2023-10-25 10:40:24 ET
Summary
- ASML Holding N.V.'s Q3 earnings have been misunderstood by many investors and the market.
- The stock has wiped out most of its gains for the year, significantly underperforming the indices.
- Everyone tends to look at the same financial figures when assessing earnings, but these are not very relevant for ASML. We explain why.
- We discuss the three most important topics: exposure to China, the demand landscape, and guidance.
Introduction
ASML Holding N.V. ( ASML ) reported its Q3 earnings last week. There have been many discussions between investors after these earnings, especially with the company’s guidance for 2024. We’ll give our point of view later in this article.
The market did not like the company’s earnings much, with the stock dropping significantly the following day:
YCharts
If we zoom out, we can see that ASML’s stock has wiped out almost all of its gains this year . The stock was up significantly at some point, but it has suffered a significant drop in the past months. It is now only up 6% year to date, significantly underperforming the indices:
YCharts
We have taken advantage of this drop and added to our position after a year without adding a single dollar. This demonstrates that, if we are patient, the market will ultimately give us opportunities to add to great companies at fair prices.
The perverse incentives of the financial system “force” many market participants to aim to outperform every year, leading to excessive trading.
Without further ado, let’s take a look at the earnings.
The ASML Q3 numbers
The headline numbers
ASML reported good headline numbers for Q3. The company beat on the bottom line but slightly missed on the top line:
Seeking Alpha
Note that these estimates are in dollars and the company reports in euros, meaning there might be misses/beats resulting from currency exchange and not the underlying performance. Analysts are not better than you or me in predicting these variables.
ASML’s top line always tends to be a bit more volatile due to the business’ nature, whereas the bottom line is more in management’s control and is where the company tends to shine against estimates.
If you have read any of our prior public articles on ASML , you should know that we are not very keen on commenting on the company’s quarterly numbers due to the volatility and the long lead cycles. We don’t think the financial metrics are the most important part of this quarter’s earnings either, so we'll focus our energy elsewhere. This quarter’s highlight or lowlight (one can see it both ways) was the guidance and the demand landscape, so we’ll spend a bit more time speaking about those in this article. We’ll also comment on the company’s exposure to China.
Breaking down net sales
Like in the last few quarters, ASML’s sales growth was in net system sales. In fact, installed based management sales were down slightly year-over-year, from €1.5 billion to €1.4 billion:
Made by Best Anchor Stocks
Note that if we look at the year-over-year growth rates, these seem pretty impressive. ASML grew its total revenue by 16%, with a 23% increase in net system sales and a 6% drop in installed based management sales.
Looking at revenue numbers is always good, but they give us little information in ASML’s case. The company still has a substantial order backlog and continues to work through it, so revenue results from orders that came in around 12-18 months ago.
Revenue can be a good indicator of demand for any company that sells products on the spot. This is not true for ASML, where revenue is a lagging indicator . To this, we have to add the change in revenue recognition the company announced last quarter around fast shipments. As we said earlier, the most important thing for ASML right now is the demand landscape, and revenue does not do a good job as a proxy of this environment.
One thing that did get some kind of backlash was the revenue distribution . There were not many changes concerning technology and end-use compared to last quarter, but there was a significant change in distribution by region, which started last quarter. China made up 46% of ASML’s Q3 2023 revenue:
ASML Investor Presentation
This rise in revenue from China worried many investors, considering the geopolitical battle between the U.S. and China on the technology front. The question to answer here should be: is that revenue at risk? Well, to answer that question, we need some context.
Management has mentioned for a while that fill rates for Chinese customers had been depressed for some time (under 50%), and now they are simply seeing a catch-up as demand in other regions falls slightly:
For system shipments this year to Chinese customers, the majority of the orders were booked in 2022 . The demand fill rate for our Chinese customers over the last two years was significantly less than 50%.
So, the Chinese customers were in fact receiving a much lower number of systems than they ordered. This was due to the fact that the demand for our systems worldwide significantly exceeded supply.
With current shifts in demand timing from other customers, we now have the opportunity to fulfill these orders to our Chinese customers. So supply is in fact catching up to demand, and we're shipping lithography systems for mature and mid-critical nodes to China, while of course complying with export control regulations.
