1 00:00:01,280 --> 00:00:04,725 And so, the difference between somebody who can live with a 10 point 2 00:00:04,725 --> 00:00:07,640 drawdown versus somebody who can live with a 20 point drawdown 3 00:00:07,640 --> 00:00:10,820 is not like 10% more, it's like way bigger than that depending 4 00:00:10,820 --> 00:00:13,735 on your time horizon. There are all these arbitrage alphas in 5 00:00:13,735 --> 00:00:17,180 the world and people get spent a lot of time thinking about predicting 6 00:00:17,180 --> 00:00:20,625 returns on the basis of those things. But the sizing is just such 7 00:00:20,625 --> 00:00:23,540 an important piece of the puzzle and thinking more about that, 8 00:00:23,540 --> 00:00:27,250 I think it's just a place that we're spending a lot of time thinking 9 00:00:27,250 --> 00:00:28,310 about and working on. 10 00:00:30,150 --> 00:00:41,175 Welcome to Top Traders Unplugged. In markets success doesn’t 11 00:00:41,175 --> 00:00:53,425 come from predicting what happens next, it comes from being 12 00:00:53,425 --> 00:01:00,775 prepared for what you can’t predict. In each episode we go deep 13 00:00:33,372 --> 00:00:39,279 with some of the world’s most thoughtful minds in investing, economics, 14 00:00:39,279 --> 00:00:45,186 and beyond to understand how they think, how they prepare, and 15 00:00:45,186 --> 00:00:50,556 how they decide, and the experiences that shaped how they 16 00:00:50,556 --> 00:00:55,926 see the world. No noise, no short-cuts, just real conversations 17 00:00:55,926 --> 00:01:00,759 to help you think better and invest with confidence. 18 00:01:03,820 --> 00:01:10,222 Hello and welcome to another episode of the Open Interest series 19 00:01:10,222 --> 00:01:17,206 on Top Traders Unplugged. This is episode number 23 and I'm your 20 00:01:17,206 --> 00:01:22,444 host, Moritz Seibert. I'm joined today by Charlie McGarraugh, 21 00:01:22,444 --> 00:01:28,846 the CEO of Altis Partners. Altis Partners was founded in 2000 22 00:01:28,846 --> 00:01:35,248 as a trend following CTA business and has since evolved into 23 00:01:35,248 --> 00:01:41,650 a macro focused trading firm with a focus on combining factors 24 00:01:41,650 --> 00:01:48,052 such as trend carry and reversion all into a single portfolio. 25 00:01:48,052 --> 00:01:54,454 Additionally, Altis functions as the sub advisor to two US domiciled 26 00:01:54,454 --> 00:02:02,020 ETFs that are managed by Simplify and this is now by far the 27 00:02:02,020 --> 00:02:08,422 largest part of the business, at least in terms of AUM. Charlie 28 00:01:05,088 --> 00:01:20,304 looks back on a long career at Goldman Sachs, including metals trading 29 00:01:20,304 --> 00:01:34,252 and then got involved in a sports betting business, which I 30 00:01:34,252 --> 00:01:44,396 personally find very interesting. We'll get into that 31 00:01:44,396 --> 00:02:00,880 a little bit later, I guess. And that one got sold to blockchain.com. 32 00:02:00,880 --> 00:02:08,488 Following this, Charlie acquired Altis Partners. So, Charlie, 33 00:01:06,898 --> 00:01:26,905 let me stop there because you can add much more detail and background 34 00:01:26,905 --> 00:01:45,373 and make it more interesting. Let me say welcome to Open Interest 35 00:01:45,373 --> 00:02:03,841 on Top Traders Unplugged. And I hope we'll have a very interesting 36 00:02:03,841 --> 00:02:08,458 and cool conversation. 37 00:02:09,450 --> 00:02:10,890 Awesome. Well, thanks Moritz. 38 00:02:10,970 --> 00:02:15,884 Excellent, look, I mean I gave a little bit of an intro. You have 39 00:02:15,884 --> 00:02:20,096 a… my understanding is you have a long career at Goldman. You 40 00:02:20,096 --> 00:02:23,957 were a partner at Goldman Sachs, working in the US, working 41 00:02:23,957 --> 00:02:28,871 in London. You are from the US but you moved over to London, then 42 00:02:28,871 --> 00:02:32,732 started trading metals. Give us a little bit of background on 43 00:02:32,732 --> 00:02:36,242 Charlie, yourself, and the sports betting piece and how you 44 00:02:36,242 --> 00:02:37,646 came to acquire Altis. 45 00:02:38,440 --> 00:03:49,764 Sure. I'll try to keep it pretty brief because the markets 46 00:03:49,764 --> 00:04:54,604 and the products and the strategies are the interesting part. 47 00:04:54,604 --> 00:06:18,896 And me, I'm boring. And boring is good in asset management. At least 48 00:06:18,896 --> 00:06:44,832 that's what we hope. So, I joined Goldman Sachs right out of 49 00:03:08,768 --> 00:03:54,260 undergrad. I majored in math and econ, did pretty well. I've worked 50 00:03:54,260 --> 00:04:35,961 on a number of different desks including emerging markets, and mortgage 51 00:04:35,961 --> 00:05:10,080 trading, and, ultimately, the commodity business. I started in 52 00:05:10,080 --> 00:06:03,154 New York and one day they said you're moving to London. And I thought 53 00:06:03,154 --> 00:06:44,855 that would be pretty fun and interesting adventure. And it was. 54 00:02:38,440 --> 00:03:21,919 We traded mortgages throughout the post-crisis world where there 55 00:03:21,919 --> 00:04:15,060 was a lot of distress in Europe. And eventually, again, as 56 00:04:15,060 --> 00:05:17,863 happens in these big firms, one day it was sort of made clear 57 00:05:17,863 --> 00:06:20,666 to me that management wanted me to move to the metals desk. This 58 00:06:20,666 --> 00:06:44,821 was 2014. And I did. It took quite a while to adjust to a trading 59 00:03:09,630 --> 00:03:40,820 environment in listed futures where the product is relatively simple 60 00:03:40,820 --> 00:04:08,891 but the market structure is extraordinarily complex. Very different 61 00:04:08,891 --> 00:04:36,962 than asset backed securities and structured finance where it's 62 00:04:36,962 --> 00:05:08,152 pretty easy to understand who's buying and selling stuff. It's 63 00:05:08,152 --> 00:05:33,104 a relatively specialized product but the products themselves 64 00:05:33,104 --> 00:06:07,413 are complicated. So, it’s a very different beast in the dynamic 65 00:06:07,413 --> 00:06:44,841 world of liquid macro. And that took a little bit of learning. 66 00:02:38,440 --> 00:03:11,656 Now, I realized that this is now the mid-2010s, the markets already 67 00:03:11,656 --> 00:03:33,800 had significantly electronified, and that if I was 68 00:03:33,800 --> 00:04:09,784 going to be good at this new thing it wasn't enough to understand 69 00:04:09,784 --> 00:04:43,000 sort of just valuation. I really had to have a better understanding 70 00:04:43,000 --> 00:05:21,752 of the dynamics of the market. And the old model where it was kind 71 00:05:21,752 --> 00:05:52,200 of voice trading, and a big salesforce making phone calls, was 72 00:05:52,200 --> 00:06:17,112 not enough really to understand, particularly in a market 73 00:06:17,112 --> 00:06:39,256 that had asymmetric information flows like physical commodities 74 00:06:39,256 --> 00:06:44,792 often do. At that time China was dominating the base metal flow, 75 00:03:10,440 --> 00:03:52,040 and it was very difficult to get a sense, as a Westerner sitting 76 00:03:52,040 --> 00:04:30,440 in London, of what was really going on the ground there driving 77 00:04:30,440 --> 00:05:12,040 base