WEBVTT - What Can Quant Trading Strategies Teach Us About Markets? 0:00:07.840 --> 0:00:10.200 Welcome to Merrin Talks Money, the podcast in which people 0:00:10.200 --> 0:00:12.960 who know the markets explain the market. I am Maren 0:00:13.000 --> 0:00:16.120 summrset Web and this week I am speaking with Simon Judes, 0:00:16.120 --> 0:00:18.000 who's the CIO of Winton. 0:00:18.239 --> 0:00:18.439 Now. 0:00:18.440 --> 0:00:22.040 Winton is a quantitrative investment management firm, headquarters in London, 0:00:22.120 --> 0:00:24.800 but with officers around the world. The firm managers around 0:00:24.840 --> 0:00:28.640 sixteen billion dollars worth of cash, mainly for institutional clients, 0:00:28.640 --> 0:00:32.080 but I think increasingly with an idea of having ordinary 0:00:32.080 --> 0:00:34.680 investors and wealth managers by their products as well. So 0:00:34.720 --> 0:00:37.080 we will talk about that along the way. Simon, Welcome 0:00:37.120 --> 0:00:38.040 to Merrin Talks Money. 0:00:38.320 --> 0:00:39.400 Thank you very much for having me. 0:00:40.120 --> 0:00:43.080 Right, let's start by trying to explain what on earth 0:00:43.120 --> 0:00:43.920 it is that you do. 0:00:44.040 --> 0:00:45.720 We hear a lot about quantitive investing. 0:00:45.760 --> 0:00:49.520 We hear about CTA, we all those model systems, et cetera, 0:00:49.520 --> 0:00:49.880 et cetera. 0:00:50.000 --> 0:00:52.320 What is it that you actually do. 0:00:52.960 --> 0:00:55.520 Well, let me talk about quant investing in a more 0:00:55.560 --> 0:00:57.840 general term, and then I'll come on to CTAs in particular, 0:00:57.880 --> 0:01:00.400 which is one type of quant investing. So the way 0:01:00.440 --> 0:01:02.279 to think about it is imagine you start off with 0:01:02.320 --> 0:01:05.160 a more traditional style of investment. Where perhaps you have 0:01:05.240 --> 0:01:07.959 analysts thinking about a company and whether they should buy 0:01:07.959 --> 0:01:09.840 shares in the company, whether they think the company is 0:01:09.920 --> 0:01:12.840 likely to do well. That would be something that a 0:01:12.880 --> 0:01:15.200 lot of effort goes into a lot of discussion, and 0:01:15.240 --> 0:01:17.240 if you want to make that decision not about one company, 0:01:17.319 --> 0:01:19.360 but about ten companies, you need to have ten times 0:01:19.360 --> 0:01:23.039 as much discussion, and that leads you to scale those 0:01:23.080 --> 0:01:26.800 organizations quite heavily and to focus on having a small 0:01:26.880 --> 0:01:29.120 number of positions that you have a high conviction in. 0:01:29.160 --> 0:01:32.680 That's how traditional investment works, and quantum investment is based 0:01:32.720 --> 0:01:35.480 on a different idea where you say, well, maybe we 0:01:35.520 --> 0:01:38.360 don't have to have that high conviction in every single 0:01:38.400 --> 0:01:40.360 investment if we can have a larger number of them. 0:01:41.040 --> 0:01:42.720 So it's a little bit more akin to how an 0:01:42.760 --> 0:01:45.360 insurance company might treat its products. It doesn't have to 0:01:45.360 --> 0:01:47.960 make money on every single policy itsels as long as 0:01:47.960 --> 0:01:50.760 it arranges things so that on average it does. So 0:01:50.800 --> 0:01:53.560 the idea is, well, what if instead of doing all 0:01:53.560 --> 0:01:56.040 this huge amount of analysis on each company that you 0:01:56.080 --> 0:01:58.760 decide to buy a share in, you instead come up 0:01:58.760 --> 0:02:02.480 with a rule that is perhaps much simpler. And it's 0:02:02.480 --> 0:02:04.600 something that's so simple that you can automate. You can 0:02:04.600 --> 0:02:06.400 get a computer to do it. So I suppose that 0:02:06.720 --> 0:02:09.119 you look at some valuation metrics or something else about 0:02:09.160 --> 0:02:11.720 how the company's performed recently, and you decide based on 0:02:11.800 --> 0:02:13.920 that you will have a position of a certain size 0:02:13.919 --> 0:02:15.760 in the