Samsung: Semiconductors over Smartphones - [Business Breakdowns, EP.147]

0:03This is Business Breakdowns. Business Breakdowns is a series of conversations with investors and operators diving deep into a single business. For each business, we explore its history, its business model, its competitive advantages, and what makes it tick. We believe every business has lessons and secrets that investors and operators can learn from. and we are here to bring them to you. To find more episodes of Breakdowns, check out joincolossus .com. All opinions expressed by hosts and podcast guests are solely their own opinions. Hosts, podcast guests, their employers or affiliates may maintain positions in the securities discussed in this podcast.

0:45This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. This is Matt Russell and today we are breaking down the technology conglomerate Samsung. It was only Apple, Microsoft, Google, and Amazon that ranked higher than Samsung in Interbrand's latest brand value rankings. Samsung being the fifth most valuable brand in the world. And it's everywhere around us. Our phones, our TVs, our refrigerators, our washing machines. You see Samsung everywhere. But it's not those finished products that are driving the majority of profits for Samsung. To break down Samsung, I'm joined by David Samra, Managing Director and Founding Partner of the Artisan Partners International Value Team.

1:29We go inside this vertically integrated technology giant. We talk about the history of the business, the manufacturing DNA, and what it means to create hardware components, and how creating those hardware components unlocks some major opportunities in the smartphone market for Samsung. We also get into the management history, what K .H. Lee took from his early days managing this business, and some of the opportunistic acquisitions that he made, and how that really shaped the culture, specifically around capital allocation and how they manage the balance sheet. This one is really interesting because we all know Samsung the brand, but I think Samsung the business is very different than what I expected.

2:06So please enjoy this breakdown on Samsung. All right, David, I am excited to break down Samsung. it is a behemoth. And I've listened to you speak before. You have a background as a value investor, and you always preach boots on the ground, really understanding these businesses, the people and the things in and around them. And Samsung just does so much. It's the things we see day to day with the TVs and the refrigerators, but it's also a lot of the components within those pieces of hardware. And I'm just curious, at the highest of level, when you think about Samsung as a business and as an investor.

2:45How do you define it? How do you think about it? How do you just categorize this one relative to all the other pure plays that we're used to seeing and things that we're used to breaking down on this particular show? Well, it's a very relevant question, mainly because the consumer electronics brand is so strong. Yet, that's not where the bulk of the profits are generated. Everybody knows the Samsung brand. In fact, Interbrand, which ranks the most popular brands in the world, this is the fourth year in a row that they've ranked the Samsung brand in the top five. So the company is well known for televisions, home appliances, and probably best known for its cell phones.

3:30Yet in most years, those businesses represent less than 30 % of the profits. Most of the profits come from the components that power these devices. That's where the other 70 % comes from. And that's how we think about and how we define all of the businesses we invest in. Where do the profits come from? Where are today's profits generated? But more importantly, where will tomorrow's profits be generated? And with a very high degree of certainty, future profits will be even significantly more generated, mainly because the handset business is mature at this point. But in the future, the components businesses will generate more profits than the consumer electronics part of the business, if I can categorize it that way.

4:18And just to provide you with a little bit of a breakdown, the memory business, Samsung is the world's largest manufacturer of DRAM and NAND generated 55 % of profit in 2022. The mobile phone business generated another 25 % of profit. And then they have another business that makes display panels. So when you're looking at your computer or you're looking at your phone, that display that you're looking at is most likely manufactured by Samsung, even your Apple phone. That generates It's another 14 % of profits. So all the other consumer electronics that everybody looks at while they're watching their football game on Sunday, it's meaningless.

4:57It generates 3 % of profits. So really, that's how we define that business. And business observers would say this is a somewhat unique company because of its vertical integration. Generally, if you look at a consumer, an old school consumer electronics company, Sony, maybe Panasonic, the big Japanese manufacturers, they would supply their own components, but definitely wouldn't have a significant amount of third -party sales from their component manufacturing. Whereas, as you can see from Samsung's breakdown, most of their profits actually come from those third -party sales, yet they're vertically integrated.

