WEBVTT - Why Playing the Long Game Still Matters with Scottish Mortgage's Tom Slater 0:00:02.720 --> 0:00:16.560 Bloomberg Audio Studios, Podcasts, Radio News. 0:00:18.040 --> 0:00:20.599 Welcome to Maren Talk's Money, the podcastings people who know 0:00:20.640 --> 0:00:23.520 the markets explain the markets. I'm Maren Zumzert Web. This 0:00:23.600 --> 0:00:26.400 week I'm speaking with Thomas Later, manager at the UK's 0:00:26.480 --> 0:00:30.000 largest investment company, Scottish Mortgage Investment Trust, something that I 0:00:30.040 --> 0:00:32.200 suspect an awful lot of you hold. 0:00:32.479 --> 0:00:33.680 If you do, you'll know that. 0:00:33.640 --> 0:00:37.080 His long term performance has been spectacular. If you've invested 0:00:37.120 --> 0:00:40.200 in two thousands, your money is up ten times since then. 0:00:40.479 --> 0:00:43.440 The last five years have not been so great. Twenty 0:00:43.520 --> 0:00:46.199 twenty two twenty twenty three were not great years for 0:00:46.240 --> 0:00:50.160 the trust, but things are now picking up nicely again. 0:00:50.280 --> 0:00:53.440 Over the last year you are up seventeen percent and 0:00:53.560 --> 0:00:55.480 a year to date ten percent. 0:00:56.400 --> 0:00:58.560 Tom's been involved with the Trust for a decade now. 0:00:58.600 --> 0:01:00.600 We've first joined Billy Gibord in two thousand and he 0:01:00.640 --> 0:01:04.440 became a partner in two thy and twelve. So a 0:01:04.440 --> 0:01:06.960 little more insight in this conversation to how Tom is 0:01:07.000 --> 0:01:09.640 looking at investing at the moment and how the trust 0:01:09.680 --> 0:01:10.240 is faring. 0:01:10.520 --> 0:01:12.240 Tom. Welcome to Merrin Dogs Money. 0:01:12.440 --> 0:01:13.560 Thank you very much for having me. 0:01:14.120 --> 0:01:14.840 Now listen in. 0:01:14.800 --> 0:01:16.920 The introduction I've said a little bit about Scottish mortgage, 0:01:16.959 --> 0:01:18.960 but I think it would be super handy given it's 0:01:19.000 --> 0:01:20.920 such a well known trust in the UK, but maybe 0:01:20.959 --> 0:01:23.600 not so much outside. Just gives a little rundown of 0:01:23.720 --> 0:01:25.800 how the trust works, what is invested in and what 0:01:25.880 --> 0:01:26.160 you do. 0:01:26.480 --> 0:01:32.360 Sure, the trust is a portfolio of what we consider 0:01:32.440 --> 0:01:35.880 to be the most exciting growth companies in the world, 0:01:36.520 --> 0:01:39.080 and we spend our time looking for these opportunities and 0:01:39.160 --> 0:01:42.480 where we find them, we try to be really patient 0:01:42.880 --> 0:01:46.520 long term owners. We're not restricted by geography, we're not 0:01:46.560 --> 0:01:50.000 restricted by whether a company is public or private. Currently 0:01:50.000 --> 0:01:53.800 are about three quarters public companies quarter private companies. That 0:01:53.880 --> 0:01:57.040 the defining characteristic is is having a really long time 0:01:57.040 --> 0:02:00.360 horizon of which to assess these opportunities. And I guess 0:02:00.360 --> 0:02:02.640 the final point is if you look at the portfolio, 0:02:02.680 --> 0:02:04.880 it has nothing really to do with Scotland or to 0:02:04.920 --> 0:02:05.800 do with mortgages. 0:02:06.120 --> 0:02:07.800 That is simply the inheritance of the name. 0:02:08.680 --> 0:02:10.040 Yeah, and it's the name you're going to be. 0:02:10.120 --> 0:02:11.800 You're going to have long term, right, a lot of 0:02:11.800 --> 0:02:13.800 trust with what you might consider to be old fashioned 0:02:13.840 --> 0:02:16.679 names have begun to change them and shift around the place. 0:02:16.720 --> 0:02:19.320 But Scottish Mortgage is so well known as a brand 0:02:19.400 --> 0:02:21.880 that you can probably hang on to this name that 0:02:21.960 --> 0:02:24.440 doesn't in any way refer to the content of the portfolio. 0:02:24.680 --> 0:02:27.320 Fine, well, I think the heritage is really really important 0:02:27.360 --> 0:02:29.919 in the history. I do think it's sad when these 0:02:29.960 --> 0:02:34.760 trusts get incorporated into sort of bland corporate nomenclature. 