Aixtron purpose of this thread
Demand for telecom base stations, converters, and charging stations is creating demand for components and devices powered by third-generation semiconductors GaN and SiC.
According to TrendForce, the GaN power market will undergo the highest magnitude of growth with revenues for 2021 to reach $83 million, an impressive 73 percent YoY increase. Going forward, annual GaN power devices revenue is expected to grow at a 78 percent CAGR and reach $850 million in 2025.
Consumer electronics, NEVs, and telecom/data centres, comprise the three largest sources of GaN power devices consumption, at 60 percent, 20 percent, and 15 percent, respectively.
TrendForce finds that about ten smartphone OEMs have released more than 18 models of smartphones equipped with fast charging capability, while notebook manufacturers are also indicating a willingness to adopt fast charging for notebook computers.
Annual SiC revenue, on the other hand, is expected to grow at a 38 percent CAGR and reach $3.39 billion in 2025, with NEVs, solar power generation/storage, and charging stations representing the top three largest source of SiC power device consumption, at 61 percent, 13 percent, and 9 percent, respectively. For the NEV industry, in particular, SiC power devices are most widely used in powertrain inverters, OBCs (on board chargers), and DC-DC converters.
IDMs from Europe, the US, and Japan control substrate supply
Due to their relative difficulty in epitaxial growth and the fact that the industry is moving from 6-inch towards 8-inch substrates as the mainstream, third-generation semiconductor GaN and SiC substrates are 5-20 times more expensive to manufacture compared to traditional 8-inch and 12-inch Si substrates.
Most substrate materials are, at the moment, controlled by major IDMs as US-based Cree and II-VI, Japan-based Rohm, and Europe-based STMicroelectronics. In response to this, certain Chinese suppliers, including SICC and Tankeblue, have successively entered the substrate market with the support of China's 14th five-year plan, with an aim to accelerate China's goal of semiconductor self-sufficiency.
Although substrate suppliers from Europe, the US, and Japan enjoy an early presence in the market and possess relatively mature process technologies, TrendForce believes that Taiwanese suppliers still hold certain competitive advantages.
For instance, not only do Taiwanese companies have vast experiences in silicon development, but Taiwan is also home to a comprehensive upstream/downstream silicon supply chain. In addition to these aforementioned advantages, Taiwan is further aided by policies that promote domestic material supply, design, and technological development. Taiwan could therefore achieve its goal of becoming a centre of advanced semiconductor fabrication that derives its strength from a gradually maturing front-end substrate and epitaxy industry chain, as well as mid- and back-end competencies in chip design, manufacturing, and packaging.
Currently, two major strategic alliances, led by Hermes-Epitek (with subsidiaries EPI and EPISIL), and SAS (with subsidiaries GW, AWSC, CWT, and ATC) are attempting to maximise their efforts in Taiwan's lacking substrate industry. Furthermore, TAISIC, jointly funded by KENMEC and TAINERGY, has submitted 4-inch SiC substrates for qualification and is actively investing in 6-inch SiC substrate R&D.
https://compoundsemiconductor.net/article/113584/...rowing_At_78_CAGR
Strong orders, demand not cooling off. I dont know the company but it seems that there is some cross-read to Aixtron etc given they talk about "next generation semiconductor devices".
Surely CWL knows better how to interpret?
ASM today also announces that the order intake in Q3 2021 is expected to exceed previous guidance.
Link:
https://www.asm.com/Pages/Press-releases/...s://www.asm.com/investors
Some interesting slides on Semi Capex spending, forecasts and applications in their IR presentation as well:
https://www.asm.com/Downloads/2021_Q2_Investor_presentation.pdf
Regards,
Fel
US memory player Micron reported last night with the sales guidance below expectations (guide 7.65bn vs consensus at 8.6bn) and stock -4% as a result.
However they raise capex spending into their next FY (i.e. 2022) to 11-12bn after spending 9.7bn in 2021, so a 15% increase at mid-point.
Unlikely that this is a direct read-x to Aixtron but I note that there have recently been concerns in the market that we are reaching the peak level in Capex for the semi industry. Aixtron (as capex customer) should hopefully be different given the shift to "next gen" semis, but nevertheless it is important to be aware of the market concerns for the sector overall.
