WEBVTT - How to Invest in Asia as the AI Trade Gets More Volatile 0:00:02.720 --> 0:00:09.560 Bloomberg Audio Studios, podcasts, radio news. Hello Marin Talks Money Listeners, 0:00:09.600 --> 0:00:12.240 it is that time of year. The Edinburgh Festival Fringe 0:00:12.280 --> 0:00:14.960 is almost here and I will once again be recording 0:00:14.960 --> 0:00:18.480 podcast episodes in front of a live audience from Panmea House, 0:00:18.800 --> 0:00:22.200 the final home of Adam Smith. This year's shows will 0:00:22.200 --> 0:00:25.120 take place on August twenty seventh, twenty eighth, and twenty 0:00:25.200 --> 0:00:28.160 ninth at two pm. Get your tickets before it sells out. 0:00:28.240 --> 0:00:41.360 The link is in the show notes. Welcome to Marin 0:00:41.360 --> 0:00:43.360 Talks Money, the podcast in which people who know the 0:00:43.400 --> 0:00:46.199 markets explain the markets. I am Maren Semsep Web and 0:00:46.240 --> 0:00:48.960 this week I'm speaking with Fioni Yang, a fund manager 0:00:49.040 --> 0:00:52.120 at Invesco. She is the manager of the Invesco Asia 0:00:52.240 --> 0:00:55.720 Dragon Investment Trust. On today's show, we talk about AI. 0:00:55.800 --> 0:00:59.320 We talked about career volatile market, very AI exposed. We 0:00:59.400 --> 0:01:02.200 talk about India Cheap or not Cheap and why has 0:01:02.240 --> 0:01:05.040 it lost so much popularity? And we talked for probably 0:01:05.040 --> 0:01:08.679 the first time on the show about the Australian stock market. Fiona, 0:01:08.840 --> 0:01:11.840 Welcome to Merin Talks Money. Thank you for having me 0:01:11.880 --> 0:01:15.280 here again. You were in earlier this year and my goodness, 0:01:15.280 --> 0:01:17.280 a lot has happened since and at the time we 0:01:17.319 --> 0:01:18.800 talked about all sorts of things, about one of the 0:01:18.840 --> 0:01:21.400 big topics that we discussed last time. And by the way, listeners, 0:01:21.400 --> 0:01:23.400 if you haven't listened to the last one, maybe go 0:01:23.480 --> 0:01:25.520 listen to that one, either before or after. This one 0:01:25.560 --> 0:01:28.280 has had lots of interesting stuff in it. I wanted 0:01:28.400 --> 0:01:33.160 to start by talking about what's happened in Korea because 0:01:33.200 --> 0:01:35.160 it was already growing great guns the last time we 0:01:35.240 --> 0:01:37.800 talked to continue to saw and saw and saw based 0:01:37.800 --> 0:01:40.240 on a couple of major AI related stocks, and then 0:01:40.280 --> 0:01:42.319 we had a little bit of a stumble, and now 0:01:42.400 --> 0:01:44.920 career is I was going to say twenty five percent 0:01:44.959 --> 0:01:47.400 of off its peak in June, but on the day 0:01:47.400 --> 0:01:49.920 that we're talking, I think that Cosby went up seventy 0:01:49.960 --> 0:01:53.600 percent even yesterday. So this is very volatile stuff. And 0:01:53.680 --> 0:01:57.919 we're also hearing about a lot of very highly leveraged 0:01:58.120 --> 0:02:01.320 ordinary career investors or amblers, I guess we should say, 0:02:01.360 --> 0:02:04.480 who've lost an awful lot of money by having very 0:02:04.560 --> 0:02:07.920 high margin debt and then being hit by falling share prices. 0:02:07.920 --> 0:02:11.200 So career is fascinating because it's exactly as you said 0:02:11.480 --> 0:02:15.360 that front row seat to AI, but it's also turned 0:02:15.400 --> 0:02:16.560 out to be very volatile. 0:02:17.720 --> 0:02:21.200 Yeah, exactly. That's a fantastic observation that you have there. 0:02:21.320 --> 0:02:23.520 On the career market, I think they take a step 0:02:23.560 --> 0:02:26.680 back quite a few things has worked out very favorably 0:02:26.760 --> 0:02:30.360 for career over the past twelve eighteen months. The memory 0:02:30.400 --> 0:02:34.760 cycle boosted by AI demand is just phenomenal. This is 0:02:34.800 --> 0:02:37.640 one of the biggest supercycle that we have seen for 0:02:37.960 --> 0:02:40.640 stocks like some se Electronics and Eessacheinis. 