TSM是一家拥有宽阔护城河和稳健增长的高质量公司，但其当前估值过高，不宜立即买入；应将其列入观察名单，等待价格回落。

I know a lot on the industry but not on the specifics for ASML and TSMC. Two vitally important input makers to the semiconductor chain. And this week's match up winner was ASML versus TSMC.

But we're going to take both these companies through the simplifier seven framework and we're going to ask ourselves which of these two stocks is the better buy today. So let us know in the chat. Are we talking to shareholders? Do you own either, both, neither, on your watch list? Let us know your interest level in these stocks.

Where is your sense for What do you think is the difficulty level of these companies? This is interesting because they have This is I don't know. I'm I'm going to say I don't know for a change. I'm going to say generally speaking I would rate these companies somewhere between medium and hard. The the technology, like what they do is quite hard to understand, but the the business models and stuff like that is on the easier side. I wouldn't put these out of the range of the average investor. These companies I don't think necessarily belong in your too hard pile.

I'm guessing Taiwan semi. Most likely, yeah.

stock number two is TSM. Also called TSMC, so that always confuses me about is this a is this C C or not? All right. This is a $2 trillion company. Also in the semiconductor equipment space. Also paying a dividend and the dividend uh the difficulty rating here is hard. Founded 3 years later, 2,000 employees, market beater over the last 5 years, over the last 10 Look, very similar returns over the last 10 years. And just uh a money-printing machine um over the over the long term. Although it looks like it did a whole bunch of nothing for a couple years there before really taking off um recently.

This is a company with a 60% gross margin, a little bit higher, a 50% operating margin. My goodness. My goodness, is that good? Net margin of 44% free cash flow margin of only 26%. So, huge difference between net margin and free cash flow margin. I'm guessing that's just massive CapEx spending uh to to grow.

Uh financials. Plenty of cash. Uh similar levels of debt. Earnings growth similar, free cash flow growth similar. Valuation. Uh 0.6 times sales. I doubt that. I doubt that that is uh correct. Uh there might be some currency things uh going on going down there. We'll check that. We'll check that more closely.

Uh all right. So, this is a company that is in the New York Stock Exchange but based in Taiwan. So, yeah, currency stuff could be playing some uh havoc with the numbers there.

TSMC is a pure-play semiconductor foundry. It does not design or sell chips of its own. It manufactures wafers and customers' designs. Around that core, it sells mask making, wafer probing, assembling and testing, design services, and advanced packaging. The company describes its business by the end markets its customers serve, high-performance computing, smartphones, connected edge devices, automotive, and digital consumer electronics. Revenue is earned per wafer produced. So, the business turns on being first to each new process and generation and on keeping expensive factories full.

That's interesting. So, it's it's on a it's I didn't realize it was a per per chip model. That's very very interesting. Uh that's far better than I thought it was going to be. Uh the company owns its its fabs and runs them in Taiwan, the United States, Japan, and Germany, which makes deciding how much capacity to build and where the central financial choice every year. What is distinctive is that it sells no chips of its own, so it never competes with the customers who hands its design.

Hence, the hence the price-to-sales ratio being 0.6. When measured not in US dollars, it does look like it's going to be a little bit cheaper. Um high-performance computing is 60% of revenue, growing 21%. Smartphones are 27% of revenue, growing 8%. Internet of Things, automotive, other.

Yeah, this has got to be the AI platform, high-performance computing, and it's growing 21% per year.

So, yeah, again, really strong growth, but the the star performer is high high-performance computing. Uh it's in the US mostly 74% China, Taiwan, Amea, and Japan. Customers. The buyers are chip companies that own a design and need someone else to build it. And they come in two kinds. Fabulous designers own no factories at all, so all the manufacturing is bought from outside. And then there's integrated device manufacturers, run their own plans, and still send work out when a design needs a process generation they cannot yet run themselves. The fabulous group carries the business. What they weigh when choosing is the company where is process technology, manufacturing excellence, customer trust, and service quality. Meaning earlier technology readiness, better quality, and faster yields. The alternatives are other foundry service providers that integrated device manufacturers that also sell capacity the company stated reach volume production on its recent process before competitors.

So it's just it's just outstanding at what it does. It doesn't compete with any of its um of businesses, and it also is it it's the manufacturing arm of fabulous chip fabless chip companies. Capacity is bought continuously rather on replacement cycle, and revenue grew

Well, I would still guess that revenue is fairly predictable. Customers who need the newest process generation have no substitutes. And look at its operating margin. Yeah, I would say that that's probably accurate.

