The Same Story Every Quarter

ASML's Q1 2025

ASML reported earnings earlier this week in a not-so-benevolent market environment (neither broadly speaking nor specifically speaking about semis). Just like it has happened to the company since god knows when, the market put a lot of weight into the quarterly numbers. If you’ve followed my work for a while, you’ll know that I don’t fixate on ASML’s quarterly numbers for a myriad of reasons.

The main reason is that they are very volatile and misleading (and getting increasingly more so). As ASML transitions to EUV, the ASP (Average Selling Price) of its products is increasing significantly and is making its financials more unpredictable and lumpy. You could ask, why? The reason is that quarterly financials and KPIs are getting increasingly exposed to the recognition of orders and revenue of certain EUV units (which retail for upward of €200 million). That shouldn’t be a problem if it were not for the fact that the timing of this recognition is unpredictable (both for the market and management) and can have a significant impact on short-term numbers.

Let me explain this with a hypothetical scenario. Say ASML receives two low-NA EUV orders (ASP of €230 million) on April 1st rather than on March 31st. This would mean that orders for Q1 are roughly €460 million lower than they would’ve been just due to a one-day difference in the timing of said orders. This is definitely an extreme example, but it serves to prove that quarterly orders are not necessarily a great leading indicator of future demand. I believe history is a pretty good guide here. Just over the last 6 quarters, ASML has experienced two quarters of significantly-above-average orders, and several that did not meet the expectations of the market:

I don’t think the market is good at forecasting order trends because…(a) analysts are flying blind, and (b) timing can mess up the numbers. I don’t blame them for getting these wrong.

These dynamics are also apparent on the revenue side, although less so due to “higher” predictability. If ASML sends an EUV system to a fab but the customer accepts this system on April 3rd rather than on March 20th (for example), then revenue is deferred to Q2. While quarterly numbers are significantly impacted by timing and have low predictive power, there’s no denying they tend to impact the company’s stock significantly (no wonder management wants to stop reporting some KPIs).

Now that I am on this topic, let’s talk a bit more about orders. Net bookings came in at €3.9 billion (€1.2 billion EUV), significantly lower than the market’s €4.5 billion expectation. This is evidently not great, but some people made the mistake of taking it as a warning sign that ASML might fall short of its 2025 guidance (which was reiterated, by the way). I believe this is far from being the case for several reasons. First, management mentioned in Q1 2024 that the midpoint of 2025 guidance (before it was lowered) was theoretically “locked-in” so long as ASML was able to achieve average quarterly orders of €4 billion in Q2, Q3, and Q4 last year. The company blew past these estimates in Q4 2024 after delivering a blowout net bookings number. This is what I wrote in my last earnings digest:

Today’s situation mirrors almost perfectly that of 2023: after a weak Q3 bookings number, the company blew past expectations (€3.5 billion) in Q4 2024, delivering €7.1 billion. This has potential implications for the 2025 guidance. Management had mentioned in Q1 2024 that the company needed an average of €4 billion in net bookings over the next three quarters to “secure” the midpoint of its 2025 guidance (which at the time was €35 billion in sales). The company exited Q3 with an average of €4.1 billion in net bookings (Q2+Q3 net bookings divided by 2), so €4 billion was all that was needed in Q4 to, theoretically, secure this midpoint; as discussed, they delivered €7.1 billion.

Management basically confirmed this during this quarter’s earnings call by claiming that EUV 2025 revenue is already 100% locked in the backlog and that DUV orders are 90% there. Considering the lead times of both systems, I would take this as good news (DUV lead times are shorter, so it’s not that worrying that orders are not yet 100% complete). There’s no denying that the mix of orders and how cancellable they are matters, but there seems to be no sign to point to ASML grossly missing 2025 guidance, as many are alluding to.

What some people are doing is annualizing the Q1 order number and adding IBM (Installed Based Management) sales on top to get to a rough estimate of 2025 revenue. This method gives a false sense of accuracy because it misses several key things. For example, it completely misses that there’s a backlog and ASML has long lead times, meaning that current orders or those that happen in the remainder of the year (Q2, Q3, and Q4) are likely to be for 2026 revenue rather than 2025. It still puzzles me why so many investors use the order number as an indication of anything when history shows that it actually has very little predictive power.

Note that the same argument of “weak orders” could’ve been used several times over the past year, but the argument would’ve quickly been put to rest with the order number of the following quarter (not saying it’ll happen this time around!). With this, I am not trying to claim that I “know better,” just that a weak order number is not indicative of anything, and I am pretty sure management has more information than we do to remain confident in delivering on 2025 guidance (which they may or may not meet). We are already two weeks into Q2, and who knows how many orders they have gotten that the market doesn’t know about.

I honestly thought the market would’ve caught up already to the low predictive power of the company’s quarterly order numbers, but seeing that it hasn’t, management has (correctly, imho) decided to stop disclosing it. Just as a reminder, recall that quarterly order bookings will stop being disclosed in Q1 2026 and that the backlog figure will only be disclosed on an annual basis. This means that this year we could work backwards to get to the backlog figure, but that this will be pretty much impossible starting next year.

What should really be considered a good proxy for long-term demand is ASML's customers’ roadmap and the ability of the company to fulfil it. This seems right on track. Everything surrounding quarterly orders makes a great headline, but it’s honestly noise when thinking long-term.