company. And that's something that you can get 0:02:15.760 --> 0:02:17.560 a computer to do for you, and you can do 0:02:17.639 --> 0:02:21.639 that with a much much larger portfolio. Now that might 0:02:21.760 --> 0:02:24.600 sound a little bit speculative, like how could you know 0:02:24.680 --> 0:02:26.560 that following a rule like that would be a good 0:02:26.560 --> 0:02:28.680 decision for you in the long run. And the reason 0:02:28.720 --> 0:02:30.560 that you can get confidence in that is because you 0:02:30.600 --> 0:02:32.520 can do something that you can't do in the more 0:02:32.560 --> 0:02:35.320 traditional style of investment. What you can do is a 0:02:35.360 --> 0:02:38.160 back test where you take that rule and you test 0:02:38.200 --> 0:02:41.119 how it would have performed over a much longer time horizon. 0:02:41.600 --> 0:02:44.880 And that is what gives you confidence in quant investing 0:02:45.440 --> 0:02:46.519 that this is going to be a good rule for 0:02:46.560 --> 0:02:48.520 you to follow. And if you think about what that 0:02:48.560 --> 0:02:51.600 then looks like, it means that when you are allocating 0:02:51.680 --> 0:02:54.040 money as a quant investor, you're not allocating or you're 0:02:54.080 --> 0:02:57.560 not really thinking directly about each individual investment. You're thinking 0:02:57.560 --> 0:03:00.800 about allocating to these rules, these ideas or algorithms that 0:03:00.840 --> 0:03:03.560 are themselves going to decide which companies they buy and 0:03:03.560 --> 0:03:06.079 which companies they sell. So that's the general idea of 0:03:06.120 --> 0:03:09.680 quant investing. It's the difference from more traditional styles of investment, 0:03:09.760 --> 0:03:12.560 and that's how you can get confidence by doing these 0:03:12.880 --> 0:03:15.320 historical studies which show you how the rule would have performed. 0:03:15.680 --> 0:03:19.080 How does that work with a long term idea that 0:03:19.160 --> 0:03:22.280 every time an anomaly or a gap in the market 0:03:22.320 --> 0:03:24.680 like that has found, it's almost immediately traded away. 0:03:26.080 --> 0:03:29.440 That's a great question. You definitely see with many types 0:03:29.480 --> 0:03:31.800 of rules like that that they work for a while, 0:03:32.600 --> 0:03:34.600 and then you can even see in this back test 0:03:34.639 --> 0:03:37.040 that they gradually start to work less and less well. 0:03:37.520 --> 0:03:39.960 And that almost certainly is always because many more people 0:03:39.960 --> 0:03:43.000 are doing it. So what can you do about that? Well, 0:03:43.200 --> 0:03:44.880 the first thing you can do is that there are 0:03:44.920 --> 0:03:49.960 some ideas which don't have that fall off in a way. 0:03:50.000 --> 0:03:53.880 Perhaps they're not based on discovering a simple trick or arbitrage, 0:03:53.920 --> 0:03:56.120 which once somebody else has discovered it, it will just 0:03:56.120 --> 0:03:59.240 stop working. They might be based on more general behavioral 0:03:59.240 --> 0:04:02.440 biases that people have, and that's really where CTAs and 0:04:02.480 --> 0:04:05.840 momentum and trendfall going to sit in this way of 0:04:05.880 --> 0:04:06.520 thinking about it. 0:04:06.560 --> 0:04:09.080 So there's two types of rules that we might follow here. 0:04:09.280 --> 0:04:12.200 One would be purely in numerical statistical, and the other 0:04:12.320 --> 0:04:13.160 might be behavioral. 0:04:14.520 --> 0:04:17.760 Well, even the behavioral ones we would still measure through data. 0:04:17.880 --> 0:04:20.240 They would still be measured through the way that either 0:04:20.320 --> 0:04:22.560 prices or other sort of days that we might use 0:04:22.800 --> 0:04:26.320 are behaving. So they're still encoded as systematic algorithms that 0:04:26.360 --> 0:04:29.080 take in data, transform it in some or other way, 0:04:29.120 --> 0:04:31.080 and then come up with a rule for how you trade. 