How recession proof is it? It says modestly. And what's their competitive position? Dominant. Dominant competitive position, semi-recurring revenue business model, globally diversified. To me, this is a five out of five business.

I completely agree. It says capital return. Revenue is positive, profits are positive, have been for years, no buybacks but they are paying a dividend. Yep, capital return, easy peasy.

Does it have a moat? Two mech- two mechanisms at that full strength, plus returns near 30%. A wide moat. Huge operating margin, the designs are node bound, moving a taped out design means re-engineering and re-qualifying it. All right, so there are some switching costs and a barrier nobody has paid. Research ran 6 to 8% revenue yearly, the top on top of the capital bill. So they're investing on their own internal capabilities, um as well.

I mean, just look at the operating margin for does it have a does it have a um a moat? Uh yeah, I would say it definitely does. A new process generation is ramping on a step change in capital spending while returns keep expanding. So the moat is widening. There's two nanometer arriving, the node reached 3% of wafer revenue in the first quarter out. Revenue is still climbing. Uh returns are still climbing and the build is funded. Capital spending stepped up to 1 trillion NT. Uh one Yeah, trillion. Uh million millions in fiscal year 2025. So they are investing heavily and it looks like their newest generation, the two nanometers, are just just just ramping now.

Very hard. Once Yeah, once you start at the manufacturing process, there's huge switching costs. Does scale help each customer nodes deepens it? This is more of a networking effect. I doubt that that kind of does. Is there a cost advantage? Yes. The industry's heaviest capital base points to a yield and process cost advantage in cost, not price, that that rivals. So, there's a cost advantage and very hard for incumbents to copy. Yeah. I agree with with the AI is saying here. Wide moat business with a with a a moat that's extremely strong.

Revenue is compounded at 19% and above 15% for a high-growth business. It's well growing way faster than the economy. The latest quarter accelerated. Second quarter revenue was up 36% and 12% sequentially. So, the trailing rate is not peaking and there's but there's no backlog. So, let's look back at history. Revenue did decline in 2022. That was the pullback year for a lot of uh tech companies and profits dipped back here, but it's pretty much been a very steep, very like look at the acceleration in revenue ever since kind of like AI took off.

Is the industry growing? For sure. Can new offerings drive growth? Absolutely. Yeah, this to me I'm going to I'm going to rate it the same way I did um ASML. TSMC solid growth. Not five out of five, but four out of five solid growth.

Five out of five. This is a great management team. Earnings beat estimates in seven of the last eight quarters. The share count has not moved in five years and the heaviest spending produced the best return. Earnings only beat analysts in seven of the last eight quarters and there was a tiny shortfall. There's no dilution. And spending that earns what does keep on its return on capital.

Morris Chang, this guy is like legendary. He's the founder of the uh of the the company as a joint venture between the Republic of China and private investors. He served as his chairman until 2018. I I read something fascinating about Morris Chang a couple days ago. He was given zero stock in the company. Zero. Yet, he is a multi-billionaire from stock and all that stock he purchased with his own money.

Hall of Fame founder? Morris Chang. Absolutely. With with with shareholder alignment. Uh C. C. Wei has been at the company CEO for eight years and was at the company for 28. I love homegrown founders. Uh $500 million with the stock. He's been the CEO since 2018, chairman since 2024. After 28 years, he came up through operations, holds a doctorate in electrical engineering.

Do employees like working here? Uh 3.3 out of five stars. Do they beat expectations? Absolutely. Are they shareholder friendly? For sure. And do they have good returns on capital? Absolutely. And again, let's check that revenue execution. Revenue execution has looked really good. Re- really really good, really consistent. Yeah. To me, five out of five.

Moderate risks. The three factors range benign to a strong against the geopolitical elevator. Okay. One geographic risk. The majority of production sits in the Republic of China and the ability to resell runs through United States export control. Yeah, I would say there's some geopolitical risk here. The other three are contained. The largest customer is 19% of revenue. No named force. I'm guessing that's that's Nvidia is the the largest um uh percent of revenue. And execution argues against worse. Earnings beats everything in the last quarter. So, customer concentration potential issues, although it's seems like it's more. The 10 largest are 78% of revenue. Just as concentrated as ASML.