metals. So, it took a little while to adjust. And I kind 78 00:05:12,040 --> 00:05:50,440 of arrived at the conclusion that I really wanted to have computers 79 00:05:50,440 --> 00:06:32,040 help me trade and that a data driven approach might be the better 80 00:06:32,040 --> 00:06:44,840 way to do it. Now, I had backed the sports betting business, 81 00:02:49,328 --> 00:03:05,660 there, a few years prior to my departure from Goldman, and this 82 00:03:05,660 --> 00:03:23,353 is now 2016. I decided to leave Goldman and work on that under 83 00:03:23,353 --> 00:03:38,324 the basic theory that sports are interesting and they lend themselves 84 00:03:38,324 --> 00:03:56,017 well to statistical analysis. This will be a good way to get an 85 00:03:56,017 --> 00:04:10,988 education in this new toolkit called machine learning and AI. Like 86 00:04:10,988 --> 00:04:30,042 many things in my life, early is the same thing as wrong. I was 87 00:04:30,042 --> 00:04:49,096 a little early on that one perhaps, but it was good to get the 88 00:04:49,096 --> 00:05:01,345 education. And the sports betting business was difficult business 89 00:05:01,345 --> 00:05:19,038 because the barrier to entry is very low and because of the seasonality 90 00:05:19,038 --> 00:05:36,731 of kind of the entropy in the business, the seasons start and then 91 00:05:36,731 --> 00:05:51,702 you learn more about the players. But it just didn't really 92 00:05:51,702 --> 00:06:08,034 capture, I guess, my heart or my desire for scalability in the 93 00:06:08,034 --> 00:06:24,366 same way. Now I know there are people who do exceptionally well 94 00:06:24,366 --> 00:06:44,781 at it, but I decided by about 2018 it was time to pivot to crypto. 95 00:02:38,440 --> 00:03:26,644 Crypto had been incredibly interesting. Crypto had this really 96 00:03:26,644 --> 00:04:25,560 interesting property that the deeper you looked at it, the more 97 00:04:25,560 --> 00:05:29,832 there was and yeah, so I sold the business to blockchain.com in 98 00:05:29,832 --> 00:06:18,036 2018, and joined there, and started working on institutional 99 00:06:18,036 --> 00:06:44,816 trading with the team there. 100 00:06:45,420 --> 00:06:48,192 Was it still a sports betting business that you sold to Blockchain 101 00:06:48,192 --> 00:06:49,578 or was it a crypto business? 102 00:06:49,580 --> 00:07:13,036 No, I mean it was more of an aqua hire. It was sort of like we're 103 00:07:13,036 --> 00:07:30,628 going to pivot it from sports betting to crypto. We had pretty 104 00:07:30,628 --> 00:07:43,822 good software engineering capability, a good concept of trading, 105 00:07:43,822 --> 00:08:02,880 and let's just change it up to something that we think can scale 106 00:08:02,880 --> 00:08:20,472 and is maybe more dynamic. Now, I know people have done really 107 00:08:20,472 --> 00:08:36,598 well on sports betting, and now prediction markets are all the 108 00:08:36,598 --> 00:08:54,190 thing. So, I guess everything moves in cycles, but at the time 109 00:08:54,190 --> 00:09:02,986 it felt like the right move. Yeah, so, I was watching Stephen 110 00:06:57,746 --> 00:07:14,078 Hedgecock, who is Altis's CEO. I'm actually the CIO of Altis. Stephen 111 00:07:14,078 --> 00:07:29,049 Hedgecock had been a board member of the betting company. And 112 00:07:29,049 --> 00:07:46,742 he called me and let me know that his partner, who had founded 113 00:07:46,742 --> 00:08:01,713 Altis, had died in an avalanche while skiing in 2019. And 114 00:08:01,713 --> 00:08:20,767 Stephen was saying to me, we have a lot of IP in this business. 115 00:08:20,767 --> 00:08:37,099 It's a pretty interesting business. Do you want to have a look? 116 00:08:37,099 --> 00:08:57,514 And so, I started looking at it in late 2019 and by 2020 I acquired 117 00:08:57,514 --> 00:09:02,958 a stake in it. The theory at the time being that the math of how 118 00:07:18,356 --> 00:07:41,900 do you trade leverage stuff competently is directly applicable 119 00:07:41,900 --> 00:08:15,908 to crypto just as much as in futures, which I think it probably 120 00:08:15,908 --> 00:08:47,300 still is. So, we thought maybe we'd build an asset manager using 121 00:08:47,300 --> 00:09:02,996 Altis's IP in a JD blockchain. But in any case, I acquired the stake 122 00:06:58,620 --> 00:07:13,310 in Altis in March of 2020, right as COVID hit, and then started 123 00:07:13,310 --> 00:07:30,260 work with the team on a carve out of my time, working on the systems 124 00:07:30,260 --> 00:07:43,820 and then ultimately ended up moving to Altis full time in ‘22 125 00:07:43,820 --> 00:07:58,510 as the crypto markets were pretty bumpy and I felt like a pivot 126 00:07:58,510 --> 00:08:04,160 to re-materialization, re-industrialization, and TradFi 127 00:08:04,160 --> 00:08:18,850 was going to be a thing, and that macro volatility was going to 128 00:08:18,850 --> 00:08:33,540 start to pick up again, which it clearly has. So, that's the story 129 00:08:33,540 --> 00:08:45,970 of the long journey into Altis. But basically, I bought a 130 00:08:45,970 --> 00:09:00,660 stake in Altis. Altis had been founded in 2000 as a trend follower 131 00:09:00,660 --> 00:09:02,920 with some... 132 00:09:03,160 --> 00:09:05,214 Yeah, tell us a bit about the history. You started, so more than 133 00:09:05,214 --> 00:09:07,110 a quarter of a century ago that was obviously not with you. 134 00:09:07,110 --> 00:09:08,690 You were maybe still in college, maybe. You were already 135 00:09:08,690 --> 00:09:09,796 at Goldman, but it wasn't with you. 136 00:09:17,160 --> 00:09:30,552 Yeah, Altis was founded the same year I joined Goldman, and Stephen 137 00:09:30,552 --> 00:09:41,712 Hedgecock, and Natasha Reeve Gray, and Alex Brunwin, and Zbish 138 00:09:41,712 --> 00:09:50,640 Hermaszewski were the four founding partners. Natasha moved 139 00:09:50,640 --> 00:10:02,916 on sometime later and then Zbish tragically died in 2019. But 140 00:10:02,916 --> 00:10:18,540 back in 2000, they… It was kind of the heyday of CTAs, and they 141 00:10:18,540 --> 00:10:31,932 built the business up over the course of that decade for 2000-2009, 142 00:10:31,932 --> 00:10:45,324 let's say. And it was really focused on trend following with a 143 00:10:45,324 --> 00:10:55,368 couple of alternative indicators in there, but really a 144 00:10:55,368 --> 00:11:06,528 tape reader, basically, with some pretty good IP around portfolio 145 00:11:06,528 --> 00:11:18,804 construction that, to this day, I think is fit for purpose. 