Data Assets & Alpha Group: The AI supply chain, with Timefolio’s Singapore CIO

[Music]

Jae Lee: I look back at the late '90s when the first internet penetrated our lives. And we would have thought then how good would it be if we started to use internet to call our friends for free or to watch films for free. But then it actually took 10 years for that to happen. I think this one, we have a much earlier life cycle because we are already seeing technologies that come up with wow factors, like the AI agent that we've seen, the translating function that we've seen. Especially on the video imaging side, it's really rapidly growing. With the amount of hyperscaler spend that we see through the Asian supply chain, we view that this could really happen in the next 24 months.

Eloise Goulder: Hi. I'm Eloise Goulder, head of the Data Assets & Alpha Group here at J.P. Morgan. And today, I'm so pleased to be sitting down with Jae Lee, who is CIO and CEO of Korean hedge fund Timefolio's Singapore business and who has built up a fundamental equity long-short strategy, leveraging the TMT and the Pan-Asian supply chain over the last six years or so. Jae, thank you so much for sitting down with us today.

Jae Lee: Thank you very much for your time.

Eloise Goulder: And Jae, I'm particularly excited to be having this conversation with you because so often on this podcast series, we interview quant or systematic investors about their use of AI as a tool in the investment process. But today, we have you as a discretionary investor speaking to us all about AI as an investment theme in itself with the Asia supply chain angle, so I'm sure it will provide a really fascinating perspective. So Jae, could you start by introducing yourself and your background?

Jae Lee: Sure. Thank you very much for the invite to this podcast. I am the CEO, CIO of Timefolio Singapore. Timefolio Asset Management is the largest hedge fund in Korea headquartered in Korea. It is a fundamental equity long-short house. It's a multi-manager pod. It's been around for the last 20 years, predominantly focusing on investing in Korean markets. So I set the business up in Singapore in 2018 to manage a fund that looks at the Korea supply chain and the Taiwan supply chain to trade regional alpha in Pan-Asia space, not only in Korea. I am originally Korean. I was brought up in the UK. I first started off my career at Lehman Brothers, managing a very similar portfolio to what I'm doing now, eventually moving on to the sell side with JP Morgan London. And then I moved to Singapore with my family to set the Singapore entity up with Timefolio.

Eloise Goulder: In your role is CIO and CEO of the Singapore entity, what exactly is your mandate?

Jae Lee: So as CIO, I oversee five multi-manager pods in Singapore of which one of the portfolio is my portfolio, looking at the Korean supply chain predominantly in tech, autos, and EVs. I also have an overlay on the risk of the five pods. And I manage the central book, where I do the hedging as well. For the CEO role, my biggest mandate is to now market the fund globally and to build out the team that we have.

Eloise Goulder: And you mentioned that you're really specializing in the Korean and the Taiwan supply chain. Why specifically those two markets within Asia?

Jae Lee: It's because we know those markets the best. My Korea team are sector specialists and sharing on the ground work, the research, supply chain knowledge that they have in Korea with our Singapore team. We can use that supply chain research to generate alpha in the region.

Eloise Goulder: That makes sense. And the sectors that you mentioned, whether it was tech, or autos, or EV, do you focus on those because the Korean supply chain is so critical to those industries?

Jae Lee: Exactly. So any supply chain where the global chain is led by Korea, we are experts. But with the Korean supply chain, it's very global. So within tech, almost the entire spectrum, the higher end tech supply chain sits in Korea. Within EVs, 95% of the high-nickel batteries that support the higher end premium models on EV sit in Korea. And that space goes on with shipbuilders, with steel, with internet gaming, media content, cosmetics.

Eloise Goulder: Very interesting. Thank you. So going back to your CIO role, how exactly would you define your investment philosophy?

Jae Lee: So we are a fundamental Pan-Asia equity long-short low net, so we don't take much beta. We really focus on the alpha. We look within the supply chain for dislocations, anomalies within the supply chain predominantly in tech, in EVs, and autos where the supply chain is really global. So return on invested capital, ROE, companies that have a very good operating leverage, margin sustainability is very important when you're investing in Asia. A lot of Asian companies can grow the revenue side very, very quickly, but what really differentiates the quality companies is to have high entry barrier within their sectors. And therefore, they can sustainably return on equity.

Eloise Goulder: That makes sense, this idea that a lot of companies have that revenue growth, that top line growth, but few have the higher barriers to entry so that they have higher margins, higher ROE, and they can sustain that. So how do you go about identifying which of those companies have higher barriers to entry?