0:04:31.160 --> 0:04:33.200 So I think those two things might end up being 0:04:33.240 --> 0:04:36.320 the same. But the second way I was going to 0:04:36.320 --> 0:04:39.599 say that you can combat that people call it alpha decay, 0:04:39.640 --> 0:04:42.440 where the idea stops working, is by doing a lot 0:04:42.480 --> 0:04:45.160 of research and continuing to find new ideas. And that's 0:04:45.160 --> 0:04:47.159 a big component of our activity as well. 0:04:47.920 --> 0:04:49.680 Okay, so there are still new ideas to. 0:04:49.680 --> 0:04:52.719 Find, absolutely, because there's new data all the time as well. 0:04:53.120 --> 0:04:56.120 Can you give us an example of a new rule. 0:04:57.200 --> 0:04:59.400 Well, that's why you might be infringing. 0:04:58.960 --> 0:05:01.040 In Okay, you don't want to do that because it's 0:05:01.040 --> 0:05:03.440 proprietary stuff. And this is one of the problems I 0:05:03.440 --> 0:05:05.440 think the retail investors often have with this type of 0:05:05.440 --> 0:05:08.240 fund is that they're told the theory, but it's not 0:05:08.240 --> 0:05:10.719 possible for them to be told the practicalities because that's 0:05:10.839 --> 0:05:12.520 your model, your proprietary data. 0:05:12.560 --> 0:05:14.159 So why don't we go back a bit. 0:05:14.520 --> 0:05:16.960 Maybe you can give us an example of a rule 0:05:17.320 --> 0:05:21.000 that you use that once worked and doesn't anymore. That way, 0:05:21.040 --> 0:05:23.520 we're not infringing on your corporate knowledge. 0:05:23.680 --> 0:05:25.040 Well, I think we can do better on that because 0:05:25.279 --> 0:05:27.720 in the CTA space actually, which is a particular type 0:05:27.760 --> 0:05:31.320 of quant investment, the basic idea of how we come 0:05:31.400 --> 0:05:34.920 up with the trades is quite well known and it's 0:05:35.240 --> 0:05:38.760 in something we can talk about very openly. So CTAs 0:05:38.839 --> 0:05:42.520 are a particular subset of quant hedge funds which apply 0:05:42.640 --> 0:05:44.719 that same philosophy that I was talking about earlier to 0:05:45.000 --> 0:05:49.599 macro markets, primarily through trading futures, so that they're trading equities, 0:05:49.640 --> 0:05:52.520 that they're trading bonds, they're trading currencies and their training commodities, 0:05:52.960 --> 0:05:55.360 and predominantly the way they're doing that is by looking 0:05:55.360 --> 0:05:57.920 at recent price momentum. So the things that have gone 0:05:58.000 --> 0:06:00.880 up over some recent period, you have a long position, 0:06:01.200 --> 0:06:03.000 and if they've gone down, then you have a short position, 0:06:03.560 --> 0:06:06.440 and that is the algorithm. It really is looking mainly 0:06:06.480 --> 0:06:07.719 at recent price behavior. 0:06:09.200 --> 0:06:12.800 Okay, so you've got a fund that trays in futures 0:06:12.839 --> 0:06:15.920 in all the asset classes based on the direction prices 0:06:15.960 --> 0:06:18.520 are already going, with the idea that you can spot 0:06:18.560 --> 0:06:19.560 where the turn will come. 0:06:20.680 --> 0:06:23.839 No, we can't really spot when the that's very hard. 0:06:23.839 --> 0:06:26.120 It's based on the idea that if the price has 0:06:26.120 --> 0:06:28.640 been going in that direction, that's more likely than not 0:06:28.680 --> 0:06:30.200 to continue going in that direction. 0:06:30.400 --> 0:06:32.120 Okay, so you've got to have an awful lot of 0:06:32.120 --> 0:06:34.200 positions for that general rule to work. 0:06:35.040 --> 0:06:38.119 Yes, you need, you know, more than one hundred really 0:06:38.480 --> 0:06:43.000 to get the benefit of the diversification, because, like I said, 0:06:43.120 --> 0:06:45.480 in general, with quantuvesting, you don't have very much conviction 0:06:45.560 --> 0:06:48.480 in any one position. So I'm slightly suspecting at some 0:06:48.520 --> 0:06:50.440 point you'll ask me what gold is going to do next? 0:06:50.480 --> 0:06:54.720 For example, Yeah, that's coming. The answer is, I don't know. 0:06:55.200 --> 