146 00:09:17,160 --> 00:09:32,214 And then the business grew and grew over the 2000s. And then, as 147 00:09:32,214 --> 00:09:42,636 you will have undoubtedly discussed on previous episodes, trend 148 00:09:42,636 --> 00:09:55,374 following was terrible in the 2010s. The amount of liquidity provided 149 00:09:55,374 --> 00:10:08,112 to the government to the market via QE and the volatility 150 00:10:08,112 --> 00:10:18,534 suppression. You basically had a governmental complex globally saying 151 00:10:18,534 --> 00:10:32,430 here's giant piles of low velocity free money and the job is 152 00:10:32,430 --> 00:10:45,168 to suppress volatility. And that was just kind of inimical to 153 00:10:45,168 --> 00:10:59,064 the returns of the trend factor. So, Altis had a tough run 154 00:10:59,064 --> 00:11:14,118 of performance in the 2010s. It took them a little while to kind 155 00:11:14,118 --> 00:11:18,750 of figure this out. By the mid 2010s they were pretty aggressively 156 00:09:37,440 --> 00:10:05,325 refactoring the system and thinking through how to be a bit 157 00:10:05,325 --> 00:10:33,210 more multidimensional in terms of the risk premia that they were 158 00:10:33,210 --> 00:11:01,095 attempting to access, and more advances on risk management and portfolio 159 00:11:01,095 --> 00:11:18,840 construction, and then they did pretty well. But then Zbish died 160 00:09:29,968 --> 00:10:14,796 in 2019 while skiing and so that kind of put the kibosh on that 161 00:10:14,796 --> 00:10:59,624 part of the firm's journey. And then I sort of showed up in the 162 00:10:59,624 --> 00:11:18,836 picture and bought the CIO stake. 163 00:11:20,951 --> 00:11:25,631 So back then, I think you've just mentioned, it was focused on 164 00:11:25,631 --> 00:11:28,751 trend following with some advanced techniques around portfolio 165 00:11:28,751 --> 00:11:32,651 construction. And then, I think after you've joined, you introduced 166 00:11:32,651 --> 00:11:37,721 a new program, a new trading strategy, which I think is the enhanced 167 00:11:37,721 --> 00:11:42,011 macro programmer, if I'm not mistaken, that's the name. So, is 168 00:11:42,011 --> 00:11:45,521 this still anchored in systematic time series momentum or 169 00:11:45,521 --> 00:11:49,421 trend following or is it now more of a combination… 170 00:11:49,461 --> 00:11:50,341 It absolutely is. 171 00:11:50,661 --> 00:11:53,949 Of all sorts of different things like a multi strat quant type 172 00:11:53,949 --> 00:11:54,497 of program? 173 00:11:55,511 --> 00:12:45,288 It's a bit of both. I'd say it's an evolution of the long-standing 174 00:12:45,288 --> 00:13:27,407 trend program that had existed at Altis called Global Futures Portfolio. 175 00:13:27,407 --> 00:14:05,697 So, it was an addition of additional factor return prediction 176 00:14:05,697 --> 00:14:40,158 systems inside the context of a similar portfolio construction 177 00:14:40,158 --> 00:14:47,816 IP layer. So yes, it is very much a sister or sibling product 178 00:12:14,277 --> 00:12:33,043 to the product that has always been Altis, which is fundamentally 179 00:12:33,043 --> 00:12:53,515 a trend CTA, but it attempts to address that problem, which I 180 00:12:53,515 --> 00:13:13,987 think is familiar to any kind of CTA and systematic trader which 181 00:13:13,987 --> 00:13:34,459 is trend following is great. It's got scale, it's got anti correlation 182 00:13:34,459 --> 00:13:51,519 to global growth beta, typically, in crisis alpha, but it's 183 00:13:51,519 --> 00:14:11,991 also episodic in its returns and it's a bit difficult to stick 184 00:14:11,991 --> 00:14:32,463 with. It's kind of like eating your vegetables. Like, you know it's 185 00:14:32,463 --> 00:14:47,817 good for you, but it doesn't always taste great. And so, like 186 00:12:02,138 --> 00:12:26,437 many people in this space, we're sort of thinking about like, 187 00:12:26,437 --> 00:12:55,154 do we want to break out of that kind of intellectual rut of 188 00:12:55,154 --> 00:13:19,453 pure trend plus nothing else kind of thing and start thinking 189 00:13:19,453 --> 00:13:43,752 about blending other types of risk premia into the system? The 190 00:13:43,752 --> 00:14:05,842 IP is quite fungible. If you've got decent predictors for 191 00:14:05,842 --> 00:14:32,350 the variants that you might see in the market then you might 192 00:14:32,350 --> 00:14:47,813 try to access them in some way. So, I kind of look at it, it's 193 00:12:10,999 --> 00:12:34,231 like a game. It's like every day there's this massive cross section 194 00:12:34,231 --> 00:12:53,591 of movement that happens across all these things that you 195 00:12:53,591 --> 00:13:16,823 could trade and mostly the market is noise. So, it's sort of 196 00:13:16,823 --> 00:13:36,183 well documented that momentum maybe has some (you should pardon 197 00:13:36,183 --> 00:13:57,479 the pun) persistent ability to explain some of that cross section 198 00:13:57,479 --> 00:14:18,775 of that noise, or that movement which is mostly noise. But 199 00:14:18,775 --> 00:14:43,943 could there be other sort of known things in there? So, that was 200 00:14:43,943 --> 00:14:47,815 one thing. And if you got decent predictions, then it's easy 201 00:12:11,171 --> 00:12:30,311 enough to start thinking about risk managing around them and sizing 202 00:12:30,311 --> 00:12:51,191 them up and blending them into the optimizer. And so that's really 203 00:12:51,191 --> 00:13:10,331 kind of the approach we've taken, which is like, hey, there's 204 00:13:10,331 --> 00:13:29,471 nothing really new under the sun. It's sort of like positioning 205 00:13:29,471 --> 00:13:52,091 momentum and value. And we want to start to do more than just 206 00:13:52,091 --> 00:14:06,011 deliver the unadulterated momentum because we think ultimately 207 00:14:06,011 --> 00:14:21,671 what clients want is returns, and anti-correlation, and capital 208 00:14:21,671 --> 00:14:40,811 efficiency. So, I'm just trying to get in there and predict 209 00:14:40,811 --> 00:14:47,771 more, basically, without overfitting. 210 00:14:48,781 --> 00:14:52,141 How many markets do you guys trade in that program? 211 00:14:53,661 --> 00:14:58,271 Yeah, we're trading something like 150 ish markets across many, 212 00:14:58,271 --> 00:15:01,498 many regulated exchanges all over the world. 213 00:15:02,231 --> 00:15:06,227 And that includes, and maybe there's this, I was asking this maybe 214 00:15:06,227 --> 00:15:10,556 out of schedule because it will lead to a question that I will 215 00:15:10,556 --> 00:15:14,552 later have related to the ETFs that you're sub advising on. So, 216 00:15:14,552 --> 00:15:17,549 in that futures portfolio you’re trading, all the markets, 217 00:15:17,549 --> 00:15:19,880 equities, FX, bonds, commodities, everything across the 218 00:15:19,880 --> 00:15:20,213 board. 219 00:15:20,311 --> 00:15:25,549 Equity index, fixed income, STIR, FX, commodities, yeah, independent 220 00:15:25,549 --> 00:15:30,787 bet count like so many others in the space. 221 00:15:31,761 --> 00:15:36,345 Yes. Okay, got it. So, this is then probably a differentiator to 222 00:15:36,345 --> 00:15:41,693 the performance in one of the ETFs or in both of the ETFs that 223 00:15:41,693 --> 00:15:45,895 you sub advise on. Because these ETFs are not allocating their 224 00:15:45,895 --> 00:15:49,715 risk across all of these markets that you've just mentioned, 225 00:15:49,715 --> 00:15:52,389 but only to a subsection of them. 226 00:15:52,561 --> 00:15:57,805 But yeah, I would say without getting too deep into any specific 227 00:15:57,805 --> 00:16:01,738 product. The investable universe that one selects can vary 228 00:16:01,738 --> 00:16:03,486 across products for sure. 229 00:16:03,901 --> 00:16:08,881 Yes, yes. Before we go into this more from a holistic perspective, 230 00:16:08,881 --> 00:16:13,861 when you look at our industry, the CTA trend following hedge fund 231 00:16:13,861 --> 00:16:18,841 industry, if you will, what would you say are the most important 232 00:16:18,841 --> 00:16:22,991 themes these days? What is changing? What is strategically important? 