0:02:40.960 --> 0:02:43.040 And on top of that, I think we had to 0:02:43.320 --> 0:02:43.840 say the. 0:02:43.880 --> 0:02:48.760 Korean government has done a fantastic job in boosting shareholder 0:02:48.800 --> 0:02:52.760 return doing the value up initiative in Korea. It's a 0:02:52.760 --> 0:02:55.959 textbook sort of example that all the other countries should 0:02:56.000 --> 0:02:58.800 really look up to for some of the holdings that 0:02:58.840 --> 0:03:02.040 we have in the portfolio, so especially the financials. They 0:03:02.040 --> 0:03:05.600 are boosting shareholder return massively over the past few years, 0:03:05.919 --> 0:03:09.760 and they start to really focus on shareholder interests, protecting 0:03:09.840 --> 0:03:14.400 minority shareholders interests, and giving us more voices in the boardroom. 0:03:14.639 --> 0:03:18.160 That's something that I think in all this data center 0:03:18.320 --> 0:03:23.040 or AI narrative that investors have forgotten. But I do 0:03:23.120 --> 0:03:25.520 see risks. There are two risks I see in the 0:03:25.560 --> 0:03:28.560 career market. The first one is the concentration risk, as 0:03:28.560 --> 0:03:33.440 you correctly pointed out, with the participation of leveraged ETF 0:03:33.880 --> 0:03:37.000 I do think some of the retail investors might sort 0:03:37.080 --> 0:03:39.120 they just bought into a product that give them a 0:03:39.120 --> 0:03:43.040 bit more return into this investing into a great company. 0:03:43.320 --> 0:03:46.680 But reality is it's a much more complex financial product 0:03:46.720 --> 0:03:51.200 they got themselves into and that led to the capital losses. 0:03:51.800 --> 0:03:55.080 And then the other risk is really earnings risk. Now 0:03:55.200 --> 0:04:00.000 the market is having super high expectations bought the memory pride, 0:04:00.040 --> 0:04:02.000 it's going to be if you look at some Sound 0:04:02.000 --> 0:04:07.920 Electronics two q preliminary result that delivered nineteen percent nineteen 0:04:08.040 --> 0:04:11.960 times increase versus two q last year for share price 0:04:12.040 --> 0:04:15.560 is actually done post that result exactly because of this 0:04:15.720 --> 0:04:20.680 expectation versus reality. Expectation of earning scrolls is super high 0:04:20.760 --> 0:04:24.279 to the degree that even fantastic earning cycle cannot that 0:04:24.440 --> 0:04:28.760 meet that expectation. That also result in share price volatility 0:04:28.839 --> 0:04:30.080 that we have seen from the peak. 0:04:31.160 --> 0:04:31.560 Interesting. 0:04:31.680 --> 0:04:33.400 I mean everything you say about the changes in the 0:04:33.440 --> 0:04:36.560 Korean market and the changes to shareholder democracy and the 0:04:36.600 --> 0:04:39.480 changes to regulation driven corporate governance, that's what we saw 0:04:39.520 --> 0:04:41.760 in Japan in a previous cycle, right, and I still 0:04:41.760 --> 0:04:45.320 think ongoing in Japan, which is interesting. So Korean regulator 0:04:45.480 --> 0:04:47.400 picked up all sorts of hints from Japan and that 0:04:47.480 --> 0:04:51.120 makes a great long term story for shareholders in the 0:04:51.200 --> 0:04:54.480 Korean market, but nonetheless in the short term, as you say, 0:04:54.520 --> 0:04:57.600 it is all about concentration. Let's get high next and 0:04:57.800 --> 0:05:00.280 Sam saying are a huge part of the market, and 0:05:00.320 --> 0:05:03.279 it's the same across the Asian market. Actually, if you 0:05:03.320 --> 0:05:07.279 think to yourself, I'm buying emerging Asia or I'm buying Asia, 0:05:07.320 --> 0:05:09.800 and you buy any kind of index or ETF product, 0:05:10.040 --> 0:05:13.640 you're actually just buying huge exposure to a couple of 0:05:13.680 --> 0:05:16.960 AI stocks. So that concentration is the thing that really 0:05:17.279 --> 0:05:20.560 matters if you're buying into any kind of index product, 0:05:20.640 --> 0:05:23.800 which means that you really have to look in detail 0:05:23.880 --> 0:05:25.760 at the second risk would you bring up there, which 0:05:25.839 --> 0:05:28.800 is the supercycle and memory? And when you refer to 0:05:28.839 --> 0:05:32.840 it as a supercycle, I wonder what you mean by 0:05:32.880 --> 0:05:35.799 that and how long you think that cycle will last, 0:05:35.880 --> 0:05:38.920 because I think for old people, old investors, people who've 0:05:38.920 --> 0:05:41.480 seen these cycles go on forever and ever. One of 0:05:41.520 --> 0:05:43.440 the first things I ever learned when I became a 0:05:43.480 --> 0:05:46.760 stockbroker in Asia, was that these cycles tend to be 0:05:46.880 --> 0:05:48.080 shorter than you think. 