0:02:34.960 --> 0:02:36.320 Now, listen, one of the things I said in the 0:02:36.360 --> 0:02:39.480 introduction is that this trust has phenomenal long term performance, 0:02:39.520 --> 0:02:41.880 but shorter term things have been a little ropie. So 0:02:41.960 --> 0:02:43.960 if we go back to twenty twenty one, that was 0:02:44.360 --> 0:02:46.640 a phenomenal year. The trust was up, the share price 0:02:46.720 --> 0:02:48.600 was up nearly one hundred percent or so, and then 0:02:48.639 --> 0:02:51.160 we went into twenty two to twenty three, and the 0:02:51.280 --> 0:02:53.640 twenty two in particular was a twenty two was an 0:02:53.639 --> 0:02:56.360 all right year. Twenty twenty three it was a horrible year. 0:02:56.840 --> 0:02:59.760 And since then things have improved, but nonetheless but not 0:03:00.120 --> 0:03:03.360 anywhere near back to the highs of those heady days 0:03:03.600 --> 0:03:06.760 of twenty twenty one. Just explain quickly for people who 0:03:06.760 --> 0:03:08.600 haven't engaged with the trust what it is that happened 0:03:08.680 --> 0:03:09.440 during that period. 0:03:09.760 --> 0:03:12.400 What happened I think as we went into the COVID 0:03:12.480 --> 0:03:15.240 period is that the world came to a lot of 0:03:15.280 --> 0:03:18.720 the companies that we owned. We owned Amazon, for example, 0:03:18.720 --> 0:03:22.000 and we all moved to starting shopping online. We owned 0:03:22.520 --> 0:03:26.680 many of the enterprise software companies that facilitated working from home. 0:03:27.080 --> 0:03:29.560 So we went through this period where the companies that 0:03:29.600 --> 0:03:32.639 we owned had this real boost to their businesses. And 0:03:32.680 --> 0:03:36.600 then as we exited the COVID period, I think habits 0:03:36.640 --> 0:03:40.440 reverted to normal a lot faster than we had anticipated, 0:03:40.960 --> 0:03:43.240 and so the tide went out for a lot of 0:03:43.240 --> 0:03:45.960 our companies, and that made things pretty difficult. And in 0:03:46.000 --> 0:03:49.120 the period you're describing, I think if you wind the 0:03:49.120 --> 0:03:52.160 clock forward to today, what we've seen is these companies 0:03:52.200 --> 0:03:55.680 dealt with very rapidly rising interest rates and changing demand patterns, 0:03:56.120 --> 0:03:59.840 and the best among them really took those signals seriously, 0:04:00.400 --> 0:04:03.160 and they thought pretty carefully about what they were going 0:04:03.240 --> 0:04:07.800 to do with their costs, with their headcount, and prioritized 0:04:07.960 --> 0:04:10.880 within their business the projects. You know, in an environment 0:04:10.920 --> 0:04:15.000 where capital had no cost, you do everything and you 0:04:15.040 --> 0:04:18.040 pursue growth. But as soon as the interest rates rose rapidly, 0:04:18.120 --> 0:04:20.680 they started to think much more carefully about prioritizing what 0:04:20.680 --> 0:04:23.719 they were doing, and so you've seen this recovery and 0:04:23.760 --> 0:04:28.680 profitability across the portfolio companies moving from sort of break 0:04:28.720 --> 0:04:32.920 even by choice to earning really pretty good margins today. 0:04:33.480 --> 0:04:35.719 And I think that's real source of optimism as you 0:04:35.760 --> 0:04:38.760 look forward, that you have that degree of robustness and 0:04:38.800 --> 0:04:43.640 adaptability that's come from that challenging period in the aftermath 0:04:43.640 --> 0:04:44.120 of COVID. 0:04:45.240 --> 0:04:46.880 And would you say that as a result of that 0:04:47.040 --> 0:04:49.560 challenging period that you've been through and your companies have 0:04:49.640 --> 0:04:53.640 been through, you have shifted anything in your investment process. 