233 00:16:22,991 --> 00:16:27,556 What's happening right now that kind of like gets your attention? 234 00:16:28,761 --> 00:17:17,397 It's a great question and there's a lot happening in the market. 235 00:17:17,397 --> 00:18:01,980 Right? And I think the structure of the business and the 236 00:18:01,980 --> 00:18:54,669 curation of the product set is as interesting as I guess, the underlying 237 00:18:54,669 --> 00:19:39,252 trading systems themselves. So, if you think of kind of like 238 00:19:39,252 --> 00:20:23,835 a design space of what kinds of investment strategies could a 239 00:20:23,835 --> 00:21:16,524 person build or invest in, it's like the thing I like to think 240 00:21:16,524 --> 00:21:48,948 about the most is reactiveness to new information. There's an entire 241 00:16:38,871 --> 00:17:15,941 school of thought out there, passive acolytes who are like, hey, 242 00:17:15,941 --> 00:17:49,641 anything you learn, the market's so smart that anything you 243 00:17:49,641 --> 00:18:30,081 learn has no value. You should not attempt to react to information 244 00:18:30,081 --> 00:19:10,521 because if you do, you'll only hurt yourself. Right? And so like, 245 00:19:10,521 --> 00:19:54,331 that would be like one extreme view, which is like the change in 246 00:19:54,331 --> 00:20:31,401 my behavior relative to a change in new information is zero. 247 00:20:31,401 --> 00:21:01,731 And that's basically like dollar cost averaging passive. And 248 00:21:01,731 --> 00:21:42,171 amazingly, something like 60% of the market is on that theory of 249 00:21:42,171 --> 00:21:48,911 the case. At the other end of the design spectrum is sort of how 250 00:17:14,505 --> 00:17:56,437 much will I react to information? I'll react a lot. Every 251 00:17:56,437 --> 00:18:42,181 time anything changes, I'll just redo my whole book. And we know 252 00:18:42,181 --> 00:19:24,113 that if you're that reactive, you're probably going to bleed out 253 00:19:24,113 --> 00:20:02,233 on transaction costs, and pay too many commissions, and slippage. 254 00:20:02,233 --> 00:20:44,165 You're also probably going to overreact and, if you're using computers 255 00:20:44,165 --> 00:21:33,721 to do it, overfit. Right? So, I guess my view is like, there's 256 00:21:33,721 --> 00:21:48,969 probably a happy middle. And interestingly enough, there's not 257 00:16:53,016 --> 00:17:51,228 that many people in that happy middle. So, CTA and futures trading 258 00:17:51,228 --> 00:18:39,738 are really interesting because futures are great for reacting to 259 00:18:39,738 --> 00:19:23,397 changing information because they're really efficient way to move 260 00:19:23,397 --> 00:20:21,609 a lot of risk around. And they're not just efficient from a 261 00:20:21,609 --> 00:21:05,268 sort of transaction cost and transparency perspective, but they're 262 00:21:05,268 --> 00:21:48,927 also efficient from a capital efficiency and leverage perspective. 263 00:16:28,761 --> 00:18:05,572 It's kind of like the only way that normal people can get access 264 00:18:05,572 --> 00:19:34,936 to borrowing as cheap as like a bank can, just synthetically, by 265 00:19:34,936 --> 00:21:04,300 virtue of the way the futures are set up, you're basically getting 266 00:21:04,300 --> 00:21:48,982 access to institutional funding, market funding. And that's 267 00:16:41,319 --> 00:17:44,109 like, that's pretty great. Most normal retail people, even trading 268 00:17:44,109 --> 00:18:53,178 a merger account, it's like pretty expensive to get access to 269 00:18:53,178 --> 00:19:55,968 leverage. So, I think that historically, maybe the CTA industry 270 00:19:55,968 --> 00:21:11,316 hasn't called out the value of that dynamic leverage enough or the 271 00:21:11,316 --> 00:21:48,990 value of that capital efficiency enough. You're starting 272 00:16:36,037 --> 00:17:19,693 to see a focus on things like return stacking and portable alpha. 273 00:17:19,693 --> 00:18:06,987 But the direction of travel for me, that I think is most interesting 274 00:18:06,987 --> 00:18:43,367 is this question of capital efficiency. It's like capital efficiency 275 00:18:43,367 --> 00:19:16,109 combined with dynamism and reactivity. It seems pretty dumb 276 00:19:16,109 --> 00:20:03,403 to just be like, my view will never change no matter what information 277 00:20:03,403 --> 00:20:47,059 I'm learning, yet that's 60% of the flows in the market. It's 278 00:20:47,059 --> 00:21:27,077 like passive financial assets. And under that theory you'd be buying 279 00:21:27,077 --> 00:21:48,905 bonds with a negative yield. Right? Also, it’s like obviously 280 00:16:53,393 --> 00:18:07,289 terrible to just trade way too much and die of transaction cost 281 00:18:07,289 --> 00:19:21,185 and overfit. Right? So, that middle case of sort of what's the 282 00:19:21,185 --> 00:20:22,765 optimal amount of reaction to changing information? It seems like 283 00:20:22,765 --> 00:21:30,503 it should be related to whatever the cost of changing your 284 00:21:30,503 --> 00:21:48,977 mind actually is. And so, from that perspective, I think there's 285 00:17:07,569 --> 00:18:05,781 a story to tell about how much adaptiveness is the right amount 286 00:18:05,781 --> 00:19:03,993 of adaptiveness and how much should I be willing to pay transaction 287 00:19:03,993 --> 00:20:07,056 costs for the ability to adapt? And what I would kind of offer 288 00:20:07,056 --> 00:21:05,268 out there for people is this world seems like it's changing a 289 00:21:05,268 --> 00:21:48,927 lot faster than it did during the QE era. So, wherever you were 290 00:16:50,469 --> 00:17:50,166 in that design spectrum before between, I never changed my mind 291 00:17:50,166 --> 00:19:00,717 at all, and I changed my mind too much. You're moving more, now, 292 00:19:00,717 --> 00:20:11,268 over the last five years, in the direction of, I want to adapt 293 00:20:11,268 --> 00:21:05,538 more and be more adaptive because the amount of fundamental 294 00:21:05,538 --> 00:21:48,954 volatility in the world is just picking up. You've got geopolitical 295 00:16:41,400 --> 00:17:23,530 stuff, crucially, versus pre COVID and really, before the election 296 00:17:23,530 --> 00:18:14,086 of Trump one and Brexit, the government was in the business of 297 00:18:14,086 --> 00:18:47,790 volatility suppression, just globally. And now the government's 298 00:18:47,790 --> 00:19:25,707 in the business of volatility inducement. Right? Starting wars, 299 00:19:25,707 --> 00:20:12,050 trade policy, tariff, you name it. They're in there and they're 300 00:20:12,050 --> 00:21:02,606 stirring the pot. And that means you need to change your mind 301 00:21:02,606 --> 00:21:32,097 because the fundamental conditions that you're operating 302 00:21:32,097 --> 00:21:48,949 under are meaningfully changing. And so, my view is adaptive, 303 00:17:01,887 --> 00:18:02,618 is interesting. And there's no place more efficient to be adaptive 304 00:18:02,618 --> 00:19:03,349 in all kinds of things that matter than the futures markets. 