5:40And the biggest business is obviously making handsets. But that's an accident of history and an accident of really the way cell phones developed many, many miles away where Apple has this closed ecosystem and they have their own operating system that powers their devices. However, Google, in an effort to grow its search business, developed the Android system where they preferred to push out its operating system to third -party handset manufacturers. And given Samsung's scale and the manufacturing of components, so that would be memory, semiconductors, application processors, and screens, they were in an incredibly strong position to grow and build a very, very large handset business.

6:33So it's an accident that they are this big, large, vertically integrated company. But I think kudos to the management of Samsung for recognizing that opportunity, recognizing the fact that they had significant competitive advantages in that business and building it out. And that's where we are today. It's really interesting to hear that breakdown in all of the research that I did. I saw divisions and segments, and most of them you're looking at sales numbers, not the profit numbers and simply presenting it in that way with memory at 55 % mobile at 25 % display. A very simple way to look at this business.

7:09And your description of what they do, it rings true to something like an Amazon with AWS and where the majority of their profits come from and some similarities in terms of how they've been able to leverage what they built inside the organization and bring it outside the organization. And you tapped into it a little bit there, But I am curious about that history. It's massive within South Korea, north of 20 % of South Korea's GDP comes from Samsung is what I read, which is incredible. So can you bring us back a little bit up until that point in time? Where did the DNA come from in terms of building up that core capability to do everything that they do today?

7:46I think that that comparison to 20 % of GDP is a little bit misleading, because I think what they're referring to is revenue. And Samsung Electronics revenue to South Korea's GDP is about 11%. And then there are other chable or group companies that probably make up the rest of it to head towards that 20 % number that you referred to. But it is somewhat remarkable when you internalize where we are in the United States. We've got a little over $23 trillion of GDP and the largest revenue company here is Walmart. And that comes to 3 % of GDP. And it's somewhat instructive to consider this. And it gets back to the history of Samsung, because stick yourself in the United States, trying to build out a consumer electronics or a semiconductor business.

8:35And you've got this giant $23 trillion marketplace where everybody speaks the same language and you have one single legal system and 320 million wealthy people. And you have a big profit pool to aim at in a roughly simple way where you understand the culture and you understand the rules of business. Put yourself in a place like South Korea, which is a trillion and a half of GDP and much smaller than the US, or in a place like France, for that matter, where companies like LVMH were built. And your domestic profit pool is tiny. So how do you take an area where you have a good understanding of the culture, you speak the language, you operate in the same time zone, and then you turn this into a massive, competitive, multinational business?

9:27That is, as Charlie Munger would say, a Lollapalooza. That's an extraordinary outcome for a company that started in a small country. And just to simplify the history of this company without spending a lot of time on it, the company really started as a component manufacturer and then evolved into manufacturing its own consumer electronics, but really entered what is their largest business today when in the 1960s, there was a company that was started in South Korea that was basically poorly financed. And the third generation here, a fellow named K .H. Lee, who just recently passed, I think was the most powerful entrepreneur in the driving force behind the building of Samsung, actually recognized the future of memory semiconductors.

10:18They had been discovered in the 1960s to be used in calculators. And as electronics grew, the use of these memory semiconductors would grow significantly. And this fellow, K .H. Lee was his name, actually personally financed the purchase of this memory semiconductor operation and embarked on a multi -decade process of aggressively building it into the giant behemoth that it is today. Today's CEO, J .Y. Lee, is the son of K .H. Lee. So family owned, despite acquiring some of the businesses that run it today, it has remained inside the family over that time period. It was listed on the stock exchange, which diluted the family's ownership.

11:02And there have been, in the early years, a couple of capital increases that were necessary to fund what is a very capital intensive business. And one of the things that KH Lee recognized, probably because he saw what happened to the original operation, right? It was poorly financed and wasn't going to survive. The one thing that he did recognize is being financially conservative in a capital intensive business that's cyclical provides you with a distinct competitive advantage. And that financial conservacism remains today as a distinct characteristic of this company. And we could talk about it later, but is actually playing into the hands again of Samsung as we go through this current downturn.

