Latest / Deep Values: Value Investing Deep Dives / ASML Deep Value Analysis: Is the Chip Monopoly an Undervalued Gem Still?
Transcript
- 0:00Welcome to the deep dive. Today we're tackling a really
- 0:03fascinating company when you probably haven't spent much time
- 0:06thinking about, but maybe you should.
- 0:09ASML. That's right, the Dutch giant.
- 0:11They don't actually make the chips you know, in your phone or
- 0:14computer, but they build the absolutely essential machines
- 0:17needed to make all the advanced ones literally indispensable.
- 0:20Exactly. If you want cutting edge AI, the
- 0:23latest phone, that chip went through an ASML machine
- 0:26somewhere along the line. For sure.
- 0:28And our sources today we've got a detailed financial research
- 0:30report, a risk analysis, a full 10 year financial breakdown.
- 0:35They let us do one key thing which is apply those classic
- 0:40strict value investing rules, think Warren Buffett style.
- 0:43We want to see if ASML is both a great company and if it's
- 0:46selling at a great price. That's the classic tension,
- 0:49isn't it? Quality often comes with a hefty
- 0:51price tag. Always the challenge, so.
- 0:52Before we jump into the numbers, just the usual quick reminder
- 0:55for everyone listening. This is purely for educational
- 0:58discussion. We're analysing the source
- 1:00materials we have, but this isn't financial advice in any
- 1:03way. Always do your own research,
- 1:05maybe talk to a professional. Absolutely.
- 1:06OK, So First things first, why ASML?
- 1:09Why indispensable? Yeah.
- 1:10What's the secret sauce? It really comes down to 1
- 1:13technology. Extreme ultraviolet lithography.
- 1:17EUV. EUV, right?
- 1:19And they are the only supplier globally.
- 1:21The only one. If you want to make chips below,
- 1:24say 7 nanometers, the different really advanced stuff, you have
- 1:28to use their machines. There's no alternative.
- 1:30That sounds like a textbook definition of a technological
- 1:32monopoly. How do they even build that?
- 1:34Well, it wasn't overnight, it was decades of honestly really
- 1:38expensive R&D, lots of trial and error.
- 1:40I bet our sources mentioned something like €6 billion spent
- 1:45just to get EUV working properly. 6 billion just on
- 1:47development. Just on development, a 20 year
- 1:49marathon, not a Sprint. And that built this well, this
- 1:53technological Moat that seems almost impossible for anyone
- 1:56else to cross right now. OK.
- 1:57So a massive R&D effort created the monopoly.
- 2:00How does that translate into, you know, the actual business
- 2:03model, the financials? It creates some really strong
- 2:06and kind of predictable revenue streams.
- 2:09The report breaks it down. About 77% comes from selling the
- 2:13brand new lithography systems, the big ticket items.
- 2:16Right, the $150 million machine. Exactly.
- 2:18But the other 23 percent is also really important.
- 2:22It's what they call installed base management.
- 2:24Installed base SO servicing the machines already out there.
- 2:28Precisely Services, software upgrades, spare parts,
- 2:31maintenance contracts. It's very stable, almost
- 2:34recurring revenue. That part alone was about 6.5
- 2:37billion in 2024. Wow, 6.5 billion just from
- 2:41services? That sounds like a great buffer
- 2:43against the usual chip industry cycle.
- 2:45It really is. It provides stability.
- 2:46But let's talk customers. Who's buying these incredibly
- 2:49expensive machines? Has to be a short list, right?
- 2:51Very short. You're talking about the Giants,
- 2:53TSMC, Samsung, Intel. A handful of players drive the
- 2:58vast majority of the orders. This.
- 2:59High customer concentration, is that a risk?
- 3:02It can be, but here's the thing, those customers need ASML.
- 3:05They have no other options for EUV if they want to keep making
- 3:08faster, smaller chips. They have to buy from ASML and
- 3:11keep paying for the service. They're locked in the cost for a
- 3:14competitor to even try and catch up.