305 00:19:03,349 --> 00:20:04,080 And for that reason there's this sort of latent demand, that 306 00:20:04,080 --> 00:20:59,290 people don't even know they have yet, for adaptive investment 307 00:20:59,290 --> 00:21:48,979 product because the world's moving faster than ever before. And 308 00:16:33,030 --> 00:17:24,258 yeah, and so, when I think about the structure of the market, 309 00:17:24,258 --> 00:18:15,486 it's like how can that value prop be delivered to which segments 310 00:18:15,486 --> 00:18:53,907 of the potential total addressable market? And it's not 311 00:18:53,907 --> 00:19:32,328 just high net worth individuals or family offices or 312 00:19:32,328 --> 00:20:15,018 institutions who buy high fee hedge fund product, although that 313 00:20:15,018 --> 00:20:57,708 is an important part of the business. There's also opportunity 314 00:20:57,708 --> 00:21:40,398 in other distribution channels for various flavors of adaptive and 315 00:21:40,398 --> 00:21:48,936 efficient, basically. 316 00:21:50,511 --> 00:21:56,858 Gotcha. What I'd like to understand a little bit better is 317 00:21:56,858 --> 00:22:03,782 how it came about or how it happened. What triggered your involvement 318 00:22:03,782 --> 00:22:11,283 as a sub advisor to the ETF space because you started in 2000, 319 00:22:11,283 --> 00:22:18,784 or Altis started in 2000 as a CTA, like many others, offering their 320 00:22:18,784 --> 00:22:24,554 products I guess to private investors, high net worth individuals, 321 00:22:24,554 --> 00:22:32,632 and so on, and so forth, in a private fund or in SMAs segregated 322 00:22:32,632 --> 00:22:40,133 managed accounts. And now most of your AUM are in ETFs you're sub 323 00:22:40,133 --> 00:22:44,172 advising on with Simplify in the US. So, how did this happen? 324 00:21:54,841 --> 00:22:06,099 Was this an intentional change to say okay, we're going to go down 325 00:22:06,099 --> 00:22:17,357 the ETF route because this is where the money's going to be, this 326 00:22:17,357 --> 00:22:29,481 is the fastest growing part of the market, or was it more like happenstance, 327 00:22:29,481 --> 00:22:41,605 you ran into the right people at the right point in time and it 328 00:22:41,605 --> 00:22:44,203 happened that way? 329 00:22:44,951 --> 00:23:12,909 Yeah, I mean it's kind of a little both. So, they say luck is 330 00:23:12,909 --> 00:23:32,879 a combination of sort of preparation and I guess something 331 00:23:32,879 --> 00:23:56,843 else, I don't even know what they say, but basically there was 332 00:23:56,843 --> 00:24:26,798 an element of luck to it. I ran into Mike Green at a macro conference 333 00:24:26,798 --> 00:24:54,756 at Bretton Woods in summer of ‘21 and Mike had been a client of 334 00:24:54,756 --> 00:25:22,714 mine at Goldman long ago and a good friend as a fellow student in 335 00:25:22,714 --> 00:25:26,708 the market. And we got to talking and he was like oh, I've 336 00:23:08,370 --> 00:23:33,918 joined this new active ETF firm and I'd really like to build 337 00:23:33,918 --> 00:24:05,853 a bunch of cool things and one of the things I'd like to build is 338 00:24:05,853 --> 00:24:31,401 a managed futures ETF. And I was like, well, funnily enough, Mike, 339 00:24:31,401 --> 00:24:54,820 I just bought like a futures platform kind of like serendipitously 340 00:24:54,820 --> 00:25:24,626 and I think it's got pretty good IP in it, so maybe we should 341 00:25:24,626 --> 00:25:26,755 collaborate. And we did and then six months later we were launching 342 00:23:02,221 --> 00:23:22,631 product on the stock exchange, so that said I was pretty open minded 343 00:23:22,631 --> 00:23:41,471 about it, and that was the preparation side which is, I had 344 00:23:41,471 --> 00:23:54,031 worked in fintech with direct-to-consumer stuff in crypto. 345 00:23:54,031 --> 00:24:11,301 I think I understood pretty well at that point (after having 346 00:24:11,301 --> 00:24:33,281 been in the tech industry for a good, I don't know what is it, 347 00:24:33,281 --> 00:24:47,411 six years after leaving Goldman) about how platforms can 348 00:24:47,411 --> 00:25:04,681 scale and what the unit economics can look like when you're 349 00:25:04,681 --> 00:25:18,811 building a scalable direct distribution business that is powered 350 00:25:18,811 --> 00:25:26,661 by the leverage of software. And I felt like I was not as precious 351 00:22:56,687 --> 00:23:09,727 about only having a single distribution channel. I just sort 352 00:23:09,727 --> 00:23:26,679 of felt like the firm has just one product which is trading IP, 353 00:23:26,679 --> 00:23:39,719 and there are lots of different means of packaging that 354 00:23:39,719 --> 00:23:54,063 IP, and not all packagings are created equal. Like some packagings 355 00:23:54,063 --> 00:24:08,407 lend themselves better to scale but are a little bit lower 356 00:24:08,407 --> 00:24:22,751 resolution. And then other packagings can get to a much more 357 00:24:22,751 --> 00:24:34,487 granular trading approach and are more suitable for complexity. 358 00:24:34,487 --> 00:24:50,135 And broadly defined, I'd say that it would be like mass market 359 00:24:50,135 --> 00:25:03,175 listed product versus hedge funds. Like there are just things 360 00:25:03,175 --> 00:25:22,735 you can do inside a hedge fund or an SMA that you just can't do 361 00:25:22,735 --> 00:25:26,647 inside an ETF. But ETFs have this incredibly low client acquisition 362 00:24:32,823 --> 00:25:26,759 friction and they scale. And so, if it's a scalable risk premia 363 00:23:25,399 --> 00:24:31,127 that you think can add value to the client base and it's comprehensible, 364 00:24:31,127 --> 00:25:26,743 then I think an ETF is a great package for it. 365 00:25:28,361 --> 00:25:35,291 So, in 2021, when you ran into Mike at Bretton Woods and you started 366 00:25:35,291 --> 00:25:41,231 to think about collaborating on an ETF project, how did you go 367 00:25:41,231 --> 00:25:43,211 about designing that ETF? 368 00:25:44,181 --> 00:25:45,501 Right. So here I'm going to… 369 00:25:45,501 --> 00:25:49,263 Because at the time is you already had replicating ETFs out 370 00:25:49,263 --> 00:25:54,051 there. And so, that would have been an option, but I think you decided 371 00:25:54,051 --> 00:25:54,735 against that. 372 00:25:54,981 --> 00:26:30,718 Yeah, it's a great question. I think this is like where the experience 373 00:26:30,718 --> 00:27:00,957 on the tech side really is relevant. In the financial industry 374 00:27:00,957 --> 00:27:39,443 we sort of talk about the word PM, and it means something very specific. 375 00:27:39,443 --> 00:28:09,682 It means portfolio manager. But all these crazy tech people running 376 00:28:09,682 --> 00:28:42,670 around, they also use the word PM, and it means product manager. 377 00:25:54,981 --> 00:26:41,781 Being a product manager I think is equally or even more important 378 00:26:41,781 --> 00:27:28,581 than being, I mean… The asset side is always important. It's a 379 00:27:28,581 --> 00:28:03,681 returns-based business. Being a portfolio manager is always important, 380 00:28:03,681 --> 00:28:42,681 and managing risk is always important. Right? Obviously, no doubt. 