11:44It's very interesting to me. I can flash back to five to seven years ago, as I was sitting in a cyclical seat. And one of my peers that covered cyclicals, or it was at least considered a cyclical industry, was the semiconductor analyst. And they view things very much in that framework. And it feels like things have shifted a bit over the years. And obviously, you've had major supply -demand dynamic shifts, and you had COVID and all different things that were playing into it. But when you think about Samsung over the course of the past maybe 25 years or so, and particularly in the memory business, has this been a business where there have been share gains that have been taking place?

12:21Have they been just naturally exposed to the growth of the industry and maybe just growing along with it? Is there anything unique in terms of what they've done within that industry as it relates to growth in their own share? The key to the business is it's capital intensive, manufacturing scale, and to be at the leading edge of technology. So let's demystify leading edge of technology, because if you can't understand something at its very basic level, it's not worth doing. And the leading edge of technology for a memory chip basically means you can manufacture a smaller chip that uses less power and delivers the same or better capability.

13:04And that requires using the latest semiconductor manufacturing equipment. It requires building factories that are very large in scale. And that means hiring top -notch engineers focused on using those machines to get to the next level of die shrink. So what that means is your chip is getting smaller, faster than the competition. And that's what really has driven market share for Samsung. So this industry, the DRAM industry specifically, has consolidated over the last three decades down from 10 or 11 players down to three. So what caused that consolidation was some of the advantages that I alluded to earlier.

13:54Some companies simply weren't financed very well. And into an industry downturn, they didn't have the financial capabilities to survive. The business is very capital intensive and it's very hard for a board of directors to keep saying, we're going to invest more and more and more money into this business. And then shareholders are saying, well, what about the dividends? When are the dividends coming? Well, the dividends are going to come when we get to scale. Samsung, perhaps because it's family owned and controlled, had a lot more patience behind the timeframe that was necessary to get to that point of scale.

14:33And they survived this consolidation. Hynex, which is the second largest player, is also a South Korean company, survived through the generosity of the South Korean government, which bailed it out. And Micron, which is the third largest player, survived through the goodness of their shareholders that kept on funding rights offerings. So the market share today breaks down in the following way. Samsung is dominant with 43 % market share. Hynix and Micron are much smaller with 27 and 23 % market share. The second type of memory called NAND, and we could talk about what the difference is between those two, is much less consolidated.

15:17NAND is technologically less demanding and has lower capital intensity. and as a result is much more competitive. And you're seeing companies losing significant amounts of money in this downturn in the end. Samsung has about a third of that market. A Japanese company called Kosha has 20%. And they've been working on a merger with a company inside of Western Digital, which has another 13%. The DRAM competitor, Hynex, has 18 % share. And also Micron trails these companies with a tiny 10 % market share. So the profit pool is spread a little bit wider in NAND than it is in DRAM. But Samsung's scale in both of those businesses allows them to generate much higher profit margins than their next largest competitor.

16:07It generates a significant cash flow, operating cash flow, which then allows them to build out capacity to feed growth in that business. And protect what they've built and build upon that. When you discuss memory chips, just discussing chips and semiconductors in general, my mind immediately goes to names that I hear about, like NVIDIA. So where does what Samsung does with memory chips, how does that compare to what NVIDIA is doing with chips? And are they overlapping in some of the same markets? This is going to be the question for some of the novices or tourists in the semiconductor space in terms of where there's overlap and where there's differences?

16:47There are all sorts of semiconductors out in the marketplace. NVIDIA, Texas Instruments, Intel, Qualcomm, they mainly make processing chips, chips that compute, whereas DRAM and NAND are chips that provide memory, meaning you can store information on these chips. And the difference between DRAM and NAND is DRAM is a chip that operates while the computer is on and as information is moving back and forth between the processing chip and a temporary storage chip, whereas NAND is effectively a solid state hard drive. And NAND is solely but surely, and that's what's driving some of the demand for these chips.