- 3:15We're talking 10s of billions. Many many years of development.
- 3:19Maybe they never even get there. The Moat is just that strong.
- 3:22OK, Ironclad Moat seems fair. Let's pivot to the financials
- 3:26then the great company test, right?
- 3:28The 10 year analysis in our sources, some of these numbers
- 3:31are just, well, incredible. They really are top tier.
- 3:33Let's start with gross margin consistently high, generally
- 3:37between 44% and nearly 53% over the last decade.
- 3:42The latest quarter, Q 3/20/25, was 51.6%.
- 3:46And the benchmark for quality is usually around 40%.
- 3:49Yeah. For a quality company, value
- 3:51investors often look for 40%. ASML consistently clears 50% and
- 3:55that just screams pricing power. It means over half of every euro
- 3:59of revenue after the direct cost of making the machine is gross
- 4:02profit. Incredible efficiency and does
- 4:04that flow down to the bottom line net profit?
- 4:05It absolutely does. Net profit margin, again
- 4:08consistently high, usually between 21% and over 31%.
- 4:12Easily clears the typical 20% quality threshold.
- 4:15OK. Strong margins.
- 4:18What about cash flow? That's key for value investors.
- 4:21Yes, free cash flow, FCF, using that as our proxy for owner's
- 4:25earnings, it's been positive every single year for the last
- 4:2810 years. Every year, even through
- 4:30downturns. Every year and not just
- 4:32positive, but averaging, you know, around 25 to 30% of
- 4:36revenue. Wait, 2530% of revenue
- 4:38converting straight to free cash flow in a manufacturing business
- 4:42like this? Exactly.
- 4:43It's extraordinary. The minimum threshold quality
- 4:46investors look for is maybe 5%. ASML blows that away.
- 4:50So where does all that cash go? Well, it fuels fantastic returns
- 4:54on capital. Their return on invested capital
- 4:56ROIC is consistently above 20%, often way higher and return on
- 5:00equity ROV even as their equity base growth has been normalizing
- 5:04around a very healthy 40%. So they're not just profitable
- 5:07or extremely efficient at using capital to generate more profit.
- 5:10Masters of capital deployment and you add to that a really
- 5:12strong balance sheet, very conservative debt levels, huge
- 5:15interest coverage. OK.
- 5:16OK. So the great company test.
- 5:18Passed with flying colors. Really, on quality metrics, it's
- 5:22a plus. Right, but quality often assumes
- 5:25things stay stable. What about risks?
- 5:28You mentioned the geopolitical angle earlier.
- 5:29Yes, that's the big one right now.
- 5:31The near term headwinds are definitely geopolitical.
- 5:33So how do the latest results look?
- 5:35Q 3/20/25. Pretty solid actually.
- 5:38Net sales were stable around 7.5 billion and new bookings, which
- 5:43tells you about future demand, were really strong, 5.4 billion,
- 5:47more than double the previous year.
- 5:49So the pipeline looks good. Yeah.
- 5:51And management is still guiding for about 15% sales growth for
- 5:54the full year 2025 S near term looks OK.
- 5:58But there's a but isn't there. There's a but, and it's a big
- 6:00one, particularly concerning China.
- 6:02Management explicitly warned about 2026.
- 6:05What's the warning? China was a huge chunk of sales
- 6:07recently, something like 46% in 2024.
- 6:10Wow, nearly half. But they expect that number to
- 6:12drop significantly, both the percentage and the absolute
- 6:15amount going to China in 2026. Or the drop export controls.
- 6:19That's the main driver, tightening export controls from
- 6:22the US and Netherlands, plus maybe some demand.
- 6:25Poland happened earlier, sort of a normalization.
- 6:27But yeah, it's a major headwind losing a big chunk of that China
- 6:31business. That sounds like it could derail
- 6:33the growth story quite a bit it. Definitely creates a challenge,
- 6:36a growth gap, but management tried to put it in context.