It’s not just the quarterly orders that are misleading; pretty much all of the company’s quarterly numbers are. ASML grew sales and EPS significantly this quarter, but these growth rates should not be extrapolated to the full year because they are very lumpy. We could also see negative Free Cash Flow in Q1, something that’s also normal for ASML, just based on cash flow seasonality. I always hesitate to share the quarterly numbers for companies like ASML, but here they are:

Source: Made by Best Anchor Stocks

The most positive thing about the release came around management’s qualitative commentary on several topics. Let’s start with guidance and tariffs. Management reiterated both the 2025 and 2030 guidance, which let’s not forget is €32.5 billion and €52 billion, respectively (using the midpoint). They also mentioned that they still expect 2026 to be a growth year based on their conversations with customers and the orders included in the backlog.

---

---

In regard to 2025 guidance, management believes there are two possible scenarios (not rocket science):

1. If AI remains strong, then they’ll get close to the high-end of the guidance

2. If AI sort of falls off, then they’ll be around the low-end

That said, based on the 2024 orders figures and management’s commentary around the existing backlog, it seems likely that ASML will land somewhere around its stated guidance. We’ll see.

Tariffs were evidently given quite a bit of airtime (as expected). Management argued that they expect to do what’s necessary to diminish the impact of tariffs on their financials. They didn’t, however, state how they would do this. I would assume a somewhat limited direct tariff impact for several reasons…

1. ASML is a monopoly and could theoretically pass these costs to customers. Management alluded that the costs of tariffs should be borne by the value chain and not ASML per se. This came on a day when rumours stated that TSMC would increase prices in the US by 30% (coincidence?)

2. We must not forget ASML is based in Veldhoven and therefore is not impacted by the US/China trade war (thus far) in its sales to China/Taiwan. Tariff impacts are currently limited to the company’s US operations and also impact the parts that are shipped from the US to Veldhoven.

3. If the US truly wants to bring leading-edge manufacturing to its home soil, it makes little sense to set tariffs on the equipment needed to build said infrastructure. Judging by the common sense shown by the current administration, we might be surprised here, but I’d say semiconductor Capex will eventually be on a special tariff regime if the plan is really to onshore chip manufacturing.

Tariffs can, however, have an indirect impact on ASML by deferring the capital investments of its customers. This seems very likely (let’s not forget about the bullwhip effect), and I honestly was surprised when management mentioned that tariffs are still not a discussion topic with customers. These impacts are pretty much impossible to discern at this stage and definitely add uncertainty.

Something that investors should really care about (definitely more than about quarterly figures) is the EUV roadmap (both low and high-NA EUV). Management shared some data points that (albeit not new) I believe are worth highlighting:

• Intel and Samsung shared some positive data points about early high-NA implementation during SPIE (lithography conference). The most important highlight was probably process simplification, which should result in higher litho intensity

• Management mentioned that the maturity level of high-NA EUV is significantly ahead of what it was at the same stage for low-NA EUV, which bodes well for its adoption

• Three customers have high-NA installed or under installation, which is helping ASML get much more data to further improve the process going forward. It’s not enough to manufacture EUV, you also need to optimize it using real-world data (this is why the moat is larger than just technology)

• The newer and more productive low-NA model (220 wph) is seeing good adoption by customers. This more productive model enables lower complexity through fewer single exposures

• Gross margins came in above expectations in Q1 due to some performance milestones reached with EUV

There’s no denying that the cost of high-NA EUV seems to be a “problem” thus far (double patterning low-NA EUV still seems to be cheaper), but we are very early into its story and should see it improve its productivity over the years, just like low-NA EUV. Management talked about three phases of EUV implementation:

• Phase I: the current phase, tests on customer fabs

• Phase II (expected 2026-2027): customers already expose some layers with the new tool

• Phase III (beyond 2027): integral in the manufacturing process for customers

Management also shared some comments about China. First, they expect China to be around 25% of sales this year, slightly above the 20% they expected last quarter. Second, and most importantly, management also shared what they think about China’s “EUV alternative” that has gone viral on social media. Christophe Fouquet (ASML’s CEO) basically disregarded it as “propaganda” (albeit without using this word):

I think there is nothing really new there. I think that we expect to continue to see news here and there on some progress with regards to EUV in China. And I think this is mostly driven by a strong wish, I think, from China to have this tool and to display some progress.

If you look at the fundamentals, even if you look at what has been shown within some picture, I think that, you know, I would consider that as research news more than product news. And therefore, you know, of course, it's always possible to generate some EUV light, may even be possible to have an EUV mirror here and there. But in no way, this is enough proof that there is a serious product on the way. So I think we are still on the same view that it will take many, many years for China to be able to make an EUV machine. And again, you should expect some more news because I think that's that's just what you do when you want to show progress.

I would also like to touch on the topic of stock buybacks briefly because I think it’s kind of misleading some people who are taking these as a signal (when they are not). So, ASML repurchased €2.7 billion worth of shares this quarter. If we look back at historic repurchases, we can see that it’s the highest quarterly repurchase ever, together with Q4 2021:

Due to the stock’s recent performance and many people’s personal views that the stock is cheap, many thought that this signaled counter-cyclical repurchases by management. While it’s good that ASML is repurchasing stock at these levels, I don’t think management should be given much credit in timing these! As I see it, ASML basically repurchases stock when there’s money to do it, not when the stock is cheap. This is pretty obvious if we plot these buybacks together with the cash position of the company and the stock price:

So, while I would love to say that management is a great capital allocator in terms of repurchases, I don’t think this is remotely true. It’s good news, though, that ASML has money now when the stock is out of favor.

Some words on valuation

Not much has changed since the last time I updated my valuation on ASML (I made a quick update last quarter). Management has reiterated 2025 and 2030 guidance, so the only thing that has changed, pretty much, is the stock price (which is lower today). This means that if I believed ASML was a good buy last quarter, I still believe this to be the case today (maybe more so).

I will continue adding to my ASML position at these prices.

Have a great easter,

Leandro