Source: Peter Wennink, ASML’s CEO, during the Q3 2023 earnings call (emphasis added).
We understand why some people would see this as worrying, but we don’t see it as such, considering most of these systems are being used for trailing-edge technology. Can the U.S. try to block China from less sophisticated systems? Obviously, they could, but we don’t think this is very likely. With these export controls, the U.S. ultimately wants to block China from using leading-edge technology for military purposes. Is this achievable by blocking leading-edge technology? Yes. Is this achievable by blocking trailing-edge technology? It seems unlikely. China can manufacture trailing edge chips for military systems even if the U.S. were to block ASML's systems today.
We must acknowledge that there is always a possibility, though. Even then, we would not be too worried about the long-term implications for ASML. Management has clearly explained that China is playing catch-up while the rest of the world is “resting” a bit from a period of heightened demand, meaning that China should not make such a large proportion of revenue in the future. In fact, management also claimed that the percentage of the backlog corresponding to orders from China has not changed much and still stands around 20%.
Many people view the geopolitical battle as one-sided for ASML, but it’s a double-edged sword . While it’s true that geopolitics pose a risk to short-term revenue based on export restrictions to China, it does also pose a significant long-term tailwind. The end-users of these leading-edge chips are well aware of the risk of disruption and are asking foundries to reduce the geographic dependability with Taiwan and China. Governments worldwide also want to manufacture leading-edge chips insider their borders.
Both of these trends incentivize foundries to build fabs elsewhere, and these fabs will require significant Capex spending on ASML’s systems, even if these companies are not as productive as in Taiwan. There’s no doubt in our mind that leading-edge chips will be built in the U.S.; at what cost is a different question, but the answer impacts the foundries (and, therefore, end customers) significantly more than ASML. Management believes that the geographical mix of revenue might change going forward but that the demand landscape should not.
Last quarter we discussed with subscribers the implications of this shift from Chinese players to the trailing edge for Texas Instruments (TXN). With the recent news of a 7nm chip that Huawei has manufactured using DUV multi-patterning, we think the risks might be overblown. Some of the DUV systems that ASML sells into the country might be being used for leading-edge nodes.
Of course, the question here is: at what cost? Foundries end up trading up to EUV because it becomes more economically feasible than DUV multi-patterning . We think this news might have spooked the US a bit, so the government has decided to apply more export restrictions. According to ASML’s management, this will not have a significant impact on the company’s sales because it’s pretty localized:
So the way we read the rules now, and of course you can imagine that, that part of the regulation, we've read pretty carefully, that the principle is that in -- that in principle, also the 1980s would fall under the export control restrictions, but only when those immersion tools are used for advanced semiconductor manufacturing .
And those advanced semiconductor manufacturing, we've been informed, only applies to a handful of fabs. Yeah. So that means that the 1980, for that -- for those handful of fabs, is off limits, but not for the vast majority of our Chinese customers, for which we don't need an export control license either. We can just ship.
Source: Peter Wennink, ASML’s CEO, during the Q3 2023 earnings call (emphasis added).
We don’t think the U.S. has any way of knowing what those DUV systems are being used for, but we have to consider that companies such as Taiwan Semiconductor (TSM) or Samsung Electronics (SSNLF), which the U.S. government trusts, also have fabs in China. Export controls don’t seem to apply to these customers. Not many companies worldwide can manufacture leading-edge chips, and if all of these are “friends” with the U.S. government, then the risk is somewhat diminished.
As for technology , just a couple of comments. The EUV weakness we saw last quarter continued this period. DUV and China continue to lead the way, and EUV keeps getting pushed out due to fab readiness . We are not worried about this either. High-NA EUV and fab readiness will be two significant growth drivers through 2025, so we should see EUV revenue as a proportion of total sales increase significantly then. EUV makes a significant portion of the company's current backlog.
The demand landscape: bookings and management’s comments
This is, by far, the most important section of this article, together with the guidance section. As discussed above, ASML has a huge order backlog, so revenue is pretty much booked and not a good proxy of demand . The best proxy of demand is the company’s net booking number and the backlog size.