381 00:25:54,981 --> 00:26:29,041 But the liability side of the business, which is how do we build 382 00:26:29,041 --> 00:26:57,861 a product that potentially services a need in a particular segment 383 00:26:57,861 --> 00:27:26,681 of the total addressable market? I think is also a really 384 00:27:26,681 --> 00:27:52,881 interesting question. And so, without getting too heavy into the 385 00:27:52,881 --> 00:28:21,701 specifics of any one ticker, again, there's a spectrum of design 386 00:28:21,701 --> 00:28:42,661 choices that you can kind of think about. One of them that I think 387 00:26:07,401 --> 00:26:32,241 is really important is if you're going to build a liquid alternative 388 00:26:32,241 --> 00:26:55,011 product, is it designed to be an uncorrelated absolute return stream 389 00:26:55,011 --> 00:27:21,921 or is it designed to be a statistical hedge of some kind or 390 00:27:21,921 --> 00:27:40,551 even a contractual hedge? So defined outcome, versus statistical 391 00:27:40,551 --> 00:27:57,111 outcome, versus and hedge versus just uncorrelated returns. 392 00:27:57,111 --> 00:28:19,881 Those are all different choices that you can make on the 393 00:28:19,881 --> 00:28:42,651 menu and turn the knobs in terms of product design accordingly. 394 00:25:54,981 --> 00:26:20,331 And one of the things that tech people will say is like, you're 395 00:26:20,331 --> 00:26:47,631 always trying to think of like jobs to be done, like what is the 396 00:26:47,631 --> 00:27:11,031 job that needs to be fulfilled by a particular product? And then, 397 00:27:11,031 --> 00:27:36,381 so, it's sort of like you have the technology at the center, then 398 00:27:36,381 --> 00:27:57,831 you have the product that you've built to hopefully solve an 399 00:27:57,831 --> 00:28:15,381 actual problem from the client's perspective. Then you have 400 00:28:15,381 --> 00:28:36,831 a go-to-market strategy, and then you have a business built around 401 00:28:36,831 --> 00:28:42,681 the go-to-market strategy. And so, for us it was a case of like, 402 00:26:18,941 --> 00:26:50,089 if we're going to try to build things that are number one, capital 403 00:26:50,089 --> 00:27:11,653 efficient, and number two, really pull out the diversification 404 00:27:11,653 --> 00:27:38,009 side of the puzzle, obviously you want more return, more diversification 405 00:27:38,009 --> 00:28:01,969 and more hedge like properties. Those are all good things. 406 00:28:01,969 --> 00:28:28,325 But everything comes in a package of tradeoffs. And it's like, 407 00:28:28,325 --> 00:28:42,701 how do you think through that? So, there was a long conversation 408 00:26:16,857 --> 00:27:00,609 between my team and Mike's team around some of the choices that 409 00:27:00,609 --> 00:27:33,423 we could make. And that involved curating the investable 410 00:27:33,423 --> 00:28:09,883 universe and thinking about the different types of alphas and 411 00:28:09,883 --> 00:28:42,697 risk management methodologies that could be on offer there. So, 412 00:26:03,808 --> 00:27:49,732 that's kind of one thing. And then I think you asked something 413 00:27:49,732 --> 00:28:42,694 about replication versus our internal stuff. 414 00:28:42,701 --> 00:28:47,681 Yeah. Because you run your own system, your own signals. 415 00:28:48,161 --> 00:28:59,861 Yeah. I mean look, the cool thing about software is it scales. 416 00:28:59,861 --> 00:29:11,561 Right? Once you build it, its marginal cost of deployment is relatively 417 00:29:11,561 --> 00:29:21,311 low, its opportunity cost, if it's a capacity constraint strategy, 418 00:29:21,311 --> 00:29:31,061 its opportunity cost of deployment may be very high. But 419 00:29:31,061 --> 00:29:42,761 if it's a scalable thing that can be delivered with software, I 420 00:29:42,761 --> 00:29:53,486 guess my view was eventually somebody else will deliver it if 421 00:29:53,486 --> 00:30:02,261 they're any good, so we might as well. Right? Again, the firm has 422 00:28:52,865 --> 00:29:09,329 one product that's trading IP and our goal is to try to put as 423 00:29:09,329 --> 00:29:21,089 much value down every distribution pipe as the pipe can 424 00:29:21,089 --> 00:29:32,849 handle. Right? From the packaging of those, sort of daily 425 00:29:32,849 --> 00:29:45,785 rebalance ETF only allows for certain types of things, but again 426 00:29:45,785 --> 00:29:59,897 maximum value for the package and then just try to create value 427 00:29:59,897 --> 00:30:02,249 for clients. And so, I wasn't very precious about it. And maybe 428 00:28:56,981 --> 00:29:08,447 also my background as sort of an outsider to the system. I've trained, 429 00:29:08,447 --> 00:29:20,795 I'd grown up on the sell side. I'd grown up cat hole structure and 430 00:29:20,795 --> 00:29:27,851 derivatives valuation stuff, not really time series optimization 431 00:29:27,851 --> 00:29:40,199 stuff. I think it just made me a little bit more of an outsider. 432 00:29:40,199 --> 00:29:53,429 And it was like, oh well, we have this IP, we might as well deploy 433 00:29:53,429 --> 00:30:02,249 it in new and interesting ways. And so, we did. 434 00:30:02,291 --> 00:30:06,971 Give us a little bit more background. I mean the first ETF 435 00:30:06,971 --> 00:30:10,481 you started incorporation with, Simplify. I think you're only 436 00:30:10,481 --> 00:30:13,991 referencing the commodity markets and the fixed income markets 437 00:30:13,991 --> 00:30:17,891 if I'm not mistaken. So, you're missing currencies and you're 438 00:30:17,891 --> 00:30:21,791 missing equities. What were the driving factors for that decision? 439 00:30:21,791 --> 00:30:25,301 Because these are important markets, they are liquid markets, 440 00:30:25,301 --> 00:30:29,981 they scale they have momentum. But you razed them from the portfolio. 441 00:30:33,042 --> 00:30:53,216 Yeah, I mean again it's a tradeoff between correlation to global 442 00:30:53,216 --> 00:31:15,224 beta and returns. It starts with a perspective of where can I 443 00:31:15,224 --> 00:31:37,232 add value incrementally to the client? Right? If the client is an 444 00:31:37,232 --> 00:31:57,406 advisor who's already got clients on their side who are chock-to-the-gills 445 00:31:57,406 --> 00:32:17,580 full of financial assets, then just stacking more equity beta in 446 00:32:17,580 --> 00:32:37,754 there, even if it has a positive return associated with it, 447 00:32:37,754 --> 00:33:01,596 may not be the way to really give them a distilled value proposition 448 00:33:01,596 --> 00:33:19,936 versus other stuff. You want to differentiate and basically be 449 00:33:19,936 --> 00:33:45,612 like, here we are. And so, for some of the products it's like basically 450 00:33:45,612 --> 00:34:00,284 capital efficient diversification is sort of what you're 451 00:34:00,284 --> 00:34:03,952 calling out. Now, I think it's a challenge in the market, and this 452 00:31:04,821 --> 00:31:33,711 is true actually on the institutional side too, not just 453 00:31:33,711 --> 00:32:08,379 in the sort of mass market channel, which is that no matter 454 00:32:08,379 --> 00:32:45,936 how much people talk about the value of alpha, and the value of 455 00:32:45,936 --> 00:33:09,048 anti-correlation, and the value of non-recourse leverage, at 456 00:33:09,048 --> 00:33:46,605 the end of the day they also want to see absolute returns. It 457 00:33:46,605 --> 00:34:03,939 just makes every conversation less complicated. And even if you're 458 00:30:41,234 --> 00:30:57,618 working with an incredibly sophisticated advisor or capital 459 00:30:57,618 --> 00:31:18,098 allocator as your client, they will invariably have some committee 460 00:31:18,098 --> 00:31:40,626 or third parties that they have to explain the complexity to. 461 00:31:40,626 --> 00:32:01,106 And so, everything is just defricationalized to the extent it's 462 00:32:01,106 --> 00:32:31,826 got a nice return on it. And that's why I think a lot of kind 463 00:32:31,826 --> 00:32:58,450 of bundled products that just have a lot of equity beta in them 464 00:32:58,450 --> 00:33:20,978 tend to be attractive for business builders in the liquid space. 