0:04:53.680 --> 0:04:58.159 I mean, looking back, for example, the investment process is 0:04:58.240 --> 0:05:02.440 very influenced by a piece of research that suggested that 0:05:02.560 --> 0:05:06.039 nearly all the returns in the stock market come from 0:05:06.080 --> 0:05:08.800 four percent of companies, right. I remember this fascinating bit 0:05:08.839 --> 0:05:11.080 of research that came out and Billy Gifford really looked 0:05:11.080 --> 0:05:13.599 at very carefully and reacted to. But one of the 0:05:13.640 --> 0:05:16.520 things possibly about looking at the market like that and 0:05:16.560 --> 0:05:18.680 thinking well, we must have those four percent of companies 0:05:18.960 --> 0:05:21.000 was thinking we must have those four percent of companies 0:05:21.040 --> 0:05:25.279 at any price, yep. And so possibly going into twenty 0:05:25.360 --> 0:05:29.160 twenty twenty twenty one, there was less of a focus 0:05:29.200 --> 0:05:32.279 on valuation inside your portfolio than there might have been. 0:05:32.440 --> 0:05:35.000 So I suppose the question is, since then is there 0:05:35.000 --> 0:05:40.280 a different approach to the valuation of companies, regardless of 0:05:40.560 --> 0:05:41.680 their amazingness. 0:05:42.240 --> 0:05:47.599 I think the philosophy that you've mentioned is core to 0:05:47.680 --> 0:05:51.000 what we do. It is just a fact of life 0:05:51.040 --> 0:05:53.600 that if you extend the timeframe through which you look 0:05:53.640 --> 0:05:57.440 at stock markets, you get a skewed distribution of returns. 0:05:57.520 --> 0:06:01.800 It's not the normal distribution that we're taught in our 0:06:01.920 --> 0:06:05.719 financial analyst training. It's that a small number of companies 0:06:06.320 --> 0:06:11.000 drive the returns, and so for us, it's how do 0:06:11.080 --> 0:06:15.000 you identify companies with that opportunity and how do you 0:06:15.040 --> 0:06:17.480 hold on to them for long enough that the return 0:06:17.720 --> 0:06:21.560 acruise for our shareholders? And in a way, it's that simple. 0:06:22.640 --> 0:06:26.360 But it also is really important that when you go 0:06:26.400 --> 0:06:29.800 through a period of tough performance that it triggers some 0:06:29.880 --> 0:06:33.280 introspection that you look at process and think about, well, 0:06:33.320 --> 0:06:37.479 what could I have done differently. One of those process 0:06:37.520 --> 0:06:41.560 outcomes for me is thinking about resilience and robustness. So 0:06:42.000 --> 0:06:45.000 it's all very well to imagine where a company will 0:06:45.040 --> 0:06:49.800 be five or ten years from today, but the conditions 0:06:49.800 --> 0:06:52.039 that it encounters along the way will be faced with 0:06:52.160 --> 0:06:56.560 today's business, So do you have the resilience and robustness 0:06:56.560 --> 0:06:59.799 to get through that in order to achieve that bigger 0:06:59.839 --> 0:07:03.400 pay chairman. That's why I started with this point that 0:07:03.440 --> 0:07:06.600 actually you've seen significant improvement and profitability in the companies 0:07:06.600 --> 0:07:10.000 we own. That for me is really encouraging because it 0:07:10.040 --> 0:07:14.600 says you have the resilience and robustness to go after 0:07:14.640 --> 0:07:17.600 that bigger picture dream. And then another point about valuation. 0:07:18.120 --> 0:07:21.920 Valuation has always been a part of our process and 0:07:21.960 --> 0:07:25.440 a really important part of our process. But what we 0:07:25.520 --> 0:07:29.440 don't do is look at short term price earnings ratios 0:07:30.480 --> 0:07:33.640 and say, well, this stock is valued at twenty times 0:07:33.640 --> 0:07:37.400 its earnings, Therefore it's cheap or it's expensive. We think 0:07:37.400 --> 0:07:41.400 there's very little information in that. So the timeframe that 0:07:41.400 --> 0:07:44.280 we think about valuation is how do we make our 0:07:44.480 --> 0:07:48.840 required level of return over the next five years, And 0:07:48.880 --> 0:07:51.840 at any given point there's a high degree of uncertainty 0:07:51.840 --> 0:07:54.720 in that. So if you take that back to what 0:07:54.880 --> 0:07:58.160 was happening in twenty one twenty two, with the benefit 0:07:58.200 --> 0:08:02.960 of hindsight, yesumber of companies had got ahead of themselves, 0:08:04.720 --> 0:08:07.119 but you also