For Aixtron, analysts currently expect 8.6% sales growth for 2022 to € 466m and 107m EBIT (23% Margin). On these estimates the stock currently trades on 20x EBIT which is in-line with its historic valuation (range of 17-23x). So in essence we have to hope for further strong orders/revenue growth in 2022 to drive the stock price further. Given that we stand at the early stage of GAN/SIC/Micro LED adoption with 3D sensing and datacenter/telco remaining stable contributors and 5G the wildcard, I would hope/think that top-line growth remains stronger than 8.6%.. but its hard to say!
Regards,
Fel
In terms of news from the industry, we also had the ASML investor day this afternoon, press release and presentations can be found here: https://www.asml.com/en/investors/investor-days/2021
Optionen
The full investment amount is actually 12B Yuan ($1.8B). To produce a total of 2.36m 4 inch wafers annually (750,000 GaAs, 1.61m GaN) would need about 100-150 G4 and G5 equivalent MOCVD. My estimate is basing on 3-4 process runs per day. The total MOCVD costs would be about $200m-300m.
While Aixtron's G5 does not have the regular blue LED market, it has the competitive edge for mini/micro blue LED manufacturing due to its uniformity and low defect films. Therefore it is also possible that Sanan would order some G5 for the blue mini/micro LED. Sanan most likely would order G4 for GaAs LED given that G4 is the tool of choice.
The Infineon CMd is a MUST WATCH event today for us. A special section on GAN and SIC. It starts in 6hrs and is available via the website. I will go through the presentation as it comes online.
Key highlight already: they raise their capex budget for next year by 50% to 2.4bn, from 1.6bn!
Given their ambition in SIC let’s hope a lot of that budget goes towards Aixtron ;)
Regards!
Fel
On GAN they quote "market forecasts" showing a market size from 47m USD in 2020 increasing to 801m in 2025, so 76% CAGR.
Focus applications are among others ( i cant read all): Chargers, 5G, wireless charging and servers/datacenters
Emerging applications in GAN: Solar, EV charging, white goods (e.g. washing machines etc).
So to me this sounds like broader applications than only fast charging which is often pointed to by Aixtron.
Will add links when the presentation is online.
Yesterday's analyst downgrade is probably rooted from that thinking. Sometimes people try to be smart, or too smart, to get ahead of the others. I think the analyst is being too smart. I have not read his report so I cannot judge what quantitative analysis and any detailed market survey he has done. If it is based on his gut feeling, or just fear, then I question the rationale.
There was a report out two days ago saying TSMC has 20 MOCVD in 2021 and planning to add another 10 in 2022. That is a 50% increase of TSMC's capacity. We know TSMC targets the GaN-on-Si mainly to supply NAVITAS and other players of course. If we use TSMC as a market gauge, then certainly it is not slowing down.
For any company that wants to gain in the Generation 3 compound semiconductors, CAPACITY is the king. If you don't have installed capacity to meet the demands coming, then forget about it. That applies to mini LED, VCSSELS, GaN, and SiC of course. Do you think they would pause at this stage?
Though not directly comparable, ASML just released a EUR 6.2bn booking number per 30 Sep (guidants.com mentions that consensus stood at EUR 4.2bn, not sure whether this was a typo).
"The demand continues to be high. The ongoing digital transformation and current chip shortage fuel the need to increase our capacity to meet the current and expected future demand for Memory and for all Logic nodes.
Source: https://finance.yahoo.com/news/asml-reports-5-2-billion-050000960.html
Optionen
https://tw.stock.yahoo.com/news/...D%E5%82%99%E6%88%B0-234733273.html
for a good start to the week, a little interview with CEO Grawert in German Magazine Wirtschaftswoche:
https://www.wiwo.de/erfolg/management/...-nichts-zu-tun/27723886.html
Regards,
Fel
Optionen
Optionen
strong Q3 results and guidance raise by ST Micro yesterday. The Auto end-market was obviously hit by production stoppages but they expect strong growth here in Q4 and overall remain very bullish in particular also on SIC.
The highlights:
- "Our SIC capacity is increasing quarter after quarter".