465 00:33:20,978 --> 00:33:43,506 Because it's just like, okay, well, you're kind of playing the 466 00:33:43,506 --> 00:34:03,986 game on easy mode in terms of putting up returns. And that said, 467 00:30:42,930 --> 00:31:25,778 my view is that, in the future, more and more of the portfolio 468 00:31:25,778 --> 00:32:02,034 construction process is going to be carried out by machines that 469 00:32:02,034 --> 00:32:31,698 are looking for mean variance, optimization of variance flavors. 470 00:32:31,698 --> 00:33:01,362 And trustworthy brands will matter and headline returns will 471 00:33:01,362 --> 00:33:31,026 matter, but also the covariance properties of your product 472 00:33:31,026 --> 00:34:03,986 will matter a lot. And probably, increasingly on the margin. 473 00:30:33,042 --> 00:31:03,974 And what's really cool about this sort of active strategy space 474 00:31:03,974 --> 00:31:40,530 is that unlike passive ETFs, which are just a race to the bottom 475 00:31:40,530 --> 00:32:17,086 on fees and just a pure economy of scale game, you can actually 476 00:32:17,086 --> 00:32:50,830 differentiate a lot as an active manager as a small guy. Right? 477 00:32:50,830 --> 00:33:24,574 Because there's a huge amount of choices to make in the design 478 00:33:24,574 --> 00:34:01,130 space of types of stuff to build and what its value prop might 479 00:34:01,130 --> 00:34:03,942 be. My theory is recommendation engines and recommendation 480 00:30:51,970 --> 00:31:16,306 systems are likely to become more important. Agentic allocation 481 00:31:16,306 --> 00:31:46,050 is likely to become more important. And in that world the 482 00:31:46,050 --> 00:32:15,794 numbers will speak. And people care about the numbers that will 483 00:32:15,794 --> 00:32:42,834 matter which are liquidity, length of track, correlation to other 484 00:32:42,834 --> 00:33:12,578 assets in the portfolio, and the higher moments of the covariance 485 00:33:12,578 --> 00:33:36,914 properties just sort of generally out there, and liquidity. 486 00:33:36,914 --> 00:34:03,954 And so, yeah, we're just trying to build for that. And you 487 00:30:43,086 --> 00:31:53,394 can do that in different ways. You can do it by curating the investable 488 00:31:52,323 --> 00:32:53,658 universe, you can do it by curating the signals, and you can 489 00:32:53,658 --> 00:33:33,834 do it by curating the portfolio construction methodologies. 490 00:33:33,834 --> 00:34:03,966 I'm sure there's other ways too. 491 00:34:04,392 --> 00:34:06,345 Or all three of them at the same time. 492 00:34:06,352 --> 00:35:10,449 Yeah, or other stuff we haven't even thought of. Right? But 493 00:35:10,449 --> 00:35:39,584 there's always more to do. But anyway, that's how I think about 494 00:34:12,071 --> 00:34:22,692 it. And I think, when you think, just again, to make a kind 495 00:34:22,692 --> 00:34:30,862 of a comment about the business on the distribution side, 496 00:34:30,862 --> 00:34:39,032 when you think about like those advisors, I would recommend 497 00:34:39,032 --> 00:34:46,385 anybody who's kind of interested in this scalable distribution 498 00:34:46,385 --> 00:34:57,006 stuff, go spend some time with some advisors on either side of the 499 00:34:57,006 --> 00:35:05,993 pond because what you'll hear, when you actually start talking about 500 00:35:05,993 --> 00:35:14,163 what's valuable for them and what's valuable to their clients, 501 00:35:14,163 --> 00:35:23,150 it's like half the job of the advisor is psychotherapist because 502 00:35:23,150 --> 00:35:32,954 no matter how well off you are, money makes you anxious because 503 00:35:32,954 --> 00:35:39,490 the cost of screwing up is really high. And so, a big piece 504 00:34:13,207 --> 00:34:26,917 of what the advisor does is this relationship management which 505 00:34:26,917 --> 00:34:41,998 is designed to assuage anxiety and that's really good thing because 506 00:34:41,998 --> 00:34:58,450 people need it. But another piece of the job, I'd say probably 507 00:34:58,450 --> 00:35:12,160 30%, is tax and estate planning. It's just being efficient, 508 00:35:12,160 --> 00:35:28,612 making sure you've done things the right way and have thought about 509 00:35:28,612 --> 00:35:39,580 stuff. Everybody's situation is unique. It's complicated. Right? 510 00:34:06,352 --> 00:34:19,302 And then there's 20% which is actual like investment selection 511 00:34:19,302 --> 00:34:32,252 and portfolio construction and allocation. And that last 20% is 512 00:34:32,252 --> 00:34:47,792 getting more and more mediated by software. So, it's kind of like 513 00:34:47,792 --> 00:35:03,332 SEO. It's like you want to build products that can get picked 514 00:35:03,332 --> 00:35:18,872 up by whatever those software systems are if you want to scale, 515 00:35:18,872 --> 00:35:34,412 at least in that distribution channel. And so, that's kind of how 516 00:35:34,412 --> 00:35:39,592 we think about it. 517 00:35:40,392 --> 00:35:52,628 Oh, and you did scale with that ETF, I think it's fair to say. 518 00:35:52,628 --> 00:36:01,368 More recently, I think Charlie, it was earlier this year, 519 00:36:01,368 --> 00:36:10,982 you started self-advising on another, I think the second ETF together 520 00:36:10,982 --> 00:36:19,722 with Simplify that has a specific focus on Chinese onshore 521 00:36:19,722 --> 00:36:28,462 commodities. Speak a little about that, like what makes these 522 00:36:28,462 --> 00:36:35,454 markets interesting? That would be my first question. And the 523 00:35:41,966 --> 00:35:50,623 second question I would have is, why a pure Chinese commodities 524 00:35:50,623 --> 00:36:00,067 focus? Why not, again, a more diverse mix that includes maybe other 525 00:36:00,067 --> 00:36:08,724 commodities or then currencies or then equities in addition to these, 526 00:36:08,724 --> 00:36:14,233 presumed, very diversifying Chinese onshore commodities. Why 527 00:36:14,233 --> 00:36:24,464 this pure play? Because to me it sounds like a little bit limiting, 528 00:36:24,464 --> 00:36:35,482 but I might be wrong. So, I'd like to hear your opinion on this. 529 00:36:36,792 --> 00:38:27,996 Yeah, so I think I'll start with, again, without going too deep 530 00:38:27,996 --> 00:40:09,933 on any one product, I'll speak in generalities first about Chinese 531 00:40:09,933 --> 00:40:28,467 commodity derivatives. So, the Chinese futures markets are, they're 532 00:36:57,316 --> 00:37:32,500 very large. They trade a lot of volume. The contract sizes are 533 00:37:32,500 --> 00:38:04,752 small, but a lot of volume goes through. And that's interesting. 