Jae Lee: So within tech, it's actually on the higher end side, for example, Korean memory or within the AI GPU supply chain, it takes five, six, seven years to build out capacity. So whether you want to break these barriers as a company, it requires a lot of time, money spent, and actual IP intelligence. The end customers, predominantly within tech, sit in the US. And the architecture of these technologies is ever-evolving. What we see in Asia as the strongest become even stronger. And the companies that have not the leading-edge technology will fall out.

Eloise Goulder: So it's clear you have to invest a lot to have that edge and then hopefully retain that edge and those margins over the longer term. And if we look at all the segments-- you mentioned technology. You mentioned EVs. In which of those areas do you find companies are benefiting from higher barriers to entry, and higher pricing power, and margins, and ROEs?

Jae Lee: Yeah. One area is on memory in your smartphones and your PCs. You need memory, otherwise, there won't be any logic within your tech hardware. Now, AI is really monetized by big GPU companies in the US. Without memory, they won't be able to have any GPUs. It's been such a consolidation within the memory space that there's only three companies in the world that can make memory of which two are Korean. And actually, the two Korean companies are maybe 70% to 80% of total DRAM market share. So that's important in terms of the higher-end spectrum. It's critical to know not only what the tier one guys are doing, but the tier three, tier four players to know the whole tech revenue stream that sits within Korea. So PCB materials, substrates, wafers, a lot of the supply chain sits within Korea. And you combine that with the knowledge that we have in Taiwan, it gives us a very good opportunity to know what the AI companies are doing and will be doing in the future.

Eloise Goulder: And can you tell me what some of these company names are, some of these large tech-focused Korean companies?

Jae Lee: Sure. So we have Samsung Electronics and SK Hynix. So these two companies are the biggest DRAM makers in the world, followed by LG Innotek, which is on the iPhone supply chain. It's very, very critical for global smartphones. And if Apple will be looking to do foldable phones, they will need to use Korean panel makers or Taiwanese hinge makers in foldable phones. We have companies like SIMCO, who do MLCC. And [INAUDIBLE] is one area where the Koreans, like Doosan, is gaining market share within the AI supply chain to provide that component to the GPUs.

Eloise Goulder: Fascinating. I mean, this question of sustainability of pricing power is arguably at the forefront of so many global investors' minds now, given the dominance of tech. And this year, for example, NVIDIA's stock price appreciation has counted for, I think, a third of the S&P 500 gains. So this question around how long these companies can sustain their market share, I think, is a critical one. Is there anything more generally on the AI theme and the dominance of these market shares that you have to say?

Jae Lee: So there's another great example of this theme, AI, which is electricity. Electricity is the biggest bottleneck of AI. We see many business leaders talk about electricity shortage. Now, where the hyperscalers are really doing massive amounts of CapEx spend in AI sit in the US. But the US infrastructure is very, very old, so they need to constantly upgrade their energy grid system to be able to provide electricity for hyperscalers to continuously spend. But they just don't have that. So where this theme becomes very interesting is within the electricity equipment makers, we have four components within that electricity, transformers, cable, wire, and also circuit breaker companies. So a few sit in Japan. Seven or eight names sit in Korea. And there's one name sitting in Taiwan. Now, because it's such a big thematic in electricity, you need every single Asian electricity transformer to be exported in the US. So you've got companies in the US traveling to Korea and saying, we would love to have all of your transformers now. We would pay a price premium. And by the way, we would want it until 2030. This is a big structural theme.

Eloise Goulder: That's incredible. And how about the tech hardware cycle? I mean, do you think that AI is going to lead to any changes there?

Jae Lee: So we've had smartphones, TVs, PCs. Replacement cycle hasn't been in the last three, four, five years. Maybe we can argue six or seven. So AI, we see text-to-video imaging to actually seeing GPT-4 zero agents. Within that spend, what we see is these AI apps, killer apps that supposedly will come in the next 24 months will be translated into your smartphones, your PCs, your TVs, which could actually create a new replacement cycle, which will be the biggest rerating story for tech hardware. But we just haven't seen that yet. AI PC is one that we are monitoring for the end of this year, but it's really down to those killer apps that are going to be developed by the big hyperscalers in the US.

Eloise Goulder: That's fascinating stuff. And I assume your point is we have to see sufficient improvements in the technology for all of us to go out and buy a new phone, a new TV, a new computer. And we haven't yet got to that tipping point. But you believe that we will get to that tipping point in the next 24 months. And that's going to be an enormous shift.

Jae Lee: Yes. So I mean, I look back at the late '90s when the internet penetrated our lives. And we would have thought then how good would it be if we started to use internet to call our friends for free or to watch films for free. But then it actually took 10 years for that to happen. But I think this one, we have a much earlier life cycle because we are already seeing technologies that come up with wow factors, like the AI agent that we've seen, the translating function that we've seen. Especially on the video imaging side, it's rapidly growing. With the amount of hyperscalers spend that we see through the Asian supply chain, this could really happen in the next 24 months.