- 6:39How so? They said even with the China
- 6:41drop, they expect total revenue in 2026 to be at least flat
- 6:46compared to 2025. Flat how?
- 6:48Where does the replacement demand come from?
- 6:50Other regions, think Taiwan, South Korea, the US places
- 6:54heavily investing in AI infrastructure, advanced logic
- 6:57chips. The underlying demand for
- 6:59cutting edge chips isn't going away.
- 7:01It's just shifting geographically because of the
- 7:02restrictions. Exactly.
- 7:04The end market pool seems incredibly strong.
- 7:06OK. So geopolitical political risk
- 7:08is managed maybe? What about the really long term
- 7:11existential stuff? Could someone invent a totally
- 7:13new way to make chips? That's always the background
- 7:16risk, technological disruption. You hear about things like
- 7:20electron beam direct, right? Maybe quantum computing
- 7:23eventually things that could in theory replace optical
- 7:27lithography? But are any of those close?
- 7:30Not really. Not at the scale and cost
- 7:32effectiveness needed for mass production.
- 7:34And ASML isn't sitting still either.
- 7:37What are they doing? They're pushing the boundaries
- 7:39themselves, developing the next generation high NA EUV for chips
- 7:43below 2 nanometers. That keeps optical lithography
- 7:46relevant for the foreseeable future.
- 7:48Plus they're moving into adjacent areas like advanced
- 7:51packaging lithography. So trying to stay ahead of the
- 7:53curve. Definitely the leadership seems
- 7:55secure for now. OK, let's pull this all together
- 7:57then. We've got ASML is a truly great
- 7:59company, phenomenal business, ironclad Moat, fantastic
- 8:03returns. Now for the tough part, that
- 8:05second test, great price. How does ASML stack up against
- 8:09traditional strict value investing price metrics?
- 8:13Well, this is where the story gets complicated.
- 8:16Or maybe simple, depending on your view.
- 8:17How so? It fails spectacularly on
- 8:20traditional metrics. Fails like what?
- 8:23Take the price to earnings ratio.
- 8:24PE Classic value investors often look for 15X earnings or less.
- 8:30ASML currently trades around 41 times trailing earnings. 41 OK,
- 8:34not even close to 15. Not even in the same ballpark
- 8:36and it's basically never traded consistently near 15X in the
- 8:40last 10 years. What about Price to book another
- 8:42classic? Same story.
- 8:43The traditional threshold might be 1.5 X book value.
- 8:47ASML is currently around 10.8 X. Wow OK so PE and PD are way off.
- 8:53Why are those metrics 15 XPE 1.5 XPB even used?
- 8:57What do they signify? They generally come from valuing
- 9:00more mature, stable, maybe slower growing industrial
- 9:03companies where growth expectations are modest.
- 9:05A 15 XPE suggests the market isn't pricing in massive future
- 9:09growth. And ASML at 41 XPE.
- 9:10The market is screaming that it expects huge continued growth.
- 9:14It's pricing in the monopoly, the tech leadership, the future
- 9:16of AI, not just its past earnings.
- 9:18So the traditional metrics might not fit this kind of company.
- 9:21That's the debate even on earnings consistency, another
- 9:24value check. ASML stumbled slightly.
- 9:27I thought Ernie's growth was strong.
- 9:28The long term trend is fantastic.
- 9:30Yes, about 21% compound annual growth in EPS over the decade,
- 9:35but it wasn't perfectly smooth. There were minor dips, flat in
- 9:392019, a small dip like 4.4% in 2022, mostly due to investment
- 9:44cycles or R&D timing. So not rising every single year,
- 9:47which is the really strict rule some apply.
- 9:49Exactly. For a purist, that tiny bit of
- 9:51volatility is a technical fail on that specific.
- 9:54Rule. OK, so the simple metrics fail.
- 9:57What about a more sophisticated valuation like discounted cash
- 9:59flow? Does that make it look cheaper?
- 10:01Well, the DCF analysis highlights just how sensitive
- 10:04the valuation is to your growth assumptions.
