I calculated the worst case for ASML: 7.4% of EPS by 2030.
The market has been behaving in recent days as if $ASML will cease to exist in a few years. A report about five Chinese lithography machines was enough, and the stock is down 12%. I built my own worst-case model, and even in the blackest scenario where China replaces ASML machine for machine, I get that it would knock earnings per share down by about 7.4% in 2030. The market is pricing something completely different. Let's break it down.
Why this matters now
ASML is the only company in the world that can make EUV lithography machines – equipment without which the most advanced chips cannot be produced. That's why nobody worried about the company for so long: it has a monopoly on the most sophisticated tier of the market. But China has been working for years on its own alternative, especially on older DUV machines (these are less advanced lithography systems that suffice for less sophisticated chips in cars, sensors, or cheaper electronics). When news emerged that a Chinese manufacturer had delivered five such machines to domestic customers, the market immediately interpreted it as the “beginning of the end” for ASML in China. Hence the drop.
The question everyone who looks at numbers and not headlines must ask themselves: how much would it actually cost ASML if this fear were fully realized?
How I arrived at this
I built my own estimate of how many machines ASML China will realistically replace with domestic production by 2030 – not cumulatively, but how many new machines it “steals” from ASML each year:
- 2026: 5 machines
- 2027: 20
- 2028: 40
- 2029: 55
- 2030: 70
Total: 190 machines over five years.
I estimated the price of a single DUV machine at 60 million euros – that matches ASML's average price for more advanced immersion machines, i.e., those that Chinese competition would realistically target first. I assumed a 60% margin, 15% tax, and 1% annual share buyback, which slightly reduces the number of shares outstanding and thus modestly lifts earnings per share even without revenue growth.
Result: cumulatively over five years, it’s 11.40 billion euros in revenue, 5.81 billion euros in net profit, and 15.61 euros in lost earnings per share. Year by year it looks like this – always the impact of that specific year, not the sum:
- 2026: loss of 0.40 euros per share, i.e., 1.0% of total profit
- 2027: loss of 1.62 euros, 3.2%
- 2028: loss of 3.26 euros, 5.4%
- 2029: loss of 4.52 euros, 6.5%
- 2030: loss of 5.82 euros, 7.4%.
The numbers are consistent – “retained” profit plus lost profit always equals the total estimated earnings per share for that year, and the percentages match the ratio. This is not a randomly invented chart; it’s an internally consistent model.
Why this is a worst case and not a forecast
Here is the most important sentence of the entire article: this model makes sense only if you accept the idea of complete 1:1 replacement – i.e., every Chinese machine that rolls out of the factory means one lost sale for ASML. I do not subscribe to that idea. It is a ceiling, not an estimate of what will actually happen.
Reality is different. ASML has its manufacturing capacity (backlog) sold out for 2026 and 2027, and very likely for 2028 as well. That means even if Chinese machines started replacing sales today, ASML has demand from customers who are waiting for their machines and have nowhere else to get them – especially for cutting-edge EUV technologies, where China is not even close. The short-term impact should thus be significantly lower than the worst-case model shows, and even the long-term impact will probably be milder, because the model assumes zero market growth and full shift of demand, which is an unrealistic simplification.
So: the worst case justifies a 7.4% decline in profit by 2030. The stock is down 12% already now, based on a report about five machines. The market today is pricing a scenario close to ASML being fully pushed out of the Chinese market – and the numbers do not support that at all.
How I see it
I have a position in ASML, so take that into account – I’m not an independent observer; I have skin in the game. But that’s precisely why I wanted to calculate the numbers for myself, not react to a headline. When I see that the worst-case scenario allows a single-digit profit decline, yet the stock has fallen by a double-digit percentage on the basis of one report about five machines, that’s a signal to me that the market is reacting to fear, not to the math. I’ve seen this pattern a few times – the market first prices in an extreme scenario, and then it waits to see whether the fundamentals catch up to the price, or the price catches up to the fundamentals.
That doesn’t mean there’s no risk. China has a clear strategic interest in becoming independent of Western technology, and the money and political will to push further than many expect. But there’s a huge difference in time horizon and certainty between “China will have a competitive alternative at the lower end of the market in ten years” and “ASML will lose 7.4% of profit in a worst case by 2030, and the market reacts with a 12% drop.”
How to act on this news
If you hold ASML or are considering it, this drop in itself is not a reason to panic – it’s a reaction to a headline, not to a change in fundamentals.
Going forward, I would mainly monitor actual orders from Chinese customers (not just the production capacity of Chinese machines), the development of ASML’s backlog for 2027-2028, and whether Chinese machines actually reach customers outside China – that would be a real red flag, because today’s model only counts the domestic Chinese market.
Be prepared that similar reports will keep coming and the market will react emotionally to them. Those who can calculate what a number really means for a company’s profit have an edge over a market that just reacts to headlines.