17:32NAND is taking market share from traditional hard drives. So the products that use these chips, the use case, are smartphones, personal computers, servers, autos, and any other IoT. People like to use these terms, Internet of Things, any other product that requires some information to be stored. And as these products become larger and more complex, they demand more storage and more processing power. So these application processors, such as NVIDIA's AI chips, are basically the highest powered processing chips. And guess what? In order for those to function well, they require much more memory. For example, just to start with history, the way the volumes are measured in the DRAM industry is in 8 gigabit chips.

18:25So that's the standard volume -based measurement for memory semiconductors. And over the past 10 years, DRAM volumes have grown at 23 % per year, While NAND, because it's grown a lot faster because it's not only growing along with these products, but also taking market share from hard drives has grown at 37 % per year. And though those growth rates look pretty high and you might be skeptical about what it looks like in the future, we think that the future here is pretty bright. For example, a chat GPT -based server requires five times more memory on average than a standard server. And an L5 fully autonomous vehicle requires greater than 30 times more DRAM and over 100 times more NAN than an internal combustion engine vehicle.

19:16And if edge computing becomes a necessity for AI, things start to get really interesting for Samsung because more memory will be required in a new generation of mobile phones and personal computers. In fact, you saw recently that Microsoft released Copilot that has to be used with Windows 10, and you need a device with a certain amount of processing and memory capacity in order to use Windows 10. So this potentially it's very interesting for Samsung because if there is a refresh cycle across the hardware endpoint spectrum, you're not only selling a lot more DRAM and NAND, but you're selling a whole new set of handsets.

20:01And as new personal computers get sold, you're selling new screens and more DRAM and more NAND. So a refresh cycle not only takes the baseline growth into this company, but also plays into the company's vertical integration. Absolutely. And when you talk about this being 55 % of the business, I immediately think about, okay, you have the memory chip makers, you have the processing chip makers, what has stopped them moving on to one another's turf? And I know these are incredibly complex, long cycles to develop these chips on both sides of the equation. But have there ever been risks associated with that?

20:42Or has Samsung ever tried to move into the processing world? Has NVIDIA ever tried to move into the memory world? And how are those relationships between what ultimately are compatible components that fit into the end output, but could obviously be seen as there's turf that you want to protect? Well, Samsung is one of the largest manufacturers of application processors in the world. And for their own handsets and for third -party mobile phone manufacturers, mainly the Chinese manufacturers. Apple designs its own chips, but outsources the manufacturing to TSMC, which is the largest foundry company in the world.

21:18And Samsung is the second largest manufacturer of CMOS chips for camera lenses, which are obviously used in their own cellular phones. Sony is the other large manufacturer in that business and also used for third -party cellular phones. And one of the interesting insights, people don't like to invest in hardware companies, product life cycles are pretty short. But one of the biggest insights that you can take from understanding both TSMC and Samsung is there's only a certain amount of leading edge manufacturing capability for application processors. There are two big capacity providers. One is TSMC, and most of their leading edge capacity is taken up by Apple.

22:04In other words, I alluded to this earlier, if Google wanted to grow its Pixel business very aggressively, there are only two places to go. One is TSMC, and guess what? They're sold out. And Apple might not take kindly at TSMC, taking in a big new competitive customer that's vertically integrated. And the only other place to go is Samsung. So how is Pixel going or Huawei or some of the other Chinese manufacturers going to build a big handset business at the top end of the handset market, which is where all the profit is? If you look at the Chinese manufacturers, they don't make any money. And actually, if they started paying the royalties for the IP that they use, they would be losing money.

22:47So really, the only profit pool in the handset business is shared between Apple and Samsung. And guess what? There's no way that you can compete with them simply because there is no manufacturing capacity available for you to manufacture those phones. So while others get worried and concerned about whether or not Samsung's handset business is going to be there in four or five years, I don't really worry about that that much. Whether or not it's going to grow, that's a question that you have to think about. But whether or not it's going to be there, it generates about $10 billion a year in free cash flow.