0:05:48.360 --> 0:05:50.640 Maybe it's best just to take a step back to 0:05:50.720 --> 0:05:54.640 look at how we ended up here, rewind the tape, like. 0:05:54.600 --> 0:05:55.480 Two three years ago. 0:05:55.800 --> 0:05:59.000 What happened to the memory industry was actually we were 0:05:59.080 --> 0:06:01.799 in a dumb cycle. If you look at NONE, which 0:06:01.839 --> 0:06:05.680 is sort of non flash, that is in deep loss making. 0:06:06.000 --> 0:06:08.960 As a result, even though the industry only have a 0:06:09.000 --> 0:06:13.040 few producers of non none of them was really expanding 0:06:13.040 --> 0:06:16.800 the capacity aggressively. Similar for d RAM, which is the 0:06:16.839 --> 0:06:20.080 other type of memory that goes into all these AI 0:06:20.160 --> 0:06:24.160 products these days, same they have gone through this commodity 0:06:24.360 --> 0:06:28.080 super up and down cycle in the priials, the data center, 0:06:28.200 --> 0:06:33.240 cloud built out era, and the profitability has deteriorated very quickly. 0:06:33.640 --> 0:06:36.159 For a company like eskaheanis they have to really go 0:06:36.240 --> 0:06:40.120 to the market to raise comfortable bonds because the balance 0:06:40.160 --> 0:06:43.160 sheet was in deep trouble. They were super leveraged, so 0:06:43.200 --> 0:06:46.400 they were not in a position to expand capex. All 0:06:46.440 --> 0:06:49.760 of this moving to where we are today is we 0:06:50.200 --> 0:06:54.240 had this period of under investment by the major players 0:06:54.279 --> 0:06:57.719 in the memory markets, and on the other side of 0:06:57.839 --> 0:07:01.599 demand we suddenly have AI agentic AI that just need 0:07:01.720 --> 0:07:04.560 a lot of memory to really record the conversation that 0:07:04.600 --> 0:07:08.720 you had with CHGBT, with Anthropic, with Claude, so they 0:07:08.720 --> 0:07:11.040 can help you to do the dust better from this 0:07:11.160 --> 0:07:16.480 point on. So that demand search is unexpected from what 0:07:16.880 --> 0:07:21.480 all these investors as well as memory producers perspective, even. 0:07:21.280 --> 0:07:22.080 Twel months back. 0:07:22.320 --> 0:07:25.480 That's why we ended up with this shortaging supply because 0:07:25.560 --> 0:07:28.600 under investment over the past couple of years as well 0:07:28.640 --> 0:07:30.640 as this sudden search in demand. 0:07:30.960 --> 0:07:32.640 So what has happened. 0:07:32.400 --> 0:07:36.840 Is really this shortage has just driven the prices of 0:07:37.000 --> 0:07:40.840 the exactly the same products up phenomenally over the past 0:07:40.920 --> 0:07:41.520 few years. 0:07:41.760 --> 0:07:44.560 Then the question is about when how long this cycle 0:07:44.680 --> 0:07:45.280 could last. 0:07:45.680 --> 0:07:48.080 I think it could last for a couple of years 0:07:48.400 --> 0:07:53.000 because of this supply to come up will take time. 0:07:53.080 --> 0:07:55.800 You need to build the clean room, you need to 0:07:56.040 --> 0:07:59.240 order the equipment and put the equipment in there. But 0:07:59.320 --> 0:08:01.560 then does that being the share price could last for 0:08:01.840 --> 0:08:05.080 two three years at the very high level, that might 0:08:05.120 --> 0:08:08.360 not lessarily be true because the expectation is already up there. 