- Now have 85 SIC programs with 70 customers, 50-50 split into automotive and industrial. In Industrial a lot of demand relates to charging stations.
- They now expect $ 1bn in SIC revenues in 2024, one year earlier than originally anticipated as EV adoption goes faster than expected.
https://investors.st.com/static-files/...-1663-4862-a806-8712d0164c83
So Outlook for SIC is to ramp over 2022/23 in terms of equipment demand in my view. That aligns with Aixtrons view that latest in 2023 we should be expecting volume orders, in my view this could start in Q3-22 given lead times from orders to installed machine.
Regards!
Fel
Transcript: https://www.fool.com/earnings/call-transcripts/...21-earnings-call-t/
Optionen
Q3 results in da house tomorrow. I have no big conviction this time and wonder if for once it will a quieter day for Aixtron shares tomorrow instead of +/- 10% intraday swings..
My expectations are: Orders at ca. 105-110m, strong sales maybe at 130m (Q2 was 68m) (as always guided by the company, Q3 better vs. Q2, Q4 better vs. Q3) and strong EBIT at ca. 35m, 27% margin.
EBIT should benefit from a) strong volumes, b) no more APEVA restructuring costs (they incured 5-10 in H1-21) and c) the beneficial EUR/USD FX rate.
I do not expect a significant change to guidance, maybe some narrowing towards the upper end of the SALES and EBIT Guidance (420-460m sales and 20-22% EBIT margin is the current guidance), consensus currently at 430m and 21.3% margin (91.7m).. so if they narrow towards the upper end there could be 8% upside on EBIT if consensus assumes 460m and 21.5% margin. However with the stock already on 20x 2022 EV/EBIT I dont think this would trigger a major move beyond +3-5%.
In any case I think the comments about SIC/GAN and 2022 Outlook are way more important for the stock than sales/EBIT for 2021. So watch for that and Q3 order intake.
Consensus assumes 5% revenue growth for 2022. I think that has some upside potential as long as orders stay solid and management doesnt talk about slowing momentum in GAN and confirms that SIC should see more volume in 2022!
As excited as ever,
Fel
many thanks for your well balanced view on tomorrow’s results – much appreciated! I tend to agree with your assessment, but feel a bit more comfortable going into the earnings. For the first time in two years it seems that we do not go into earnings with lofty expectations, i.e.: i) share price is some 20% below this year’s peak, underperforming the market in recent weeks, ii) Aixtron’s management has well flagged that 2Q21 order entry will not be repeated, iii) Oddo downgrade had its impact (on expectations) as well. As you mentioned, I do hope that there will be a positive narrowing in terms of revenues/EBIT guidance – there are only 8 weeks left for this year and the EUR/USD FX guidance rate of 1.25 compares to ~1.18 average ytd EUR/USD rate (spot <1.16).
I also do have the hope that the “headline” growth numbers – maybe revenues +100% and EBIT +300% vs. 3Q20 – might be well perceived by the market. As you indicated, a quieter trading day would be very much welcomed, nobody needs the usual quarterly results drama…
Optionen
Optionen
Auguste Philip Richard -- Northland Capital Markets -- Research Division-MD & Senior Research Analyst
Got it. And then on the compound semi, I think you've mentioned power, RF and photonics. Now I was just wondering if you could give a little bit of color on the MOCVD side, how much of your business these days is photonics and how much of it is power? And how much of it is RF?
William John Miller -- CEO & Director
Sure. Let me try to take that apart a little bit. I would say, Gus, our MOCVD business is really operating at historically low revenues today post exit of their commoditized LED business. We did a lot of work restructuring the business and pivoting toward new markets and new opportunities. And as you know, we go after the gallium nitride applications with our propel single-wafer reactor, and that can be used for both power electronics, RF devices as well as disruptive silicon-based micro LED applications. And we have a second product for arsenide phosphide called Lumina for applications in photonics, such as indium phosphide lasers, VCSELs as well as red micro LEDS. I would say we are starting to gain tractions in MOCVD. I would say in the near term, we expect to see revenue growth in early stage micro LED next year. And from a number of opportunities, including AR/VR as well as other photonics applications. Also, as I mentioned, we did see the 8-inch GaN-on-Silicon power market with multiple customers and placed an eval for 8-inch power electronics at a large foundry. We also, as I mentioned, placed an eval tool for red micro LED. So to answer your question, I would say, this year, we did have a fair amount of power electronics business with a single-wafer reactor at 8-inch. But I would say we're not seeing most of the production today is on six inch. So I would say we're in a bit of a seeding waiting game for that market to develop. So as I look forward, I would say, our business is largely more predominantly going to be in micro LED applications, whether that's disruptive or traditional red micro LED going forward. So I don't know if John, if you have any more color you want to add.