23:18I think that's a pretty steady, just slow -growing number. Very interesting dynamics across the space, just in terms of how the industry is structured. You've seen this consolidation where, to your point, there just isn't that much capacity, or the capacity is controlled by select suppliers, but also select customers when it comes to Apple. Qualcomm's another huge Foundry customer. And you see this dynamic. So one of the things that we haven't talked about at Samsung is that it has a Foundry business. It's only 10 % of the profits. So we don't talk about it that much, but it is growing 20 % per year and really is the only other viable competitor to TSMC out in that space.

23:55So companies like Qualcomm are increasingly using Samsung to manufacture their leading edge application processors and other semiconductors. And how much disparity is there in the margin profile between divisions? I know the sales numbers and just knowing the sales numbers relative to what you gave me in terms of the profit numbers, there is a pretty big gap. Maybe just take the memory business. What does that margin profile look like for them? And has that swung much just in terms of trends over time? Has it been fairly stable? Well, the manufacturer, and part of the reason why you're able, Buffett says, I'd rather have a volatile 15 rather than a steady 10.

24:38And Samsung epitomizes the volatile 15 because it's a cyclical business. So the DRAM business, towards the top of its profitability level, tends to generate margins at 35 % or 40%. And towards the bottom generates margins of 5 % to 10%. Very rarely does it lose money in DRAM. While Hynex and Micron, as they are today, will lose significant amounts of money at the bottom of the cycle. And the difference there is what I alluded to earlier, really is scale and manufacturing capability. Capacity in the business tends to be added in large increments because you want to manufacture at scale. And the business has its ups and downs.

25:19And the smaller competitors, Hynex and Micron, really only have the capital and shareholder enthusiasm behind investing in new capacity when things are very good. So everybody tends to add capacity at the same time. So you get these big increments of capacity that come in. And the lead times in this business are very long. It takes many months between the time you start manufacturing a semiconductor with this big piece of silicon to the time the actual chips come out at the end all fully packaged and ready to go. And between that and these big increments of capacity, if there are any wobbles in terms of demand, you tend to get these very big down cycles.

26:01So that's what causes these ups and downs in the business. And of course, that hurts price. So earlier, we talked about the volume growth, and volume growth has continued all the way through the downturn, but it does hurt price. In fact, the selling price for an 8 gigabit DRAM last year fell significantly from a little over $3, almost $3 .25, down to $1 .70. And NAND went from $0 .71 to $0 .39. sense. But what that does, and we've seen this happen over several cycles, is it kicks in the traditional dynamic that exists in any commodity -oriented business. When the price of a commodity falls below the marginal cost of production, that does two things.

26:46One, it induces demand, and two, it reduces supply. And let's talk again about Samsung's competitive advantage. In this case, in this downturn, in the current downturn we're going through, both Hynex and Micron are losing significant amounts of money and have had to cut back their capital spending. Samsung, on the other hand, has that handset business that generates $10 billion worth of free cash flow and carries about $100 billion worth of net cash on the balance sheet. We go back to K .H. Lee. What did he discover when he bought this business? That in a downturn, you need to be very well capitalized.

27:27And so what's happening is Samsung is investing in the next generation of manufacturing technology. These machines are called EUV. There's only a single manufacturer of those machines called ASML, ASM lithography. And interestingly, many years ago, Samsung and a few other large semiconductor manufacturers helped capitalize ASM to provide them with enough money to develop these EUV machines. And they also helped subsidize their R &D. What do they get in exchange? Well, they get the first EUV machines that come out. And these machines aren't cheap. These machines cost 170 million bucks each, roughly.

28:09Take four airplanes to move them. That's how complicated they are. And bespoke manufacturing facilities to operate them. And they're incredibly complicated to operate. And this business is all about manufacturing yields. And it takes months, if not over a year to try to figure out how to get these things to work. And as a result, going into this downturn, Samsung had started the process of building out its EUV manufacturing capability and have continued to invest all the way through the downturn. So as you come out of the downturn and customers are looking for the next leading edge of technology, smaller chips that use less power with more capability, Samsung stands to not only gain market share, but because you're selling the leading edge chips, which sell at a higher price, you will not only gain market share, but your level of profitability will be much higher.