Eloise Goulder: It's incredible when we think about what that's going to mean for all of us and the idea that we'll all have an AI agent helping us navigating the world in our everyday lives.

Jae Lee: Yeah. We might be in a world where maybe we don't need to learn a foreign language anymore, where teachers may have to change the way they teach. Maybe we need to really change the way we invest or look at things. It will change rapidly. And as long as the regulations are there to support the industry, I think this will be a real game changer.

Eloise Goulder: So going back to the subsegments that have the higher barriers to entry, where they really have invested and they're benefiting from that market share, at least today. You've mentioned Korean memory. You've mentioned electricity equipment makers. Are there any other segments where you really see higher barriers to entry at this stage?

Jae Lee: Lesser so higher entry barrier, but one real structural theme that we are very interested in now is Asian consumer products penetrating into the US. But this structural trend of US consumers, especially the younger generation, being exposed to Korean media, Korean films through avenues like Netflix-- we see some of the biggest US celebrities advertising Korean products. But these are all lower pricing products that actually have very good quality. So now they're really kicking off in the US. There's obviously inflation in grocery prices. A lot of these thematics have made US consumers try to invest cheaply. But also, when they're spending on lower pricing products, they also look for quality. And I think that's where Asian products come in. So we see revenues of certain Korean cosmetic companies growing five times every month in the US. Now, when skincare penetrates globally, it can be really sticky. The US cosmetics market is 10 times bigger than Korea, and they will have margin sustainability because these medical devices are actually quite cheap. And therefore, they are able to raise prices once they capture that audience.

Eloise Goulder: It's really interesting, and it sounds like there are various structural drivers helping these themes outperform. One of them is the fact that the ad market has become so global. And I'm thinking about Instagram, TikTok, the idea that this is very much an engagement-led and a pull-led advertising rather than a push and that you now have many Korean products penetrating, as you say, the US and the European markets. But also you mentioned inflation, cost of living crisis trading down. Presumably that is also helping for the lower price segments.

Jae Lee: Yes, absolutely.

Eloise Goulder: So we've covered those segments with higher barriers to entry, higher margins, higher ROEs. Can we turn to the opposite? You're an equity long short fund, so you obviously take shorts. So which of those businesses that actually really have low margins at this stage?

Jae Lee: Sure. So single stock short alpha is, for us, as important as our long alpha. And within the short side, we look for companies that don't have pricing power. One big example is Asian panel industry, where we've had vast amounts of oversupply, especially in the LCD industry, and therefore the margins have been compressed to low single digits. Another example is within the EV material space, especially within separators. Again, they have no pricing power. Their margins used to be double digits. And now it's very hard to forecast those margins higher than 5% And where there is overcapacity, competition is great, but they have no real edge in technology.

Eloise Goulder: Really interesting. So can we turn to some of the threats to this growth that you've mentioned? You mentioned regulatory concerns. What's the regulatory landscape going to be like for tech innovation, AI, et cetera? The other concern out there is trade barriers globally. We're obviously going through a year of multiple elections where decisions around trade could change. What are your views on these threats to the structural growth story?

Jae Lee: So trade war, import tariffs, these are big threats to all the industries that we trade in. But companies within Asia are actually taking opportunities, saying, you know what, I actually want a factory in the US now because we will get tax credits. Spend to import and export will be lower because we can just make these products in the US and sell domestically in the US. But we always look for alpha both on the long and short side. We look at dislocations within the space. So we'll always be trying to go long companies that can take advantage of this scenario versus the other ones who will lose market share.

Eloise Goulder: That makes sense. So shifting gears, I always ask our investing clients how they leverage data in the investment process. Now, there's one set of answers if you're a systematic quant investor. I'm curious from your lens, as a discretionary supply chain-led investor, how you're really leveraging data in the investment process.

Jae Lee: Sure. So a lot of our industries in Asia are export companies. And therefore, checking import/export data in respective countries and scrapping data is actually very, very important because, like you said, we first need to see upside in the top line. And therefore we are constantly checking if these products are actually selling well every week-- not only every month but every week. So it's like a bit of a puzzle matching game where you've got to constantly scrap data to support your investment case. So we really go deep down into our fundamental research in checking the supply chain. And we're checking, within tech or in autos, EVs, whether everything matches within each sector so that, ultimately, we have to translate it to, does this company have operating leverage? Does this company have margin sustainability?

Eloise Goulder: You're very much using data from a bottoms-up fundamental perspective to shape your revenue forecasts and then to shape your margin forecast and your earnings forecasts.