0:06:58.520 When the question comes, I don't know, but we believe 0:06:58.520 --> 0:07:00.680 that there's a greater than fifty percent chance that it 0:07:00.720 --> 0:07:03.200 will continue to go up, which is not saying very much, 0:07:03.480 --> 0:07:05.680 but as part of a large portfolio, it's still a 0:07:05.760 --> 0:07:06.640 very valuable thing to do. 0:07:07.240 --> 0:07:09.320 Okay, before we move into a CTR. What does CT 0:07:09.480 --> 0:07:10.760 stand for? By the way it. 0:07:10.720 --> 0:07:13.800 Sounds for Commodity Trading advisor, which is it's a regulatory 0:07:13.880 --> 0:07:18.400 term from the seventies in the US, so it doesn't 0:07:18.400 --> 0:07:18.880 have very much. 0:07:19.160 --> 0:07:21.640 It doesn't really the words don't really mean anything anymore. 0:07:21.640 --> 0:07:24.360 The acronym means something, but the words no longer mean anything. 0:07:24.720 --> 0:07:25.160 Correct. 0:07:25.440 --> 0:07:27.080 Okay, let's go back to the. 0:07:29.520 --> 0:07:33.040 Pre CTA conversation about the different rules and evaluations. I 0:07:33.040 --> 0:07:35.960 am actually really really keen to nail you down a 0:07:36.000 --> 0:07:38.320 little bit on the kind of thing that you mean 0:07:38.560 --> 0:07:40.720 when we talk about all these different rules, all these 0:07:40.760 --> 0:07:45.600 different diversification, different models, etc. For the purpose of the 0:07:45.640 --> 0:07:49.120 retail investor who is not used to this kind of 0:07:49.200 --> 0:07:53.240 fund and is used to more traditional equity funds where 0:07:53.240 --> 0:07:55.080 we're talking about or we're doing growth and this is 0:07:55.080 --> 0:07:57.000 what growth means, and we're doing value and there's more 0:07:57.080 --> 0:07:59.560 value means, and we're doing growth through a reasonaal price 0:07:59.560 --> 0:08:01.320 and that's what this means. And we're doing tech and 0:08:01.320 --> 0:08:03.960 that's but this means we're doing mining. That's what this means, etcetera. 0:08:04.040 --> 0:08:06.040 These this stuff is easy to get a handle on, 0:08:06.320 --> 0:08:10.280 but what you do is not. So a neat little 0:08:10.280 --> 0:08:15.720 example of one type of strategies, less models, less valuation metric, etc. 0:08:16.040 --> 0:08:18.520 That works or has worked for you would be super handy. 0:08:18.920 --> 0:08:21.000 Absolutely. I think value is a good one to focus on. 0:08:21.480 --> 0:08:25.160 So if you think about traditional value investors, Warren Buffett 0:08:25.200 --> 0:08:26.800 entering gain this kind of thing there is. 0:08:26.800 --> 0:08:28.520 We take issue with that one more on Buffet and 0:08:28.600 --> 0:08:29.800 value investing. Come back to that. 0:08:29.880 --> 0:08:32.199 Maybe, well, think of somebody who's doing a lot of 0:08:32.240 --> 0:08:35.160 effort into assessing a company and deciding if the current 0:08:35.200 --> 0:08:38.880 market price represents good value. So maybe if that's the way, so. 0:08:38.880 --> 0:08:41.600 That now describes every single investor there is, because no 0:08:41.640 --> 0:08:43.680 one buys something which they don't think is good value, 0:08:43.760 --> 0:08:46.440 even a growth investor. If they're paying seventy times for something, 0:08:46.480 --> 0:08:48.320 we'll think is good value because they expected to go 0:08:48.400 --> 0:08:50.840 up right absolutely same, same, same. 0:08:51.120 --> 0:08:52.880 So yeah, okay, a lot of people would would fit 0:08:52.920 --> 0:08:55.320 themselves into that that way of describing things. And now 0:08:55.400 --> 0:08:57.800 imagine instead of putting a huge amount of effort into 0:08:57.840 --> 0:09:00.320 each company into looking at what the management of said, 0:09:00.640 --> 0:09:03.320 into looking all over the balance sheet, into looking at 0:09:03.360 --> 0:09:05.480 the drivers of the company's profits and so on, and 0:09:05.520 --> 0:09:07.480 work out what the price should be Imagine that all 0:09:07.480 --> 0:09:09.640 you do is you look at the price siarnings ratio 0:09:09.960 --> 0:09:12.760 of every company that you can find, and you just 0:09:12.880 --> 0:09:15.160 rank them, and you just look at the ones that 0:09:15.600 --> 0:09:17.160 have lower than average and the ones that have higher 0:09:17.160 --> 0:09:19.880 than average price sarnage ratios, and you just decide to 0:09:19.920 --> 0:09:22.719 go along the ones that have lower than average PE 0:09:22.800 --> 0:09:24.880 ratios and short the ones that have higher than average. 