0:08:08.560 --> 0:08:12.120 Stock market is about expectation versus reality. The expectation is 0:08:12.160 --> 0:08:15.480 already there's a sort of an Ai demand that cannot 0:08:15.480 --> 0:08:17.800 be mad and the supply will not come up in 0:08:17.840 --> 0:08:20.320 the next two to three years. If we do see 0:08:20.360 --> 0:08:23.640 any sort of vacuum in the demand side of things, 0:08:24.120 --> 0:08:27.360 or supply could come up just faster than expected because 0:08:27.400 --> 0:08:30.520 suddenly all these companies have raised huge amount of money, 0:08:30.680 --> 0:08:33.040 not just for the Korean ones, but also the Chinese 0:08:33.080 --> 0:08:37.480 competitor is doing an IPO in the coming days. All 0:08:37.520 --> 0:08:40.520 of that could change the supply dynamics quite a bit. 0:08:41.120 --> 0:08:44.720 Then it's about when does that price or the customers. 0:08:44.800 --> 0:08:47.120 By the other day, you have a handful of hyperscalers 0:08:47.120 --> 0:08:49.440 paying for all these products. If they do see demand 0:08:49.559 --> 0:08:53.480 weakening slightly and the supply going to come up quite massively, 0:08:54.000 --> 0:08:56.960 the bargaining tip going to go back from the memory 0:08:57.000 --> 0:09:02.000 producers to the payers, and we could see prices start 0:09:02.120 --> 0:09:03.520 to fall from this point on. 0:09:03.840 --> 0:09:07.160 Yeah, let's talk about demand. Then the supply conversation is 0:09:07.280 --> 0:09:09.440 slightly easier in a way, isn't it, because you can 0:09:09.480 --> 0:09:11.640 see the timeline that's required, as you say, to build 0:09:11.640 --> 0:09:14.160 the clean rooms, et cetera. But when you look at demand, 0:09:14.200 --> 0:09:19.839 the assumptions for rising demand are pretty pretty enormous, pretty enormous. 0:09:20.520 --> 0:09:22.720 There are all sorts of things that could affect that, 0:09:22.800 --> 0:09:24.760 one of those being people changing the way that they 0:09:24.880 --> 0:09:28.560 use llms at the moment, people changing from using the 0:09:28.600 --> 0:09:32.280 services provided by the big American hyperscalis to using Chinese products, 0:09:32.280 --> 0:09:35.440 et cetera. So there's lots of things that could shift 0:09:35.679 --> 0:09:36.920 inside the demand equation. 0:09:38.040 --> 0:09:41.640 There are quite a few moving parts in that demand equation. 0:09:42.320 --> 0:09:45.280 How I see. I think demand is here to stay 0:09:45.520 --> 0:09:48.520 is just a question of how fast it's going to grow. 0:09:49.000 --> 0:09:52.440 The market expectation is no one really has a crystal 0:09:52.440 --> 0:09:54.480 ball of what's going to happen in the future, but 0:09:54.600 --> 0:09:57.480 the narrative in the market is that demand are going 0:09:57.520 --> 0:10:00.600 to grow at an exceptional pace. I do see a 0:10:00.640 --> 0:10:03.679 couple of risks in there. The first one is really 0:10:03.720 --> 0:10:09.720 the return on those Hyperscaler's investment. It is really, given 0:10:09.760 --> 0:10:12.760 the current sort of annual round rate of the revenue 0:10:13.120 --> 0:10:17.040 produced by open AI and Tropic and not to mention 0:10:18.120 --> 0:10:23.080 on certain paths towards profitability, the ROI of this Hyperscalar 0:10:23.120 --> 0:10:28.240 investment is still questionable. On this huge amount of capex, 0:10:28.320 --> 0:10:30.880 we're going to see all we expect to see, So 0:10:30.920 --> 0:10:34.360 that's one risk. The other risk is competition. As you 0:10:34.440 --> 0:10:38.520 pointed out, the Chinese large language model viewed up on 0:10:38.720 --> 0:10:46.120 cheap power, cheap electricity, cheap land, and cheap hardware producers 0:10:46.520 --> 0:10:50.720 because China is yeah, the technology could be generators behind, 0:10:51.080 --> 0:10:55.320 but once they get to a certain technology level, the 0:10:55.440 --> 0:10:58.439 cost of producing those products is just significantly cheaper as 0:10:58.480 --> 0:11:01.800 compared to the developed market. So as the result, their 0:11:01.920 --> 0:11:04.640 large language model is actually prized at a fraction of 0:11:04.679 --> 0:11:07.400 the cost of