ASML’s Q3 2023 net bookings came in at €2.6 billion , significantly below last quarter. These were primarily comprised of DUV, with EUV net bookings around the €0.5 billion mark:
Made by Best Anchor Stocks
We can conclude two things from the chart above. First, it’s pretty clear that the cycle is now impacting orders . Net bookings have gone from a high of €8.9 billion to a low of €2.6 billion in four quarters. ASML operates in a cyclical industry, and while its huge order backlog allows its financials to weather tough times, the company can see these cycles in its backlog.
Note that the backlog is only a result of demand exceeding supply for some time now. If, and when, ASML manages to couple its supply to demand, we should see the backlog slowly disappear, and the financials become increasingly exposed to the cycle. This doesn’t worry us much as long as the secular long-term trend is intact. In fact we welcome some kind of cyclicality. We have been increasingly looking at great companies in semi-cyclical industries with long-term tailwinds behind their back. The rationale is that these companies tend to give more opportunities because there will always be times when nobody will be willing to hold them.
The second thing we can see from this net booking number is that ASML continued eating into its order book. This is the case because recognized revenue was ahead of net bookings. Nothing worrying here, either. It’s obvious the company’s order book is now smaller, but it’s still at a very healthy €35 billion , or equivalent to more than 1 year of revenue:
Made by Best Anchor Stocks
What we think many fail to realize here is that a semiconductor company has a backlog covering one year’s worth of revenue while we are close to the bottom of the cycle:
The industry seems to be passing through the cycle trough . There has been some improvement in end market inventory levels downstream, although inventory levels upstream remain elevated. As a result, our customers continue to moderate wafer output by running at lower utilization levels.
While lithography tool utilization are still running at levels lower than normal relative to last quarter, tool utilization in Logic continues to show signs of improvement, while Memory has yet to turn. We concur with our customers that still expect to see an inflection point, indicating the start of a recovery by the end of the year, although the shape and slope of the recovery remains uncertain.
Source: Peter Wennink, ASML’s CEO, during the Q3 2023 earnings call (emphasis added).
We’ll discuss the implications of this forecast in the guidance section later on.
Breaking down profitability
There is not much news on the profitability front compared to last quarter. ASML’s gross margin continued to improve sequentially while operating margin remained somewhat stable:
ASML Investor Presentation
Higher than expected gross margin resulted from the DUV product mix and some one-off cost effects. The good news is that management will not cut on important investments in 2024 despite a muted growth environment:
So obviously in the current environment, we are frugal, right, as you might expect us to do. So we're definitely controlling our SG&A expense there. On the R&D side, that's long-term, and I think we would be ill-advised to now go cut our R&D roadmap, and that's not what we're doing, right? So we are continuing to execute on the R&D roadmap.
Source: Roger Dassen, ASML’s CFO, during the Q3 2023 earnings call (emphasis added).
This is great to see. Despite the head start the company has, it should continue innovating. Recall that ASML creates a virtuous circle through its technology. More advanced technology drives increased technological adoption and new use cases, which in turn require more chips. This virtuous cycle has been discussed by management many times.
Cash flows - Continued weakness as expected
ASML’s cash flow weakened significantly year-over-year, but there’s nothing that should take us by surprise here:
ASML Investor Presentation
Management mentioned last quarter that customers are being more diligent with their cash flows in the current environment , which obviously impacts ASML. Note that when demand was significantly above supply, many customers prepaid amounts to secure the supply of these systems. Now that the demand-supply imbalance is not as evident, these customers are saving this cash to weather the storm.
We don’t think this is really worrying because ASML still generates a significant amount of cash and has a very solid financial position. Management continues to return excess cash to shareholders despite their willingness to hold more cash during such period:
ASML Investor Presentation
Guidance - Where the confusion lies
With the demand landscape, guidance might be the most important part of this quarter’s release. This is what ASML shared with investors:
ASML Investor Presentation
Up to here, nothing worrying, right? Management confirmed its 30% growth expectations for the year with a slight gross margin expansion . So, what spooked the market? Management said the following during the Q3 2023 call (emphasis added):
Therefore, based on our current view, we expect the revenue next year to be similar to 2023 . As such, we see 2024 as a transition year…
But hold on, something is missing here…
…but also as an important year to prepare for the significant growth that we expect in 2025.
