in the future, something the founder has denied in the past.  Huawei has been in a difficult position, a result of US trade sanctions, and has seen some of its business segment shrink rapidly.  The letter to employees was specific about the effect the creation of false accounts might have on one of the company’s businesses entering the capital markets, stating that falsifying accounts would move from a disciplinary issue to a legal one, with dismissal the immediate result.
The letter preceded Huawei’s full year financial report, which saw sales growth of 3.8%, the slowest in 10 years, with negative growth in the company’s overseas markets, and indicated that the company will encourage talented employees to take positions outside of the Mainland over the next 3 to 5 years to bolster its sagging international prospects, and made a number of references to its aggressive employees, hundreds of whom volunteered to be demoted when the US sanctions were first introduced.  “To go to the battlefield, to shoot, to be injured, is always our priority” was the mantra in the letter, reiterating his philosophy, but his references to a potential IPO seem to take more weight outside of the company.

Perhaps the title of this note is not completely accurate, as there were really no losers in the panel space in March.  The large panel segment grew 11.9% m/m in terms of revenue and 8.5% in terms of shipments, while large panel ASPs grew 3.1%.  On a y/y basis, revenue grew 56.6%, with shipments up 26.7% and ASP’s up 23.6%.  While the y/y growth was impressive in March, the industry is still comparing against a period of declining panel prices near the onset of COVID-19, so we would expect y/y comparisons to remain unusually strong for the first half, while becoming progressively less abnormal as the year progresses

On a regional basis Taiwan saw the largest increase m/m at 20.9%, while Japan saw the largest increase on a quarterly basis, but Fig. 2 points out that Japan’s share of the LCD large panel market is miniscule compared to other regions.  Fig. 2 also points out the negative March y/y and 1Q q/q and y/y comparisons for South Korean large panel producers, as both Samsung Display (pvt) and LG Display (LPL) have decreased their large panel capacity through plant sales or closings

On a company basis, the table below looks at those panel suppliers that are primarily large panel producers, as a number of small panel producers also produce small quantities of large panels.  This can make large panel performance for these types of producers very volatile, given that small changes in such numbers can lead to large percentage changes, but little impact on the industry as a whole.  Companies we have excluded from the table below are:
Hannstar (6116.TT)
InfoVision (688055.CH)
Japan Display (6740.JP)
Tianma (000050.CH)
 
Given the relatively unusual circumstances of last year, most of the y/y comparisons don’t give much new information, and the negative results from Samsung Display and Panda have more to do with the sale of LCD fabs than actual performance, with BOE and Chinastar being the receivers of those fabs.  That said, Sharp stood out with its best monthly revenue (Feb./Mar.) since mid-2014, and Chinese large panel producers HKC and CHOT saw their best quarters since inception.  While capacity increases are embedded in some of the more positive monthly and quarterly results, higher utilization and large panel price increases would be the root for much of the gains 

Large panel shipments did not see growth in 1Q, at least on a q/q basis (-1.8%), however looking at the 1Q 5 year q/q average for large panel shipment, which is -9.7%, the relatively small drop in large panel shipments in 1Q is unusually strong.  Subsequent q/q growth will compare against higher averages as noted in the table below.  Shortages in the display space have become more of a problem for large panel producers as the year has been progressing, and with relatively little foundry capacity increases (drivers, TCONs), and tight supply of glass and raw material price increases, we expect large panel price increases will be the only way large panel producers will see increasing revenue on a monthly or quarterly basis.  There will be some capacity increase at Gen 7 and above fabs (+4.1%), but the bulk of any improvement will come from further panel price increases, against progressively stronger 2020 results, particularly in 2H, and higher 5 year averages.

In our 2/2/21 note we mentioned that Tianma had capped its new small panel OLED fab in Xiamen, 50 days ahead of schedule, which, the project manager said was the case, despite ‘a number of both physical and material based obstacles.  As it turns out, there seems to be a bit of hyperbole in that statement as the image below shows that while part of the roof has been constructed, general terminology in the fab construction space would mean the fab is ‘tight’ and is not subject to outside environmental forces, which it is not in the current image.  According to the latest information, the roof decking has been partially completed and the full cap is expected to be completed at the end of this month, with equipment move in beginning in July.
While terminology might be at fault here, there was a part of the fab that was capped in February, both workshops and dormitories, but not the main fab, so according to our model the project remains on a normal schedule under which we model small scale production beginning in September of 2022 and reaching full capacity (30,000 sheets/month) in 2023.  We expect a phase 2 project to begin at the end of 1Q ’23, albeit a smaller line consisting of 18,000 sheets/month.