the leading edge model that we have 0:11:07.440 --> 0:11:09.920 seen in the Western world. I was just having lunch 0:11:09.960 --> 0:11:12.720 with one of my friend who operates one of his 0:11:12.840 --> 0:11:16.760 own software company in the Asia region. He talks about, yes, 0:11:16.800 --> 0:11:19.400 we just switched between the models. Certain tasks, we can 0:11:19.559 --> 0:11:23.440 just use the ninety percent cheaper Chinese model, and certain 0:11:23.480 --> 0:11:27.320 tasks the most advanced masks, maybe we use the Western model, 0:11:27.440 --> 0:11:30.080 and that's a perfect way out for us to both 0:11:30.120 --> 0:11:33.440 save one cost and really get to the desired outcome 0:11:33.600 --> 0:11:37.400 for coding and etc. So that just there's two risks 0:11:37.440 --> 0:11:41.040 that I have taken to consideration thinking about the demand 0:11:41.280 --> 0:11:44.160 and how quickly that demand going to grow from this 0:11:44.240 --> 0:11:44.559 point on. 0:11:44.920 --> 0:11:46.880 Yeah, see, that's interesting. So a lot of the demand 0:11:46.880 --> 0:11:49.200 assumption to assume that everyone will always want to use 0:11:49.240 --> 0:11:51.599 the best model, but that's not really the case. Is 0:11:51.640 --> 0:11:54.400 that people will use the good enough model depending on press, 0:11:54.920 --> 0:11:58.719 depending on task. Yeah. Okay, So if you are now 0:11:58.760 --> 0:12:01.960 investing in this region and you're looking at this concentration, 0:12:02.080 --> 0:12:05.320 and you're looking at these risks that we've just talked about, 0:12:05.600 --> 0:12:07.640 and I know that a lot of people will say 0:12:07.720 --> 0:12:10.600 to you that the big companies, the Samsung's and the 0:12:10.840 --> 0:12:13.920 high Next, etcetera, that they still look reasonably cheap. But 0:12:14.360 --> 0:12:17.800 of course that's the way cycles always work. That you're 0:12:17.840 --> 0:12:20.200 supposed to sell these companies when they look cheap and 0:12:20.200 --> 0:12:22.000 buy them when they look expensive, because that's how the 0:12:22.000 --> 0:12:25.560 cycle works. So if you're an active investor, now is 0:12:25.600 --> 0:12:29.360 the time to diversify away from those peer plays, is 0:12:29.400 --> 0:12:32.959 my guess, and start looking at companies that are not 0:12:33.000 --> 0:12:35.160 so much the picks and shovels, but the companies that 0:12:35.240 --> 0:12:38.480 will benefit in other ways from using AI. If you're 0:12:38.520 --> 0:12:40.640 still going to talk about this as an AI trade, 0:12:40.760 --> 0:12:42.600 start looking at the companies that will be able to 0:12:42.679 --> 0:12:45.840 use AI to improve their productivity, their customer service, etc. 0:12:46.160 --> 0:12:48.920 So you're moving out from those peer plays. 0:12:49.240 --> 0:12:53.040 It is exactly you know for active bottom up farm manager. 0:12:53.120 --> 0:12:55.640 That's my job day to day. So two angles. I 0:12:55.840 --> 0:12:56.320 look at this. 0:12:56.920 --> 0:13:00.720 One is really outside of the AI stuff in Korea. 0:13:01.200 --> 0:13:04.199 If you are SMSUM employee, you get this huge bonus 0:13:04.240 --> 0:13:06.680 because you just negotiated to get ten percent of the 0:13:06.679 --> 0:13:10.000 company's op to operating profit to be paid in bonus. 0:13:10.120 --> 0:13:11.440 What are you going to do with it? You're going 0:13:11.480 --> 0:13:12.040 to buy a house? 0:13:12.080 --> 0:13:13.320 Can we just stop on that? What do we get 0:13:13.320 --> 0:13:15.079 people getting in the way of bonuses? And I've read 0:13:15.120 --> 0:13:17.240 about this one hundreds and hundreds of thousands of dollars 0:13:17.280 --> 0:13:19.960 being paid out and bonuses to ordinary employees. What's the 0:13:20.000 --> 0:13:20.720 scale here? 0:13:21.200 --> 0:13:23.840 This is going to be massive people the sale side. 0:13:23.880 --> 0:13:28.079 The most foolish expectation is next year, as Kahnis could pay, 0:13:28.679 --> 0:13:32.000 could get to about four hundred, four hundred and fifty 0:13:32.760 --> 0:13:36.640 trillion Korean one. If you have ten percent of that 0:13:36.960 --> 0:13:39.680 paid to the employees, it is going to be huge 0:13:39.679 --> 0:13:41.360 as a percentage of the country's GDP. 0:13:42.080 --> 0:13:44.480 Of course that's expectation. I'm not saying we will get there. 