So, ASML’s management expects a flat year in 2024 and significant growth in 2025. The market was unhappy because many (ourselves included) expected 2024 to be a growth year. So, what’s going on? Management argued that they are preparing for the worst-case scenario in 2024, where the cycle recovers but does so slowly. The 2025 growth expectations are based on several reasons:
Now, based on discussions with our customers, we currently expect 2025 to be a strong year, driven by a number of factors. First, the secular growth drivers in the semiconductor end markets, which we have previously discussed, such as energy transition, electrification, and AI. The expanding application space, along with increasing lithography on future technology nodes drives demand for both advanced and mature nodes.
Secondly, the industry expects to be in the middle of a cyclical upturn in 2025, starting in 2024. And lastly, as mentioned earlier, we need to prepare for the significant number of new fabs that are being built across the globe. These fabs are spread geographically, are strategic for our customers, and are scheduled to take our tools. It is essential that we keep our focus on the future and build capacity to be ready for this ramp.
Source: Peter Wennink, ASML’s CEO, during the Q3 2023 earnings call (emphasis added).
In short, ASML expects 2025 to be the meeting point between a recovery cycle and new fabs , which should drive net system sales significantly. 2024 growth did not disappear; it just seems to have been pushed out to 2025.
We saw many people confused with the timing here. In fact, the market also seemed quite confused with it. The day after ASML reported earnings, TSMC held its earnings call and mentioned they also expected the cycle to start turning in a couple of months. This is consistent with ASML’s message, and it makes sense because ASML gets these messages from its customers, TSMC being the largest one.
However, there was one striking difference, but not really, if you understand how this works. While ASML does not expect growth in 2024, TSMC does expect it to be a growth year . This difference got many people confused, but the message is still consistent. To enjoy growth, TSMC can simply adjust quickly to the growing demand by increasing the utilization of its existing tools. ASML can’t do this due to its long lead times. When customers are confident that the recovery is underway, they will most likely place orders, and ASML will be able to fulfill them around 2025, consistent with management’s message of a strong 2025.
Some investors expect ASML’s stock to be dead in 2024 due to its no-growth guidance, but we honestly believe it will move with orders, not revenue. Once the recovery is underway and orders start to flow, the market will look forward and will believe management’s 2025 guidance. Never forget that the market is forward-looking. We’ll know beforehand if 2025 expectations are on track :
But indeed, you are right. If 2025 is the kind of strong year that we expect, then indeed you would have to see the order recovery in the first half of 2024, absolutely.
Source: Peter Wennink, ASML’s CEO, during the Q3 2023 earnings call (emphasis added).
It was nevertheless funny seeing ASML drop the day it reported earnings and rise when TSMC did, considering the messages given in both earnings calls were strikingly similar!
All in all, we are not worried at all about ASML’s no-growth guidance for 2024 . This is a cyclical industry, and we knew it would eventually show up in the company’s numbers. The company has been the last to suffer the downcycle and will probably be the last to recover from it due to its long lead times. The good news is that management has all of 2024 to prepare the company for the significant growth ahead, which will inevitably come.
We would also be cautious about taking management’s guidance at face value regarding the cycle's timing. If the last quarters have been an indication of anything, it is that management is not accurate in predicting the cycle despite all the information they have. This should make you even more cautious when you see a random person trying to forecast it. Why not focus on the long-term trend? It is less hassle and probably a more profitable thing to do over a long period.
Conclusion
We hope this article helped you understand ASML’s quarter, especially what we consider the most important topics:
-
China exposure
-
Demand landscape
-
Long term expectations.
ASML always has some sort of cloud on top of it, be it geopolitics or the cycle. While this might sound bad, it’s actually good for long-term investors who know that these clouds are temporary and will be eventually dissipated by the long-term secular trend.
In the meantime, keep growing!
For further details see:
Important Takeaways From ASML's Q3