Disruption a threat. A missed process generation named a threat in the company's currently heading in this cadence. Yeah, yet another company Justin that's that's dealing in bits uh in the atoms, not bits. So, harder to disrupt. How much is outside of their control? Definitely with um geopolitical stuff. With excellent financials. How do you want to rate this one for riskiness?

I think it really nailed it for me as the moderate. I think that there are just because I don't see a ton of we'll call it if risks were bullet points, I don't see a long list of risks. But the political risk is so high that that's the only reason why it would go down to moderate from very low.

Yeah. Yet another company that if if there was no geopolitical, no customer concentration, no doubt would be a low risk company. It's basically it's basically a monopoly, but yeah, when you add those in, yeah, there I would say moderately risky is is is appropriate here.

All right, valuation. Taiwan Semi two out of five. Uh PE is 30. It's above its historic range. The dividend yield is 0.77. And same. So, PE ratio is about 30. You can see that the average is about 25. So, the {quote} "fair value" using a 5-year PE would be about 336 per share. This is 416. Got down to as low as Wow, it was Oh, it was 11 times earnings back in 2022. And if we look over a 10-year period, towards the higher end, but not insane.

Let's do free cash flow. Similar, although free cash flow consistently runs at a uh higher number than earnings. And if we go dividend yield, the dividend yield was almost 3% back in 2022. Wow, what a bargain it was uh back then. Free cash flow yield uh very similar. And if we do that same reverse DCF, uh yeah, this is going to be off. This is using the um This is using the other currency. Uh that that's why there. Um it's it's comparing a market cap in dollars to um free cash flow in new Ti- new Taiwanese. That's why you're getting that kind of weird number there. Um so, multiple analysis is definitely going to be uh the way to go here.

And yeah, looking really >> It's funny when you go back 5 years, you can just put your marker on November of 2022 when uh ChatGPT came out. And it's basically since then straight up. >> Yep. This is when ChatGPT came out. Yeah. Sentiment was quite low, and then boom, got more uh important.

Yeah, to me, very very similar though. Uh I I would rate it as I think I rate it Taiwan's uh ASML is very expensive. I would rate this as um >> And you put >> expensive. >> You put the other as two as well. >> Oh, I did. Okay. >> Mhm. >> I'll rate this. Look at this. Like almost the exact same composite score.

And this is a company in phase four, um great business, wide moat, strong moat direction, a good growth, fantastic management team, moderately risky, also expensive, also expensive valued. I'm also going to put this, I would say, on on a watch list. That's kind of where it's worth on. And for a price alert, I would go, let's just say, yeah, 333 would certainly be something.

All right, Justin. So, to review, both businesses, awesome. Both businesses, more expensive than you would want to kind of pay. Both businesses have good growth, and both are facing the same geopolitical risks, the same disruption risks, and the same customer concentration risks.

No pushback from me. I I don't think you're going to go wrong with either company. Either company is is really fantastic. If I had to choose, I'm going with Taiwan Semi. I like going with larger companies. It's a two trillion versus 600 billion. I like the business model of Taiwan Semi um bit better. Um the risk I think is is prob- is the risk is higher for Taiwan Semi um absolutely, but to me you're getting compensated for that with a a better valuation.

I think Taiwan is more expensive, but I would rate it as a two out of five. So, to me very similar businesses, very similar growth, very you're taking on risk either way, but if I had to choose one, I think I would give the slight edge to Taiwan Semi uh here is what is what I would choose. But again, two absolutely absolutely fabulous businesses.

Yeah, this really in my head this is about as close as you're going to get to a [laughter] coin toss comparing the two because you could just go down the individual lines and it's ooh, this is a I like this. This is a good comparison right here.

And then here's the one-page overview of uh Taiwan Semi uh that we have. Some of the things you can do here, you can check out the financial statements and they're and they are set to the Taiwanese uh market there. So, you can see just how like consistent the revenue growth uh has been. From a capital return perspective, both of these companies are paying dividends. You can see that uh Taiwan Semi really been paying a dividend for uh quite some time. Both of them have been buying back stock. Uh both uh uh excuse me, Taiwan Semi has not been buying back stock, but the stock-based compensation, even though this looks like a lot in absolute terms, this is very very small. Like look at the dilution rate. The dilution rate has been effectively zero for uh for 10 years. So, I love management teams that kind of that that kind of uh do that. So, yeah, my choice Taiwan Semi, but I don't think you can go wrong with either.