0:13:44.960 --> 0:13:48.320 But with that huge amount of money, then people will 0:13:48.360 --> 0:13:51.000 want to spend it. They want to buy house, they 0:13:51.000 --> 0:13:54.800 want to buy handbags. Actually, it's already happening. People talking 0:13:54.800 --> 0:13:59.679 about the incestuable demand for luxury handbags as well as jewelry. 0:14:00.120 --> 0:14:03.440 In Korea, the department store is just packed. If you 0:14:03.559 --> 0:14:05.720 go to the VCA in the morning, you have to 0:14:05.800 --> 0:14:09.280 get a queue ticket the first moment it opens, otherwise 0:14:09.360 --> 0:14:10.880 you will not be able to visit it for the 0:14:10.920 --> 0:14:11.360 whole day. 0:14:11.559 --> 0:14:15.720 You have to cute. You have too cute store to 0:14:15.840 --> 0:14:19.200 go to one of the jewelry shops. Yeah right, everyone 0:14:19.240 --> 0:14:20.120 should come to Edinburgh. 0:14:20.160 --> 0:14:22.120 There's no queues outside our luxury shops. 0:14:23.160 --> 0:14:25.440 I think they're gonna do that absolutely. 0:14:25.800 --> 0:14:28.880 But your holidays here I needs employees were waiting. 0:14:30.480 --> 0:14:32.920 So those are the things that we expect to see. 0:14:32.960 --> 0:14:35.240 We already started to see signs of that. But once 0:14:35.240 --> 0:14:37.760 that money get into the pocket of the employees, that's 0:14:37.840 --> 0:14:40.160 exactly what they're gonna do. They're gonna go overseas to 0:14:40.240 --> 0:14:42.680 travel as well, so we want to benefit from that. 0:14:42.800 --> 0:14:46.040 I think the banks it's going to see acceleration in 0:14:46.120 --> 0:14:49.600 long world the welsh mana in product demand for introduced 0:14:49.600 --> 0:14:52.240 by the banks. And then also on the other side 0:14:52.320 --> 0:14:55.400 is Peppex gonna benefit a lot of the construction company 0:14:56.200 --> 0:15:00.040 in the country. So that's around within the career, the 0:15:00.080 --> 0:15:04.960 secondary and tertiary beneficiary from the AI trend, and then 0:15:05.000 --> 0:15:07.240 outside of that is the point of RAI is about 0:15:07.360 --> 0:15:11.440 who're going to use this AI technology to really benefit 0:15:11.560 --> 0:15:15.760 structurally for the future. So some of the winners that 0:15:15.800 --> 0:15:19.000 we identify actually the Internet space. We like some of 0:15:19.040 --> 0:15:23.160 the game developers in China or in the region. If 0:15:23.160 --> 0:15:25.680 you think about online games, how you develop in the past, 0:15:25.760 --> 0:15:28.680 you spend millions of dollars trying to produce the content 0:15:28.800 --> 0:15:32.840 to make the AGI looks absolutely amazing. That comes at 0:15:32.840 --> 0:15:37.200 the expense of hiring thousands of programmers graphic designers. But 0:15:37.280 --> 0:15:39.840 now using AI, you can just make it so much simpler. 0:15:40.000 --> 0:15:43.800 It saves the costs and really enhance the gameplay and 0:15:43.840 --> 0:15:47.120 you can even tailor made the products to every individual 0:15:47.200 --> 0:15:50.400 user have a completely different gameplay versus a guy sitting 0:15:50.440 --> 0:15:52.400 next to me because we just like different things. We 0:15:52.440 --> 0:15:56.280 want different experience in the virtual world. That just increased 0:15:56.520 --> 0:16:00.200 time span on the platform and lower the cost some 0:16:00.280 --> 0:16:02.760 of the stocks that we have held in the portfolio 0:16:02.800 --> 0:16:07.840 for the longest time. It's just underappreciated by this AI beneficiary. 0:16:08.080 --> 0:16:11.120 Thematic Can you give us the names and tell us 0:16:11.120 --> 0:16:12.760 some of the details of some of the other stocks 0:16:12.800 --> 0:16:16.120 you're holding in Korea that have exposure to these themes. 