have to remember that valuation is always 0:08:07.120 --> 0:08:10.600 a relative game, and so with interest rates at zero, 0:08:11.160 --> 0:08:13.280 with growth rates that were being delivered by some of 0:08:13.320 --> 0:08:16.920 these companies revenue growing fifty or sixty percent, then actually 0:08:18.080 --> 0:08:23.480 the relative valuation didn't look particularly challenging, and with the 0:08:23.480 --> 0:08:25.880 benefit of hindsight, the sensible thing to do would have 0:08:25.920 --> 0:08:28.120 been to have gone into a lot more boring types 0:08:28.160 --> 0:08:33.360 of stocks, which themselves look pretty expensive relative to their history. 0:08:34.000 --> 0:08:38.880 But that for us isn't That doesn't fit the philosophy 0:08:38.880 --> 0:08:40.400 of what we're trying to do. It would not have 0:08:40.440 --> 0:08:44.600 been consistent with the very clear objectives that we set 0:08:44.640 --> 0:08:48.000 out for shareholders. So yes, it was a mistake, but 0:08:48.080 --> 0:08:51.839 it's I think staying true to the philosophy is really 0:08:51.880 --> 0:08:54.800 important and deviating from that would have been a bigger mistake. 0:08:55.320 --> 0:08:56.960 Yeah. Well, you could also say that it's a little 0:08:56.960 --> 0:08:58.640 early to say whether it was a mistake or not. 0:09:00.760 --> 0:09:04.160 We'll see the benefits have been long term, exactly, and 0:09:04.280 --> 0:09:04.680 I suppose. 0:09:04.720 --> 0:09:04.880 Yeah. 0:09:04.960 --> 0:09:07.000 The other question to ask really is that we know 0:09:07.040 --> 0:09:09.040 you're answer to this, but a lot of people would 0:09:09.080 --> 0:09:11.480 say that if it is true that four percent of 0:09:11.520 --> 0:09:14.199 the companies drive the majority of returns, which moment of 0:09:14.240 --> 0:09:16.040 course it is because you've done very long term research 0:09:16.080 --> 0:09:19.600 into this. Then the answer is not necessarily to hold 0:09:20.280 --> 0:09:23.080 a trust or put your faith in a manager who 0:09:23.120 --> 0:09:25.640 believes they can find that four percent, but simply to 0:09:25.720 --> 0:09:28.240 hold an ETF that covers the whole market, so you 0:09:28.320 --> 0:09:30.600 definitely get them. So you have to have a lot 0:09:30.640 --> 0:09:33.240 of faith in a manager to believe that they're the 0:09:33.240 --> 0:09:36.000 ones who can find the four percent. So I think 0:09:36.000 --> 0:09:37.959 what we should do now is move on to talking 0:09:38.000 --> 0:09:40.920 about how you find those companies. How do you look 0:09:41.000 --> 0:09:43.240 at the market and say this is the one that 0:09:43.640 --> 0:09:45.960 will survive. I mean, even in something like AI, which 0:09:45.960 --> 0:09:48.520 obviously going to be a phenomenally transformative thing already is 0:09:48.559 --> 0:09:51.440 transformative for the economy and for our for everyone's lives, 0:09:51.760 --> 0:09:53.400 there are a lot of companies doing this. How do 0:09:53.440 --> 0:09:55.280 you look at a company and say, this is the 0:09:55.280 --> 0:09:57.800 one that will be in that four percent that will 0:09:58.160 --> 0:10:01.600 almost definitely drive the returns that we've promised are invested. 0:10:02.120 --> 0:10:04.079 I suppose the first thing is that I agree with 0:10:04.440 --> 0:10:07.400 your your premise there. I think either you just stone 0:10:07.400 --> 0:10:09.880 the whole market, and that's a pretty sensible thing for 0:10:10.360 --> 0:10:12.480 people to do, especially if you can do it at 0:10:12.600 --> 0:10:16.120 a very low cost. You will capture those companies. But 0:10:16.400 --> 0:10:18.120 you do have to accept the price of doing that 0:10:18.280 --> 0:10:21.320 is that you will be diluted by the other ninety 0:10:21.360 --> 0:10:23.720 six percent of companies that don't matter. 0:10:24.520 --> 0:10:25.960 So in the way you should you should do one 0:10:26.080 --> 0:10:26.480 or the other. 