0:09:25.320 --> 0:09:29.000 That is an example of an algorithmic rule, which in 0:09:29.040 --> 0:09:34.080 and of itself is not very sophisticated or clever, obviously, 0:09:34.160 --> 0:09:37.160 because it doesn't involve any interesting analysis of what those 0:09:37.200 --> 0:09:39.439 companies are doing. It just takes one piece of data 0:09:39.480 --> 0:09:41.960 per company and update it roughly once a quarter. Depending 0:09:42.000 --> 0:09:43.720 on how from these numbers are updated, it. 0:09:43.720 --> 0:09:45.160 Would rebalance once a quarter. 0:09:45.400 --> 0:09:48.560 Generally, yes, exactly. So this is an idea that many 0:09:48.559 --> 0:09:53.000 people have traded for decades. It doesn't work that well 0:09:53.360 --> 0:09:56.280 anymore as well as it used to, and many people 0:09:56.840 --> 0:09:59.520 have adopted more sophisticated tweaks on it. You know, perhaps 0:09:59.800 --> 0:10:02.199 you don't just rank every single company. Maybe you look 0:10:02.240 --> 0:10:05.680 at the ranking of a company within a sector. Maybe 0:10:05.720 --> 0:10:08.080 instead of looking at the trailing earnings, you look at 0:10:08.080 --> 0:10:11.480 analysts forecasts of earnings maybe you try to forecast what 0:10:11.520 --> 0:10:13.680 the next earning announcement will be yourself. You know, there 0:10:13.679 --> 0:10:15.800 are lots of ways to make the idea still quantitative 0:10:15.800 --> 0:10:18.600 and systematic, but more sophisticated than that original base idea. 0:10:19.520 --> 0:10:22.080 But ultimately, what they're all doing is something that is 0:10:22.200 --> 0:10:25.080 much less in depth than what a traditional investor would 0:10:25.080 --> 0:10:26.960 do to assess whether a company's good value or not. 0:10:27.640 --> 0:10:30.200 And on the flip side is that you can do 0:10:30.280 --> 0:10:32.240 it very easily for a very large number of companies, 0:10:32.520 --> 0:10:35.400 and you can back test it through history and you 0:10:35.440 --> 0:10:37.200 can see what following this rule would have done. 0:10:37.400 --> 0:10:39.920 Okay, so you can do all that, and you have 0:10:39.960 --> 0:10:43.439 maybe a one hundred different strategies running in in a fund. 0:10:43.440 --> 0:10:44.240 Do you said earlier? 0:10:45.360 --> 0:10:48.160 It varies. We have different products. So some of the 0:10:48.200 --> 0:10:51.240 CTA products, for example, are really based around the properties 0:10:51.640 --> 0:10:53.640 of momentum of trend following on macro and markets, and 0:10:53.840 --> 0:10:55.800 there are good reasons for that because that has a 0:10:55.800 --> 0:11:00.360 specific function for investor portfolios, which we can talk about products. 0:11:00.400 --> 0:11:02.920 We have combined a much larger number of different strategies 0:11:02.920 --> 0:11:05.400 where we're really trying to benefit from the diversification you 0:11:05.440 --> 0:11:07.000 get from a large number of different ideas. 