​We have noted many times that the psychology behind component buyers is a complex one.  The delicate balance between not having enough and having too much is what gray’s the hair on buyer’s heads, but aside from the COVID-19 pandemic, which has increased demand for certain CE items, is there something else that has changed?
Fig. 5 indicates that China has been increasing its semiconductor imports at a steady pace over the last year (actually longer) and March set a new semiconductor import record.  More telling are the 1Q comparisons, with 2021 seeing a 29.8% increase in semiconductor imports in China vs. last year’s 10.5%, so there is little question that China has been more aggressive than usual when it comes to semiconductor imports, but we do not believe that the underlying force that is driving this behavior is COVID-19 demand, but more of a mindset change brought about by the trade restrictions placed on Chinese companies by the previous US administration, particularly those placed on Huawei (pvt), which have had a devastating effect on companies that have found themselves unable to produce products that were routinely sourced in earlier years.
Over the last two years, we believe such trade restrictions have changed the way Chinese CE managements look at inventory.  When sourcing was unrestricted, many Chinese CE companies operated under JIT inventory rules, maintaining low inventory levels to keep costs low and avoid end-of-year write-offs.  Once it became apparent that the US trade restrictions would have a real effect on sourcing, Chinese buyers became more aggressive and began building inventory in components that were specified in the US restrictions.  After early January, when it became apparent that the new administration was in no rush to change the government’s stance on such restrictions, Chinese CE companies and OEM/ODMs stepped up ordering to build inventories further, unbalancing the supply chain even further than it already was.
While this might seem a rather simplistic explanation as to why semiconductor shortages seem to get further out of control this year, we have seen almost every level of component, from raw materials and pre-cursors, passive components, PCBs, and semiconductors of many types, rapidly escalate in price, which we believe is based to a degree on increased demand, but more so by a need to increase component inventory levels by Chinese CE companies, who might have been less aggressive in previous years..  Whether this is a ‘new’ psychology that remains a part of the CE mindset in China, or whether such buyers return to JIT sourcing if demand slows is unanswerable, but our real concern is how quickly such a change might be effected. 
We expect the initial change would be slow with orders that have been pushed out by months  being brought back to shorter lead times, but if demand stays weak, it would seem logical that some of those orders would be ‘postponed’ and CE companies would work down existing inventory.  All in, the CE space faces a higher than normal risk level once semiconductor demand slows or turns negative, with a similar situation facing the display space.  We have mentioned a high wire act before as the poster child for CE in 2021 and that seems even more so as the year progresses.  JOHO.
 

There is a semiconductor shortage, but until recently much of the expansion talk by large foundries such as Taiwan Semiconductor (TSM), Samsung (005930.KS), and UMC (UMC) was concerning major investments in 7nm, 5nm and 3nm nodes.  EUV equipment for such fabs has been in short supply and much ado about trade restriction concerning Chinese semiconductor fab purchases for these nodes dominated the press.  That seems to have changed recently, as semiconductor shortages continue to grow, but not nearly as aggressively at 7nm and 5nm as in more mature nodes, where the bulk of semiconductor production lies.
TSM recently indicated that it would be spending ~$2.88b to expand its 12” 28nm fab 16 line in Nanjing, which makes logical sense in terms of filling current and future demand, but has been maligned in the Chinese press for using “Mainland policies, resources, hydropower, and low-cost talent”, and further “dumping at low prices in the Mainland, and squeezing out China’s emerging chip manufacturing companies.”  Much of this comes from public opinion rather than industry, which removes those who would understand why a global semiconductor shortage is devastating to all industrialized economies and leaves those who see the question from a very different perspective, but it is unusual to see China engaging in a debate about protecting one of its fledgling industry’s against foreign incursion. 
China’s largest semiconductor foundry SMIC (688981.CH) has already approved a plan to build a new 28nm fab in Shenzhen that will produce 40,000 wpm when completed and existing SMIC fabs at or above 28nm are building out additional capacity, which we expect is the root of the anger toward TSM’s expansion, especially since SMIC is not considered to be at the same technical level as TSM, despite both working at 28nm.  China is expected to be able to produce ~1m wpm this year (all nodes combined), which keeps the impact of the TSM expansion, which will not be completed until 2023 to a minimum.
The dumping issue is more one of misunderstanding how the semiconductor industry works, as chip level prices decline as less mature nodes come on line and mature fabs reach fully depreciated status., and since the same shortages face Chinese companies, TSM’s expansion in China will help to alleviate that position for many Chinese ODM’s who are currently vying for off-Mainland capacity against other large ODMs.  Should the Chinese government to try to limit TSM’s 28nm expansion by not offering the same financial and physical resources that it offers to its own semiconductor companies, would be a bit like cutting of one’s nose to spite one’s face.  In the long run it will hurt, but defending China’s economy from semiconductor ‘dumping’ is a bit ironic after years of accusations of China’s product dumping on other economies.

On Wednesday we noted that Apple had announced a new 12.9” iPad Pro model using mini-LED backlighting.  With expected demand for the new product, which is expected to be released in May, of ~5m units, Apple has set up a supply chain to produce the newly configured backlights.  This supply chain is also involved in Apple’s mini/micro-LED development project in Taiwan, which is based on mini-LED chips produced by Ennostar (3714.TT) (formerly Epistar), and involves a number of Taiwanese and South Korean companies.  Once the chips are produced at Epistar, they are sent to both TSMT (6278.TT) and Yenrich (pvt) a subsidiary of Ennostar, where they are mounted directly on to PCB boards via SMT (surface –mounting technology), rather than more typical thru-hole mounting.
Given the size of the mini-LEDs relative to more standard size LEDs, surface mounting is the logical technology, especially given the number of LEDs, which in this case is over 10,000.  That said, moving and mounting such small LEDs is a daunting task that carries with it inherently lower yields, and the removal and replacement of those mini-LEDs damaged in the process is a chore that adds to both the cost and the average production time.  It has been said that TSMT has been plagued with low yields that have been a cause for potential short supply or extended delivery times for the iPad Pro, and reports that much of the SMT work has been shifted to Yenrich until the TSMT yields can be improved.
The flexible PCB that is the substrate for the mini-LEDs is produced by South Korean supplier Young Poong Electronics (000670.KS) and LG Display (LPL) and Sharp (6753.JP) are supplying the display, with GIS (6456.TT) supplying touch components and Heesung Electronics (pvt) assembling the modules.  While there are bound to be glitches in the new supply chain, we expect it will have little overall difference on how successful the 12.9” iPad Pro is in the long run, and we expect initial reviews to be positive, which we expect will be the case based on both the M1 chip and the new display.  All in, we would expect most iPad Pro users that are ready to replace older machines, will wait an extra month or so if necessary.