0:10:26.679 --> 0:10:29.080 You shouldn't. You shouldn't mess around in the middle. You 0:10:29.080 --> 0:10:32.120 should either have either have as someone who has the 0:10:32.160 --> 0:10:35.800 whole market or have someone who's making a genuine effort 0:10:36.040 --> 0:10:37.400 to have that that four percent. 0:10:37.800 --> 0:10:39.000 In my view, you should do both. 0:10:39.559 --> 0:10:41.680 You know, I think it's a perfectly sensible thing to 0:10:41.720 --> 0:10:44.719 have that broad exposure at very low cost. And then 0:10:45.040 --> 0:10:48.160 I think depending on your risk appetite, we are a 0:10:48.240 --> 0:10:52.160 volatile portfolio. We embrace that volatility in the pursuit of 0:10:52.200 --> 0:10:54.800 long term growth. But that's not for everybody. But we 0:10:54.920 --> 0:10:58.920 do think for those that want to go after that 0:10:59.120 --> 0:11:03.000 higher level of appreciation, that there's the role for both, 0:11:03.200 --> 0:11:06.120 you know, concentrated active management and index exposure. 0:11:06.480 --> 0:11:09.960 But key in this is keeping costs to a minimum. 0:11:10.080 --> 0:11:10.280 You know. 0:11:10.400 --> 0:11:12.840 It's it's if you're going to pay for active management, 0:11:12.920 --> 0:11:16.359 then pay as little as you can and get genuinely 0:11:16.520 --> 0:11:19.120 active portfolio for your money. 0:11:19.280 --> 0:11:23.000 Yeah, okay, so how hell. 0:11:22.600 --> 0:11:25.400 First of all, it's it's it's go back to what 0:11:25.559 --> 0:11:29.160 is the scale of the opportunity and is it growing 0:11:29.240 --> 0:11:33.080 or shrinking? And what is the likelihood of capitalizing on 0:11:33.160 --> 0:11:36.480 that opportunity? And is that increasing or falling? I think 0:11:36.559 --> 0:11:40.280 those are the two starting points for process. There are 0:11:40.320 --> 0:11:45.240 lots of great businesses out there, strong franchises, great combative positions, 0:11:45.520 --> 0:11:47.720 But can you really say there is the opportunity for 0:11:47.760 --> 0:11:50.200 this business to be many times at size in the future. 0:11:50.920 --> 0:11:53.079 And so for us, that's a question that narrows down 0:11:53.200 --> 0:11:57.040 the universe pretty quickly. And then is that opportunity growing 0:11:57.120 --> 0:11:59.760 or shrinking. So if you take Amazon, which is one 0:11:59.760 --> 0:12:03.400 of our most successful investments of the past twenty five years, 0:12:03.440 --> 0:12:05.120 I think we bought it in two thousand and four, 0:12:05.160 --> 0:12:08.360 two thousand and five, so that's twenty years, not twenty five. 0:12:08.400 --> 0:12:12.760 But if you look at that business and through our ownership, 0:12:12.960 --> 0:12:15.320 what we were continually able to say is that the 0:12:15.320 --> 0:12:19.040 opportunity was increasing in size. It started as a bookseller, 0:12:19.640 --> 0:12:23.280 it went into ebooks, it went into media more broadly, 0:12:23.320 --> 0:12:26.880 it went into general merchandise, it went into selling computing infrastructure. 0:12:27.320 --> 0:12:30.480 Every time we revisited the case, you could say, I 0:12:30.480 --> 0:12:34.240 think this opportunity is big, and it's getting larger. And 0:12:34.280 --> 0:12:36.800 then as you went through time, you could also see 0:12:36.840 --> 0:12:40.120 that its competitive position was getting stronger. We talked about 0:12:40.160 --> 0:12:43.840 its competition relative to Bonds and Noble and other physical 0:12:43.920 --> 0:12:47.800 book retailers. We talked about its competitive position relative to 0:12:47.840 --> 0:12:50.880 eBay in online commerce. But over the years it just 0:12:50.960 --> 0:12:53.559 became clear and clearer that this was business which had 0:12:54.120 --> 0:12:57.560 a higher and higher likelihood on capitalizing on an opportunity 0:12:57.600 --> 0:13:01.559 that was growing. And so when you looking at opportunities today, 0:13:02.040 --> 0:13:04.000 some of the companies we've bought in the past year 0:13:04.080 --> 0:13:10.520 would be New Bank and Revolute to digital online banking businesses. 0:13:10.880 --> 0:13:15.960 New Bank focused on Latin America, more focused on Europe. 