0:11:07.240 --> 0:11:09.400 This sounds to me like the kind of thing that 0:11:09.679 --> 0:11:12.320 AI would be remarkably helpful for. 0:11:13.040 --> 0:11:16.720 AI certainly is helpful for us. It's not something that's 0:11:16.760 --> 0:11:20.640 extremely new in a sense, it's the latest example of 0:11:21.320 --> 0:11:25.760 a fairly transformative technology or set for techniques. It probably 0:11:25.760 --> 0:11:28.120 wouldn't surprise you, given that we have a huge technology 0:11:28.120 --> 0:11:30.480 component to our business and the people we employ are 0:11:30.679 --> 0:11:33.200 people with backgrounds and software engineering and sciences and so on, 0:11:33.280 --> 0:11:35.880 that this has always been something that's very important for us. 0:11:36.160 --> 0:11:39.040 Before AI, it was machine learning, and it was all 0:11:39.080 --> 0:11:40.640 times of data, and it was big data, and it 0:11:40.640 --> 0:11:43.520 was cloud technology, and it's very important for us to 0:11:44.440 --> 0:11:46.400 keep up with the way that all of these ideas 0:11:46.400 --> 0:11:50.160 develop and certainly the recent developments in large language models 0:11:50.160 --> 0:11:51.280 have been very helpful for us. 0:11:51.600 --> 0:11:53.880 Does it make everything move faster, faster and faster, so 0:11:54.000 --> 0:11:56.160 the edge that you might get from each new idea 0:11:57.200 --> 0:11:58.960 is removed much more quickly? 0:12:01.440 --> 0:12:03.720 Well, to some extent. I mean there's a general truth 0:12:03.720 --> 0:12:08.200 in investing that nothing you know ever totally transforms the 0:12:08.240 --> 0:12:10.600 business because at the end of the day, of everyone 0:12:10.640 --> 0:12:16.080 discovers something, then its value reduces because ultimately there's a 0:12:16.120 --> 0:12:19.880 certain number of dollars to be made out of forgiven idea. 0:12:20.120 --> 0:12:21.680 The more people who do it, the more they have 0:12:21.760 --> 0:12:24.199 to share it. But that doesn't mean that you can 0:12:24.600 --> 0:12:27.000 just not follow it. You still have to do these things. 0:12:27.400 --> 0:12:28.840 All right, Let's move on for what I know you 0:12:28.840 --> 0:12:30.720 really want to talk about, which is the CTA stuff 0:12:30.720 --> 0:12:31.760 and the trend following fund. 0:12:31.880 --> 0:12:33.079 Explain how that works. 0:12:33.520 --> 0:12:36.319 Yeah, absolutely so. This is a quant strategy that trades 0:12:36.520 --> 0:12:41.800 the macro markets, primarily through futures, and it has a 0:12:41.800 --> 0:12:45.560 particular role in investor portfolios, which has always been there, 0:12:45.600 --> 0:12:49.360 but it's become increasingly important really since twenty twenty two. 0:12:50.320 --> 0:12:52.160 There have been other moments as well where investors have 0:12:52.160 --> 0:12:54.880 become particularly interested in CTAs. The financial crisis was one, 0:12:54.880 --> 0:12:58.400 particularly where equities obviously fell a lot and lots of 0:12:58.400 --> 0:13:01.480 other asset classes that did poorly, and people start to 0:13:01.520 --> 0:13:03.120 look around for what things had done well. And at 0:13:03.120 --> 0:13:05.480 that time, the two things that had done well were 0:13:05.559 --> 0:13:09.800 bonds and CTAs, and so there was a big surge 0:13:09.800 --> 0:13:12.040 of investment into both of them after the financial crisis. 0:13:12.080 --> 0:13:16.000 And then in twenty twenty two, the remarkable thing that happened, 0:13:16.360 --> 0:13:19.480 which is a headache not just for the retail investors 0:13:19.480 --> 0:13:22.240 that you mentioned, but actually for all investors, is that 0:13:22.320 --> 0:13:27.160 bonds and equities fell together for a year. The reason 0:13:27.320 --> 0:13:31.160 is the problem is because a lot of allocation philosophies 0:13:31.240 --> 0:13:33.719 had relied on the idea that bonds could provide divestification 0:13:33.760 --> 0:13:37.040 to equities, and that wasn't a crazy idea. You'd had 0:13:37.040 --> 0:13:40.120 a twenty year period leading up to there where bond's 0:13:40.120 --> 0:13:43.079 and equities were negatively correlated, and not every time that 0:13:43.120 --> 0:13:46.120 pretty much every time stocks had gone down, bonds had helped, 0:13:46.160 --> 0:13:47.720 as they did in the financial crisis. 0:13:47.720 --> 0:13:50.400 So the traditional sixty to forty portfolio worked for everybody 0:13:50.440 --> 0:13:52.520 for a very long time and then suddenly kind of 0:13:52.559 --> 0:13:53.920 stopped working exactly. 