0:16:16.600 --> 0:16:19.560 Right, So in Korea we own some of the banks 0:16:19.560 --> 0:16:21.640 like k BE Financial is one of the biggest bank 0:16:21.840 --> 0:16:25.000 in the region. We also own some Sound Farmering. It's 0:16:25.040 --> 0:16:29.080 one of the insurance company that has just seen shareholder 0:16:29.120 --> 0:16:33.720 return being boosted and into the capexpanding cycle of the 0:16:33.760 --> 0:16:36.800 Samsung Group. We do own some sound na It is 0:16:36.840 --> 0:16:39.240 one of the captives sort of construction company for the 0:16:39.240 --> 0:16:43.560 Samsung Group that could continue to cater to the company's 0:16:43.600 --> 0:16:47.320 future buildout of their clingroom in the country. And then 0:16:47.360 --> 0:16:51.200 for the other theme about the beneficiary of AI using 0:16:51.240 --> 0:16:54.600 AI technology, we do net Eas and Tencent in China. 0:16:54.720 --> 0:16:57.440 They are the game producers in China, and because of 0:16:57.480 --> 0:17:01.000 the regulation that you need to really pup games in China, 0:17:01.160 --> 0:17:04.320 it's not something that AI can't actually do for you 0:17:04.440 --> 0:17:08.560 to really negotiate with the regulator, what sort of games 0:17:08.560 --> 0:17:11.359 can be brought to the market that still needs the 0:17:11.880 --> 0:17:14.960 humans and humans, yes exactly. 0:17:35.400 --> 0:17:39.600 That much of Asia is now beginning to look relatively expensive. 0:17:39.640 --> 0:17:42.200 So we've talked about career and I know Japan is 0:17:42.200 --> 0:17:44.400 a new area, but that's not the pure value play. 0:17:44.880 --> 0:17:47.440 It used to be India, welcome onto India in a 0:17:47.440 --> 0:17:51.000 little bit, but expensive and some Vietnam, for example, used 0:17:51.040 --> 0:17:54.400 to be so cheap, isn't it anymore? But China remains 0:17:54.520 --> 0:17:58.879 pretty inexpensive, still a very unpopular market. Tell me about 0:17:58.880 --> 0:18:00.720 that from point of view. 0:18:00.720 --> 0:18:04.440 Yeah, I think unfortunately, China has just been used as 0:18:04.440 --> 0:18:08.800 a funding source to buying into the AI hardware space. 0:18:09.160 --> 0:18:13.800 So the Chineses have many of the popular holdings among 0:18:13.920 --> 0:18:17.639 Asia and emergeny market investors. When you think about investing 0:18:17.680 --> 0:18:21.119 the region, you usually own stocks in China because they 0:18:21.119 --> 0:18:24.600 are perceived as the best. And now with the funding 0:18:24.720 --> 0:18:28.440 flow coming out of these popular holdings and going into 0:18:28.520 --> 0:18:33.760 Korea and Taiwan, is just have depressed that market valuation. 0:18:33.920 --> 0:18:36.600 So people have just been selling their holdings in Chinese 0:18:36.600 --> 0:18:39.280 markets to buy into the exciting stuff and career in 0:18:39.320 --> 0:18:40.760 Taiwan exactly. 0:18:40.800 --> 0:18:43.640 I think that's definitely play an important role in that. 0:18:43.920 --> 0:18:47.720 And if we look at the index composition of MSCI China, 0:18:48.080 --> 0:18:51.160 there are a lot of Internet names in there, and 0:18:51.480 --> 0:18:55.399 many would argue for internet companies, maybe you need to 0:18:55.560 --> 0:18:59.960 reinvest yourself, reinvent yourself to have a new business model 0:19:00.280 --> 0:19:04.359 to really survive in the new ai era. Maybe you 0:19:04.400 --> 0:19:07.320 want to increase your capech spending, you need to have 0:19:07.359 --> 0:19:12.119 the most competitive large language model so as to wing. Really, 0:19:12.640 --> 0:19:16.280 as a result, it's not sociue whether those companies are 0:19:16.359 --> 0:19:21.400 still the leaders in the region, so they get sold 0:19:21.440 --> 0:19:24.000 off because the FUM flow and the question mark of 0:19:24.080 --> 0:19:27.640 how they're going to evolve themselves in this new technology world. 