0:13:16.400 --> 0:13:19.760 And it seems to us that for the next billion 0:13:19.840 --> 0:13:21.880 consumers who are going to have a bank account, they're 0:13:21.920 --> 0:13:24.080 going to come through one of these digital platforms, not 0:13:24.120 --> 0:13:28.199 one of the traditional banks. They have a huge advantage 0:13:28.400 --> 0:13:31.640 over traditional banks with the in terms of their cost structure, 0:13:32.160 --> 0:13:35.199 not having branches and all that goes with that, and 0:13:35.240 --> 0:13:38.960 their digital tools have a huge reach, so they can 0:13:39.120 --> 0:13:43.680 get better faster than those that are limited to a 0:13:43.720 --> 0:13:46.800 single market and a smaller number of customers. So that's 0:13:46.800 --> 0:13:50.240 an opportunity that's big. It's an opportunity that's growing, and 0:13:50.480 --> 0:13:52.640 the longer time goes on and you don't see any 0:13:52.640 --> 0:13:59.680 competitive response from from those traditional regional branch based banks, 0:14:00.080 --> 0:14:02.720 the more likely it is we think those companies will 0:14:02.760 --> 0:14:04.160 capitalize on that opportunity. 0:14:05.800 --> 0:14:08.720 Okay, so when we look at themes, a lot of 0:14:08.720 --> 0:14:10.959 what you invest in is obviously at tech. So it's 0:14:11.000 --> 0:14:15.160 around the AR revolution, it's around digitalization. Where are the 0:14:15.200 --> 0:14:18.839 most interesting parts inside that tech revolution? It seems to 0:14:18.880 --> 0:14:21.000 be moving faster I think than even you might have 0:14:21.040 --> 0:14:21.920 thought a few years ago. 0:14:22.120 --> 0:14:25.480 The way that I look at this is that it's 0:14:25.800 --> 0:14:29.080 tech is not really a very helpful label. You know, 0:14:29.320 --> 0:14:32.680 we could talk about the companies that are using electricity 0:14:32.840 --> 0:14:36.680 or the companies that use telecoms. The fact that they 0:14:36.720 --> 0:14:41.840 are using technology is neither a helpful way of categorizing 0:14:41.880 --> 0:14:44.160 them or really understanding their advantages. 0:14:44.640 --> 0:14:46.280 The way we prefer to think. 0:14:46.200 --> 0:14:49.640 About it is, if you take stick with the example 0:14:49.680 --> 0:14:53.320 of Amazon, it's part analyzing the resale business and it's 0:14:53.400 --> 0:14:55.800 part analyzing the infrastructure business. 0:14:56.920 --> 0:14:57.800 If you look at. 0:14:57.720 --> 0:15:01.960 Meta, the owners of Facebook and Instagram, it's more about 0:15:02.160 --> 0:15:05.840 media and share of attention than anything else. It's about 0:15:05.880 --> 0:15:11.040 selling advertisements. But I think you know, where tech becomes 0:15:11.680 --> 0:15:15.320 really relevant is in the in the leadership of these 0:15:15.360 --> 0:15:21.400 businesses and the understanding of how technology is driving change 0:15:21.440 --> 0:15:22.359 in their industries. 0:15:23.080 --> 0:15:24.480 Take your example of AI. 0:15:25.560 --> 0:15:29.920 If you think about how a typical enterprise encounters AI 0:15:30.080 --> 0:15:33.280 and thinks about deploying it, senior management will say, right, 0:15:33.720 --> 0:15:37.480 there's something to this whole AI thing. Let's find somebody 0:15:37.560 --> 0:15:40.960 from the I T department who knows about it, and 0:15:41.040 --> 0:15:43.600 then you know, if they get increasing conviction, they maybe 0:15:43.640 --> 0:15:48.160 bring in a consultant to tell them about AI. Then 0:15:48.280 --> 0:15:50.840 perhaps you do an experiment. Let's find some little bit 0:15:50.840 --> 0:15:53.600 of our business where we can develop some of these tools, 0:15:53.640 --> 0:15:57.280 and let's see what happens. And you know, maybe we 0:15:57.320 --> 0:15:59.840 can take some people out of the customer service department 0:15:59.880 --> 0:16:02.200 because we can use some of these tools to blow 0:16:02.240 --> 0:16:06.640 AI and that whole process, that feedback loop just takes 0:16:06.760 --> 0:16:11.560 a long time. Now you compare that to let's stick 