0:13:54.000 --> 0:13:55.600 And in fact, it didn't have to be sixty forty. 0:13:55.600 --> 0:13:57.680 You know, any reasonable combination of bonds and equities would 0:13:57.679 --> 0:14:00.839 have performed extremely well over that time period. And then 0:14:01.679 --> 0:14:06.160 that diversification disappeared. Really starting twenty twenty, the correlation flip 0:14:06.200 --> 0:14:09.080 to being positive, which meant that the potential for them 0:14:09.080 --> 0:14:10.600 both to go down at the same time was there, 0:14:10.640 --> 0:14:12.520 even if it hadn't actually happened yet. But then in 0:14:12.520 --> 0:14:14.760 twenty twenty two that potential was realized and they both 0:14:14.800 --> 0:14:17.440 did go down together, and that really drove home that 0:14:18.520 --> 0:14:21.360 you can't just rely on bonds to be your diversified equities, 0:14:21.400 --> 0:14:24.200 and so people started again looking at at CTA strategies, 0:14:24.200 --> 0:14:26.280 which had done very well that year. So why is 0:14:26.320 --> 0:14:29.800 it that CTA strategies provide this diversification. There are basically 0:14:29.880 --> 0:14:33.680 two reasons. The first reason is that among the futures 0:14:33.760 --> 0:14:36.200 markets that they trade, there are a variety of things 0:14:36.240 --> 0:14:39.280 which are very uncorrelated. Exposures are things which are quite 0:14:39.280 --> 0:14:42.320 difficult for people to benefit from otherwise. So an example 0:14:42.360 --> 0:14:44.400 of that would be, for example, the big rise in 0:14:44.440 --> 0:14:46.760 the price of cocoa which took place in late twenty 0:14:46.800 --> 0:14:49.440 twenty three and early twenty twenty four, and then subsequently 0:14:49.480 --> 0:14:52.040 the big drop in the price of coca that happened 0:14:52.080 --> 0:14:54.440 in the last few months. That's a trend that you 0:14:54.480 --> 0:14:56.760 can benefit from by trading the futures markets, but it's 0:14:56.840 --> 0:14:59.960 very difficult to get that exposure through stocks, for example. 0:15:00.600 --> 0:15:03.000 So that's one reason. The other reason that you can 0:15:03.000 --> 0:15:05.440 benefit a lot from CTAs in a portfolio is that 0:15:05.480 --> 0:15:07.960 because they trade futures markets, and futures markets have this 0:15:08.080 --> 0:15:10.480 amazing property that it's just as easy to be short 0:15:10.480 --> 0:15:13.320 as it is to be long. You can benefit from 0:15:13.320 --> 0:15:16.360 short positions in the major asset classes. It's just as 0:15:16.360 --> 0:15:18.680 easy to maintain a position that short equities or short 0:15:18.720 --> 0:15:20.680 bonds as it is to be long equities and long bonds, 0:15:20.920 --> 0:15:23.120 and so trivially in the year like twenty twenty two, 0:15:23.640 --> 0:15:26.280 the strategy is short in those asset classes and it 0:15:26.320 --> 0:15:26.760 does well. 0:15:27.040 --> 0:15:29.200 So that was a good year for CTA. 0:15:29.520 --> 0:15:32.320 That was a period when you're the assets that you 0:15:32.360 --> 0:15:35.320 hold kind of peak, right, so that there's winter and 0:15:35.360 --> 0:15:37.960 as a whole managed less money now than it did 0:15:38.000 --> 0:15:40.880 at its peak. What's driven that is that because people 0:15:40.880 --> 0:15:43.400 looked at the successes in times like that and then go, well, 0:15:43.440 --> 0:15:45.120 since then, long equity has been marvel. 0:15:45.120 --> 0:15:46.760 It's very easy to make money. And why do you 0:15:46.800 --> 0:15:47.880 need to mess around the edges? 0:15:48.840 --> 0:15:51.720 Oh, our peak was at a different time from that, well, 0:15:51.800 --> 0:15:52.000 was it? 0:15:52.320 --> 0:15:52.760 But earlier? 