- 10:06The current stock price is around $985 per share, give or
- 10:10take, right? If you plug in a really
- 10:12conservative growth rate for free cash flow, say, it stalls
- 10:15out at only 5% per year. Which seems unlikely given the
- 10:18demand, but OK. Right as a stress test, the DCF
- 10:20spits out an intrinsic value of only about $403 per share.
- 10:24Meaning hugely overvalued if growth disappoints badly.
- 10:27Massively. But now let's use our nominal or
- 10:30sort of base case assumption from the analysis. 15% FCF
- 10:33growth per year, pretty strong, reflecting the AI boom.
- 10:36What value does that give? Around $850 per share.
- 10:39Still below the current $985 market price.
- 10:43Correct. Even with robust 15% growth
- 10:46baked in, the analysis suggests it's still maybe 16% overpriced.
- 10:50No margin of safety there for a value buyer.
- 10:52So what growth rate does justify the current price?
- 10:54You need to assume even higher growth if you plug in a really
- 10:58optimistic 25% FCF growth. Which would be phenomenal
- 11:02sustained growth. Then the value jumps to about
- 11:03funder $755 per share, so the current market price around
- 11:07$985. It's essentially baking in
- 11:10sustained FCF growth in the high teens, maybe 1718% per year for
- 11:15the next decade. So the market is pricing in
- 11:18continued dominance and strong growth.
- 11:20It understands the quality. It's just demanding a premium
- 11:22for. It precisely, it reflects the
- 11:24confidence in Asml's unique position and you have to
- 11:27remember management's role here too.
- 11:28How so? The leadership, folks like the
- 11:30new CEO Christophe Fouquet, the CFO Roger Dasson, they have a
- 11:34proven track record. They keep investing heavily in
- 11:36R&D, that 4.6 billion in 2024 to maintain the lead.
- 11:40And returning cash. And returning cash, big
- 11:43dividends 6.4 to use it for 2023 and massive share buyback
- 11:48programs. There's a 12 billion
- 11:50authorization running now. They're reinvesting for growth
- 11:52and returning capital, very shareholder aligned.
- 11:55OK, so wrapping this all up, what's the key take away for
- 11:58you, the listener? Well, ASML is, without death an
- 12:01absolutely phenomenal business. It nails the great company
- 12:04criteria, the mode is real, the returns are world class, but it
- 12:08consistently trades at a very rich valuation.
- 12:11It just fails. The strict price discipline of a
- 12:14traditional value investing approach always looks expensive
- 12:16on paper. So it forces a choice.
- 12:18It really does. Do you stick rigidly to your
- 12:21price rules? Maybe looking for that 15 Xpe
- 12:24and potentially miss out on what could be decades of compounding
- 12:27from a unique monopoly. Or do you decide that this level
- 12:30of quality that's kind of locked in growth trajectory in a
- 12:33critical industry justifies paying that premium that maybe
- 12:36the traditional rules don't fully capture this specific
- 12:39situation? Which leads us perfectly to our
- 12:41final provocative thought for you to consider ASML's next
- 12:44generation machines, the high NAUV.
- 12:47They're essential for pushing chips below 2 nanometers.
- 12:51That's the future of AI, Advanced computing, everything.
- 12:53The absolute cutting edge. So if that demand truly
- 12:57explodes, if AI requires computing power we can barely
- 13:00imagine today, will the investors who waited, who
- 13:04insisted on buying ASML only if it hit some arbitrary low PE
- 13:08ratio, will they look back and realize they missed out on one
- 13:12of the truly great compounding machines of our era?
- 13:15Does extreme quality in rare cases like this actually make
- 13:18the traditional concept of a fair price almost irrelevant?
- 13:21Something to definitely think about, indeed.
- 13:23And if you found this deep dive interesting and want to hear
- 13:25more analysis like this, definitely subscribe to our
- 13:28channel, Deep Value Investing on Apple Podcasts or Spotify.
- 13:31We'll be back soon with another one.
- 13:32Until next time.