0:16:11.600 --> 0:16:15.640 with Meta and Mark Zuckerberg, somebody who has a real 0:16:15.760 --> 0:16:20.480 deep understanding of this technology. You can see immediately where 0:16:20.480 --> 0:16:23.520 it has application in the business. You know, how do 0:16:23.640 --> 0:16:27.560 we gain competitive advantage from deploying these tools? How do 0:16:27.640 --> 0:16:30.440 we drive revenue in the business from deploying these tools, 0:16:30.720 --> 0:16:33.320 and then you can see all these projects that are 0:16:33.360 --> 0:16:36.040 springing up the way that they're transforming what they're doing. 0:16:36.880 --> 0:16:40.240 So if you look at time spent on Facebook, for example, 0:16:40.240 --> 0:16:42.720 which is a pretty mature platform at this point, it 0:16:42.840 --> 0:16:45.800 was up about eight percent last year, and that's because 0:16:45.840 --> 0:16:48.920 they were using AI tools to surface content that their 0:16:49.040 --> 0:16:51.640 users love them, were engaged with, and so you can 0:16:51.640 --> 0:16:55.040 see the immediate impact in the business. Shopify would be 0:16:55.080 --> 0:16:59.040 another great example. That's a business that creates tools for 0:16:59.240 --> 0:17:02.840 retailers to operate their business. So whether that's having their 0:17:02.840 --> 0:17:08.280 own website, you know, selling through platforms like metas platforms, Instagram, 0:17:08.400 --> 0:17:11.800 et cetera, helping their physical stores doing point of sale. 0:17:12.160 --> 0:17:15.159 But what you have as a technical leader who immediately 0:17:15.240 --> 0:17:19.320 understands how AI can make the business better and deploys 0:17:19.359 --> 0:17:22.600 the technology to increase the relative advantage that they have 0:17:22.760 --> 0:17:26.520 to other players in that industry. So go back to 0:17:26.720 --> 0:17:29.760 the philosophy. It increases the size of the opportunity because 0:17:29.800 --> 0:17:32.520 you can do more with this technology, and it increases 0:17:32.560 --> 0:17:35.239 the likelihood of success because you deploy it faster than 0:17:35.280 --> 0:17:40.240 your competitors. And so when we're looking at technology, we 0:17:41.040 --> 0:17:43.200 don't get excited about you know, this is a great 0:17:43.240 --> 0:17:46.240 widget or this is a you know, a great tech. 0:17:46.560 --> 0:17:49.159 It's more, these are great businesses and they can deploy 0:17:49.240 --> 0:17:51.280 this to improve their relative position. 0:17:51.400 --> 0:17:52.800 That's what gets us really excited. 0:17:53.840 --> 0:17:56.840 Okay, interesting, And one of the things that you've written 0:17:56.840 --> 0:17:59.480 about and talked about before is how if we want 0:17:59.480 --> 0:18:02.639 this gen rational AI to be truly transformational, which we do, 0:18:02.760 --> 0:18:04.560 we want it to drive or expecting it to drive 0:18:04.560 --> 0:18:07.800 a productivity revolution across the world. Finally, thank goodness, we've 0:18:07.800 --> 0:18:10.199 been wasting a long time for that productivity revolution, haven't we. 0:18:10.480 --> 0:18:12.800 That If that's to happen, it's got to be ubiquitus. 0:18:12.880 --> 0:18:15.119 So everyone has to have chips, So you can't have 0:18:15.200 --> 0:18:17.520 these very expensive chips. It's not going to continue that 0:18:18.200 --> 0:18:21.080 the producers have these very very high margins. And one 0:18:21.119 --> 0:18:24.159 of your most successful positions over recent years has been 0:18:24.200 --> 0:18:27.320 a video of course, but that's going to suffer as 0:18:27.320 --> 0:18:31.920 a consequence of the commoditization of the products right well. 0:18:32.560 --> 0:18:35.840 And video is absolutely at the heart of this AI 0:18:36.560 --> 0:18:39.560 revolution that we are seeing and we're going to see. 0:18:39.840 --> 0:18:42.920 But I do think that we have to bring down 0:18:42.920 --> 0:18:46.119 the cost of these systems massively. If you look at 0:18:46.520 --> 0:18:50.479 the deployment of technology in previous waves, whether that was 0:18:51.080 --> 0:18:58.159 the mainframe computer, the personal computer, mobile devices, online, the cloud, 0:18:58.720 --> 0:19:03.000 and all of them brought prices down and that was 0:19:03.040 --> 0:19:06.240 a really important component of the spread of the technology. 