0:15:53.240 --> 0:15:56.600 Yeah, well in twenty sixteen. But I think there's there's 0:15:56.640 --> 0:15:58.800 a peak of a single manager, which has more to 0:15:58.840 --> 0:16:01.600 do with what's going on in that organization. But the 0:16:01.640 --> 0:16:04.680 peak of the industry is a different question. So in 0:16:04.760 --> 0:16:07.280 terms of how much people allocating to CTAs, it's a 0:16:07.320 --> 0:16:11.160 little more difficult to tell because some CTAs are public 0:16:11.760 --> 0:16:15.560 like we are. Others are offered through products that perhaps 0:16:15.600 --> 0:16:18.080 banks offer, and is not publicly known how much is 0:16:18.400 --> 0:16:21.520 being traded there. But what we're certainly seeing is that 0:16:22.520 --> 0:16:24.920 the interest relates strongly to the performance of the strategy, 0:16:25.040 --> 0:16:26.080 as you'd expect. 0:16:25.840 --> 0:16:29.200 Of course, okay, and the performance of the strategy long term. 0:16:30.120 --> 0:16:33.720 Well, look, long term, it's been a great addition and diversifier. 0:16:33.800 --> 0:16:35.840 So you can take any combination you like of bond's 0:16:35.840 --> 0:16:39.360 and equities, and it's been beneficial to add an allocation 0:16:39.440 --> 0:16:42.520 to CTAs as well over the long term. Beneficial like her, 0:16:43.280 --> 0:16:46.120 Beneficial in the sense that performance has improved, and particularly 0:16:46.120 --> 0:16:50.160 the drawdowns that you get when you have these big 0:16:50.200 --> 0:16:53.320 moments like twenty eight or twenty twenty two are mitigated 0:16:53.360 --> 0:16:53.880 to some extent. 0:16:54.280 --> 0:16:56.800 Let's say, for example, you're a retail investor and you 0:16:56.880 --> 0:16:59.600 make five percent of your portfolio CTA. 0:17:00.600 --> 0:17:02.680 How would that how would that have helped your performance 0:17:02.680 --> 0:17:04.240 over I say a decade long period. 0:17:05.200 --> 0:17:07.480 Well, you know, the exact number is very depending on. 0:17:07.560 --> 0:17:08.920 I know, I know the exac num is very bad. 0:17:08.960 --> 0:17:10.560 But you know, we need we need to know that 0:17:10.640 --> 0:17:13.040 it would make a positive difference of more than a 0:17:13.080 --> 0:17:15.400 couple of basis points here are that it does. 0:17:15.760 --> 0:17:17.240 It makes a positive and you can measure in a 0:17:17.280 --> 0:17:18.639 couple of ways. So you can say, well, there's an 0:17:18.680 --> 0:17:21.520 annualized number which might go up by one percent or 0:17:21.520 --> 0:17:24.080 something depending on exactly the waiting you give to it. 0:17:24.400 --> 0:17:27.359 But then what is also important is not just that 0:17:27.440 --> 0:17:30.760 sort of average, but the fact that historically, at least 0:17:30.840 --> 0:17:33.200 the CTAs have performed well in the periods which haven't 0:17:33.200 --> 0:17:37.200 been good for equities. And so it's added the profits 0:17:37.240 --> 0:17:40.640 at particular moments when other things were doing badly, which 0:17:40.640 --> 0:17:45.040 isn't as easily quantified by just adjusting the annual return number, 0:17:45.200 --> 0:17:47.320 but it's a very important component of why people are 0:17:47.320 --> 0:17:48.119 allocating to it. 0:17:48.400 --> 0:17:51.040 Yeah, can you tell from of course you can, But 0:17:51.200 --> 0:17:54.840 looking back, what how much of your performance comes from 0:17:54.880 --> 0:17:57.320 being long asset classes and how much comes from being 0:17:57.359 --> 0:17:58.640 short asset classes? 0:17:58.760 --> 0:18:01.320 There any any differential there? 0:18:01.359 --> 0:18:03.640 Are you equally good at both or do they both 0:18:03.680 --> 0:18:05.040 equally feed into performance? 0:18:06.040 --> 0:18:08.679 You get good performance from both. So I'll give you 0:18:08.720 --> 0:18:12.480 some examples. So if we look at two thousand and 0:18:12.520 --> 0:18:15.399 eight at the financial crisis, that's a really good example 0:18:15.440 --> 0:18:17.800 to play through. So that year, a good portion of 0:18:17.840 --> 0:18:20.479 profits did come from long positions in bonds. Bonds did 0:18:20.560 --> 0:18:24.040 very well, but about a third of profits came from 0:18:24.119 --> 0:18:27.439