Jae Lee: Yes. When you have those data in hand, that's when a qualitative investing comes in where you have to talk to management. You have to talk to sell-side analysts and predict who will be right, who will be wrong and capturing the tone of the management. And that's where experience comes in.

So we are, of course, not 100% data driven, but data is the starting point. And then you have to do the qualitative work.

Eloise Goulder: And how about multiples and valuation? Growth as a factor has done so well over the last 10 plus years now. And so many companies that are demonstrating this revenue growth and strong margins, strong ROE, earnings growth have been outperforming globally. But at what price do you pay for that? So let's say you have a company that does demonstrate that strong growth, does demonstrate those high barriers to entry and that strong earnings growth, but is trading on a very, very high multiple today. How would you think about that?

Jae Lee: Valuation is one of the most important metrics that we look at. But we combine valuation with global positioning and also buyside expectations. And style matters a lot. But that's where, if you are a style biased-- and you can easily become growth biased in Asia, as growth, like you said, has worked for a long time-- then that's when you become more of a beta or a momentum fund. So if we are long growth, we like to be short the same growth in the same subsector. So we are playing the actual alpha, not the beta in that pair trade. And valuation comes to that when we're measuring whether we want to be in this growth stock or value stock. But we have to combine it with positioning and buyside expectations.

Eloise Goulder: So where positioning is long, where crowdedness is long, where buyside expectations are too high, you will want to trim that exposure, even if your fundamental growth expectations are also high.

Jae Lee: Yes. So we would be in the structural themes for two, three months. But then soon they become very crowded. And so that's when we like to take money off the table.

Eloise Goulder: And it's so interesting to hear about how you're hedging that growth exposure at a sector level. Totally makes sense that you're really trying to capture that idiosyncratic, stock-specific alpha rather than the beta or the factor. But presumably there are some periods where you really feel you're leaving something on the table.

Jae Lee: Yes, of course. So we have to hedge. We think of hedging as a car insurance. So when you're paying car insurance, you know that you have to pay a certain price every month, which is the same thing as you're leaving that on the table. And especially in Asia, where it's such a hard environment to trade, especially with severe macro conditions-- and that's what we've seen in the last two, three years-- if you're not hedging and you're not properly doing factor adjustments when you're hedging, it means that your performance will be very, very volatile. But there are years like last year and this year where, OK, we are hedged from growth to long growth to short growth. But we actually size up on the conviction idea.

Eloise Goulder: Yes, it's a great point. So while you're hedging a lot of that factor risk, you can somewhat compensate for that by really sizing up your higher conviction ideas. So you do have a lot of weight and a lot of exposure to those segments where you've done your research. So when I think about your investment philosophy, from what I'm hearing, you have a lot of regional, local expertise in these markets. And I would assume barriers to entry to have that level of domain-specific local knowledge are pretty high, and that's what gives you your edge. How do you think about that? And what's the competition for this sort of alpha in your space?

Jae Lee: That's a great question. So combining Korea to Taiwan and Japan is very hard, especially on the Korean side, because it's a very local market. And you need a very big team. And not only do you need the size of the team to be sitting in Korea, but you need that team to have possibly worked in that same space for the last 10, 20 years. So we have that with Timefolio Korea with their experience for the last 20 years. But what's also important is we also have a alternative investment team, a pre-IPO team that actually monitor companies that are not listed as well. And to have that local knowledge of knowing the companies even before their listing is critical. And that's where our supply chain checks really begin with when we are checking the tier 3, tier 4 unlisted supply chain. That's how we get the first edge.

Eloise Goulder: And that pre-IPO team, I'd imagine they create value in a number of ways. Partly, if those companies then go on to list, you can participate. But even if they didn't, it gives you that sense of the competitive landscape that might help you to analyze the listed peers.

Jae Lee: Absolutely.

Eloise Goulder: Thank you so much, Jae. This has been such an insightful discussion. And as I said at the start, we so often interview quant investors about their use of AI in the investment process on this podcast. But this time round, you've told us all about AI and the edge you have within Asian supply chains around AI as an investment theme. So we very much turned our typical conversation on its head, and I've learned so much in the process. And I hope our listeners will too. So thank you very much for taking the time to speak with us today.

Jae Lee: Thank you very much. Thank you.

Eloise Goulder: Thank you also to our listeners for tuning in to this bi-weekly podcast series from our group. If you'd like to learn more about Jae's work and Timefolio, then please do reach out to your JP Morgan sales representative. Otherwise, if you have feedback or if you'd like to get in touch, then please do go to our website at jpmorgan.com/market data intelligence, where you can reach out via the Contact Us form. And with that, we'll close. Thank you.

[End of episode]