0:19:28.080 --> 0:19:30.399 Yeah, but evaluations reflect that this is one of the 0:19:30.400 --> 0:19:33.800 few places in the region where you can still say 0:19:34.160 --> 0:19:35.359 definitely not paying. 0:19:36.320 --> 0:19:39.119 As the contrary, investors, when other people look away, we 0:19:39.200 --> 0:19:41.399 always want to look more into it because there are 0:19:41.440 --> 0:19:44.639 a lot of gems in that market, and investors are 0:19:44.760 --> 0:19:47.680 fixated on a lot of the macro issues about weak 0:19:47.760 --> 0:19:54.520 economic growth or still underwhelming consumption data. But you know what, 0:19:54.920 --> 0:19:58.920 there are companies that can actually thrive in this macro environment. 0:20:00.119 --> 0:20:03.320 Extraordinary, isn't it What a small part of the indices 0:20:03.359 --> 0:20:04.760 the Chinese market takes up. 0:20:05.680 --> 0:20:06.679 Definitely, Definitely. 0:20:06.760 --> 0:20:09.080 I think it's more of the case of how quickly 0:20:09.280 --> 0:20:12.640 Taiwan and the Korea have gone up, and then the 0:20:12.680 --> 0:20:16.320 market cap of this mega company, the concentration that they 0:20:16.400 --> 0:20:19.440 have in the Asia in decks is mind blowing. 0:20:20.240 --> 0:20:23.160 Yeah, I'm back to concentration and be careful what you buy, right, 0:20:23.359 --> 0:20:26.280 all right, let's look at India, which was incredibly popular 0:20:26.359 --> 0:20:29.920 level age as everyone loved India and we were constantly 0:20:29.960 --> 0:20:31.840 being told you didn't need to worry about price, just 0:20:32.040 --> 0:20:34.880 by India because of the growth and the exciting companies, 0:20:34.920 --> 0:20:38.160 et cetera. And that's over. No one's died in India anymore, 0:20:38.320 --> 0:20:41.080 fallen out of favor. But you've recently taken a trip 0:20:41.080 --> 0:20:42.000 to India. 0:20:42.320 --> 0:20:45.320 Yes, so India is a very interesting one. 0:20:45.640 --> 0:20:47.960 Two years ago, it was one of the most popular 0:20:48.359 --> 0:20:51.959 country to investing for the whole region, and here we 0:20:52.000 --> 0:20:55.119 are it is completely being put into the bottom of 0:20:55.160 --> 0:20:56.800 the jaw because it doesn't. 0:20:56.520 --> 0:20:59.719 Really AI stocks in there all. 0:20:59.760 --> 0:21:03.320 The Internet company again similar like the Chinese ones, being 0:21:03.400 --> 0:21:06.679 questioned about their survivability for the long term and the 0:21:06.800 --> 0:21:09.120 growth in the long term. So I just thought it's 0:21:09.119 --> 0:21:11.600 an interesting time to definitely take a trip and visit 0:21:11.640 --> 0:21:15.800 our portfolio companies and identify whether they are more interesting ideas. 0:21:16.119 --> 0:21:19.040 I think the conclusion is still the market, despite on 0:21:19.080 --> 0:21:23.280 the performance versus everywhere else, is still expensive. 0:21:24.160 --> 0:21:25.720 Unfortunately, is. 0:21:27.200 --> 0:21:30.680 Valuation is something that we can't get the hurd or over, 0:21:30.840 --> 0:21:33.240 but doesn't mean we can't own stocks in the market 0:21:33.280 --> 0:21:36.640 because they're still selected parts of the market looks extremely 0:21:37.240 --> 0:21:40.000 valuable for us and if you look at the growth 0:21:40.280 --> 0:21:45.280 compounding power of some of the portfolio company is definitely impressive. 0:21:45.800 --> 0:21:48.760 So one of my favorite in India and during the 0:21:48.800 --> 0:21:52.360 trip is Truan. This is a non bank financial company 0:21:52.800 --> 0:21:56.520 operating India for extended period of time. They basically serve 0:21:56.640 --> 0:21:59.439 to say, the truck drivers helping them to get a 0:21:59.480 --> 0:22:02.240 second give them a loan to get a secondhand truck. 0:22:02.560 --> 0:22:05.800 They also help the small businesses giving them some working capital, 0:22:05.880 --> 0:22:08.879 loan and etc. So the business has thrived over the 0:22:08.880 --> 0:22:12.360 past years because of the underpenetration of the formal banking 0:22:12.400 --> 0:22:15.960