https://www.fool.com/earnings/call-transcripts/...21-earnings-call-t/
Auguste Philip Richard -- Northland Capital Markets -- Research Division-MD & Senior Research Analyst
Right, right. No, I get it. And then just -- this probably isn't a fair question, but any color on who is ahead the arsenide phosphide on gallium nitride or the micro LED on silicon, the red on silicon, which program looks like it's advancing more quickly.
William John Miller -- CEO & Director
I would say, Gus, that we're actually -- it looks like, at this time, it looks like both are proceeding. We are seeing activity in the traditional -- the traditional approach of red, green and blue pixels for a luxury TV type markets. It looks like that's starting to happen. And we're seeing AR/VR applications with nontraditional kind of GaN-on-Silicon type solution. So right now, I would say it's fairly hard to handicap where we are. But I would say it looks like they're both moving forward.
Auguste Philip Richard -- Northland Capital Markets -- Research Division-MD & Senior Research Analyst
And I'll -- my last question. So it sounds like it's more application-specific in terms of which [Indecipherable] is going to be used.
William John Miller -- CEO & Director
Today, I would say that seems to be the case. I'd love to see kind of how it all plays out overall.
Order intake some 7m above your estimate, EBIT margin in line with your estimate (but some 2-3ppts better than eg the DB estimate).
I don‘t fully get why they haven‘t narrowed the revenue guidance…there are only some 8 weeks left and service revenues shouldn‘t be that volatile - why keeping the EUR 40m spread?
Optionen
Optionen
just briefly as the Conf call just finished. Overall a very solid call, clearly guiding towards the upper end of the order range, so Q4 should be around minimum 100, I think rather 110m (Q3: 114), with management ruling out the lower end (would imply 65m in Q4) - no idea why they did not narrow the range. Sales on track in Q4, margin will be good as well.
- NO DECELERATION in GAN - is the key message from the call. It started with consumer electronics (fast charging) but now the applications are clearly broadening out and adoption rising, that means customers adding capacity. They see this as a stable trend also into 2022. New applications are telco base stations, datacenters (sounded more advanced) but also automotive starting, e.g. BMW announced using GAN in onboard chargers. That is rare as usually Auto takes long in adoption.
- All other end-markets (Laser, datacenter/datacom, ROY LED) remain on solid levels and are expected to stay that way.
- SIC. Very strong and much clearer comments vs. 9 months ago. Expect volume orders in H2-22 and "significant volume" in 2023. Customers are placing orders now, takes 6-9 months of delivery (so deliver in Q2) and 2 months of installation. That means customer can work with tool and then decide for orders in Q3/Q4-22. Very credible framing in my view.
- Micro LED: Are in collaboration with other players. Volume ramp might happen from 2023 onwards.
- Question: "Can we expect growth in 2022?" We have a strong backlog that will drive H1 sales and expect order intake to continue at this level. "That means 22 will be a strong year again".
Other highlights:
- no issues in the supply chain
- SIC market growing rapidly.
- 15% tax rate only.
- 330m in cash on the balance sheet.
- confirm FY Guidance. Expect Q4 sales above Q3 level.
My view: 2022 should see 10% sales growth and a bit of margin expansion (no APEVA restructuring costs, new equipment with higher price/gross margin), so maybe 115m in EBIT vs. current consensus on 105m. Stock trades on 20x EBIT on the 105m consensus which is in-line with history, so not expensive. I think the stock can trade sideways a bit but clearly it should not drop towards 16-17.. the investment case is fully in tact and will gain more speed as SIC and Micro LED ramp.
Cheers!
Fel