0:19:06.880 --> 0:19:09.520 Now that has been happening in AI. 0:19:09.680 --> 0:19:12.600 If you look at the you know, the cost of 0:19:13.400 --> 0:19:17.040 processing tokens in these models is coming down very rapidly, 0:19:17.640 --> 0:19:21.439 but there is this real constraint, this bottleneck in that 0:19:21.520 --> 0:19:26.800 we just don't have enough compute and that's led to 0:19:26.800 --> 0:19:29.879 two very high chip prices. You can see it in 0:19:30.440 --> 0:19:36.240 Nvidia's margins because they have been almost a monopoly provider 0:19:36.359 --> 0:19:41.399 of the silicon that's needed for this and for me, 0:19:41.680 --> 0:19:46.479 I think if if we're really going to see this 0:19:46.560 --> 0:19:49.679 technology reach its true potential, you are going to see 0:19:49.880 --> 0:19:52.840 more than one provider of the silicon to do it. 0:19:52.960 --> 0:19:56.399 As you move from the training phase for these systems 0:19:57.880 --> 0:20:00.439 to the deployment phase, so you're running in who are 0:20:00.440 --> 0:20:04.400 actually using the technology. It's not clear that Nvidia's edge 0:20:04.600 --> 0:20:07.960 is as strong in that inference market as it has 0:20:08.040 --> 0:20:10.760 been in the training market. So inspite to how much 0:20:10.800 --> 0:20:12.439 exposure do you want to have, and the decision we 0:20:12.480 --> 0:20:16.000 took last year is that we would reduce our direct 0:20:16.000 --> 0:20:19.320 exposure to Nvidia. We bought and then grew a new 0:20:19.359 --> 0:20:23.480 holding in TSMC, which is the biggest manufacturer of these 0:20:23.560 --> 0:20:26.200 chips and is a bit more agnostic to who is 0:20:26.280 --> 0:20:29.640 designed them. But we're also putting money into companies like Matter, 0:20:29.720 --> 0:20:31.800 so we don't want to reduce our overall exposure to 0:20:31.840 --> 0:20:33.879 this area, which is is you know, it's going to 0:20:33.920 --> 0:20:37.879 define the next twenty years of investing, but it's we 0:20:37.920 --> 0:20:40.680 don't think it's going to be as concentrated in Nvideo 0:20:40.760 --> 0:20:44.240 as it has been over the past two or three years. 0:20:44.400 --> 0:20:47.720 Yeah, I mean the other big bottleneck in the AI 0:20:47.840 --> 0:20:50.880 story is the energy part of it, right, So we're 0:20:50.880 --> 0:20:53.240 going to have all these huge data centers and this 0:20:53.320 --> 0:20:57.080 is incredibly energy hungry. I'm sure become less energy hungry 0:20:57.080 --> 0:20:59.320 over the years, but nonetheless we do have this constraint 0:20:59.359 --> 0:21:02.240 on it at the moment to overlook at investing in 0:21:02.440 --> 0:21:04.960 the energy sector. And we've talked on a podcast a 0:21:04.960 --> 0:21:07.800 few weeks ago about how extraordinary it is that people 0:21:07.800 --> 0:21:09.560 are paid to the moon and back for a company 0:21:09.560 --> 0:21:12.320 that is connected to AI, but absolutely nothing for the 0:21:12.359 --> 0:21:15.520 mining companies that produce the materials, without which none of 0:21:15.520 --> 0:21:16.600 these things can exist. 0:21:17.920 --> 0:21:22.880 Yeah, I think energy is a significant constraint, and we're 0:21:22.920 --> 0:21:27.119 moving from an environment where you've seen very little growth 0:21:27.160 --> 0:21:30.480 in energy demand for quite a sustained period in Western 0:21:30.520 --> 0:21:34.240 economies to one where you can see that demand growth 0:21:34.320 --> 0:21:37.320 coming through. And actually it's pretty easy to pick up 0:21:37.320 --> 0:21:40.040 on because, particularly in the US, the utilities have to 0:21:40.080 --> 0:21:42.720 submit these ten year plans, and so you can see 0:21:42.720 --> 0:21:45.080 the assumptions in those plans changing over time. 0:21:46.880 --> 0:21:49.280