534 00:38:04,752 --> 00:38:34,072 And China is something… Yeah, these numbers change through time. 535 00:38:34,072 --> 00:39:12,188 But let's say it's order of magnitude a third to half of the 536 00:39:12,188 --> 00:39:41,508 just overall molecular and energetic flows of the human economy 537 00:39:41,508 --> 00:40:16,692 on planet Earth. It's not small and it is interesting and it's 538 00:40:16,692 --> 00:40:28,420 also heavily policy driven. And because the systems between the 539 00:37:01,176 --> 00:37:37,752 West and the East are financially not that interlinked, 540 00:37:37,752 --> 00:38:10,264 although physically highly interlinked, there's a chance for 541 00:38:10,264 --> 00:38:46,840 some segmentation in the market to create some independent 542 00:38:46,840 --> 00:39:31,544 bet to make. And that's interesting. I think there's a strong 543 00:39:31,544 --> 00:40:08,120 argument to be made that there's real diversification benefit 544 00:40:08,120 --> 00:40:28,440 in some of this stuff. The second thing is, I think historically, 545 00:36:51,667 --> 00:37:12,917 the market structure in China was just completely different than 546 00:37:12,917 --> 00:37:38,417 the market structure in the West. And as a result different types 547 00:37:38,417 --> 00:37:57,542 of risk premia pertain, like in particular short-term momentum 548 00:37:57,542 --> 00:38:25,167 seem to work better in China than it did in Western markets. People 549 00:38:25,167 --> 00:38:48,542 love short-term momentum when it works because it's like a high 550 00:38:48,542 --> 00:39:11,917 book turnover, the number of bets is high, which directly translates, 551 00:39:11,917 --> 00:39:41,667 if you have a positive edge, to a decent ROE that's not very capital 552 00:39:41,667 --> 00:40:00,792 intensive. And of course, longer wavelength risk premia require 553 00:40:00,792 --> 00:40:22,042 longer commitment to capital, and therefore are more capital intensive. 554 00:40:22,042 --> 00:40:28,417 So that's desirable. Now, some of our research suggests that the 555 00:36:57,384 --> 00:37:20,550 Chinese commodity markets momentum factor is extending in its 556 00:37:20,550 --> 00:37:51,438 duration and that might be a function of both the markets becoming 557 00:37:51,438 --> 00:38:17,178 more sophisticated and maybe incrementally a little bit more interlinked 558 00:38:17,178 --> 00:38:45,492 with the West. Although that second thing I think they're likely 559 00:38:45,492 --> 00:39:08,658 to maintain sophistication and get more sophisticated. But their 560 00:39:08,658 --> 00:39:34,398 degree of interlinkage with the Western system I think could 561 00:39:34,398 --> 00:40:02,712 easily go backward, not just forward, as we've seen with some 562 00:40:02,712 --> 00:40:28,452 of the trade policy stuff in the last couple years. And that's 563 00:36:44,920 --> 00:37:33,688 another reason I think to possibly be in there, which is like 564 00:37:33,688 --> 00:38:14,328 if you get this rip deglobalization impulse in the system, 565 00:38:14,328 --> 00:38:54,968 you might see pretty heavy divergence between onshore and offshore 566 00:38:54,968 --> 00:39:39,672 markets and, again, get some independent bets to make which is 567 00:39:39,672 --> 00:40:28,440 never a bad thing in the context of an active trading book. 568 00:36:36,792 --> 00:37:04,584 So that's the first, I guess the first question which is they're 569 00:37:04,584 --> 00:37:25,428 interesting, they're different and they're active in their scale 570 00:37:25,428 --> 00:37:53,220 there. Now, why do it as a single isolated product? That is 571 00:37:53,220 --> 00:38:16,380 more about a client side perspective of user choice. It's 572 00:38:16,380 --> 00:38:46,488 like here it is. Right? It's not the easiest thing to have on. 573 00:38:46,488 --> 00:39:09,648 It's expensive to access it. Generally speaking there's some structure 574 00:39:09,648 --> 00:39:35,124 in the market. In order to navigate the complexity of cross 575 00:39:35,124 --> 00:39:55,968 border financial arrangements you typically have to trade these 576 00:39:55,968 --> 00:40:23,760 things via swaps with other dealers and stuff. And none of that 577 00:40:23,760 --> 00:40:28,392 is frictionless. So, we wanted to basically pull that out and basically 578 00:37:51,522 --> 00:39:21,198 be like if you want a concentrated ability to kind of dial 579 00:39:21,198 --> 00:40:28,455 your own exposure in it, then here it is. And, again, it's like 580 00:36:52,236 --> 00:37:23,124 what does software do? Software unbundles and rebundles, 581 00:37:23,124 --> 00:38:01,734 and in this instance, because it's sufficiently out there offering 582 00:38:01,734 --> 00:38:44,205 for Western allocators of any flavor to decide to allocate to, 583 00:38:44,205 --> 00:39:38,259 we wanted to pull it out as a sort of standalone, as a standalone 584 00:39:38,259 --> 00:40:24,591 thing to give people that choice of unbundling it from a product 585 00:40:24,591 --> 00:40:28,452 perspective. 586 00:40:29,032 --> 00:40:35,224 Is this one a pure time series momentum trend following model or 587 00:40:35,224 --> 00:40:41,416 does it also have other factors and flavors on top of classic 588 00:40:41,416 --> 00:40:42,448 systematic trends? 589 00:40:43,262 --> 00:40:49,094 It has some relative value stuff on top of the trend factor. 590 00:40:50,862 --> 00:40:54,613 Cool, Charlie, well, before we close it down, anything new, any 591 00:40:54,613 --> 00:40:58,023 new initiatives, new ideas, stuff that you're working on, maybe 592 00:40:58,023 --> 00:41:02,115 you cannot share it, but things that keep you excited at the 593 00:41:02,115 --> 00:41:02,456 moment? 594 00:41:02,462 --> 00:41:21,305 Yeah, I am, again, just obsessed with this concept of capital 595 00:41:21,305 --> 00:41:40,148 efficiency and leverage. I think the market just does… The average 596 00:41:40,148 --> 00:41:58,991 advisor and the average retail person in the market generally thinks 597 00:41:58,991 --> 00:42:17,834 of leverage as like the 3X Levered MicroStrategy or Tesla. It's 598 00:42:17,834 --> 00:42:34,964 like gambling product that is sentenced to death by volatility 599 00:42:34,964 --> 00:42:38,390 drag, basically. But capital efficiency is something that everybody 600 00:41:08,860 --> 00:41:18,000 likes and I'm really interested in sort of packaging that 601 00:41:18,000 --> 00:41:28,054 capital efficiency and getting more focused on CAGR for people and 602 00:41:28,054 --> 00:41:37,194 sort of like, where's that efficient frontier sort of prudent 603 00:41:37,194 --> 00:41:45,420 leverage management? I think most systematic traders, when they're 604 00:41:45,420 --> 00:41:55,474 packaging up client strategies for clients, at least sort of fungibly 605 00:41:55,474 --> 00:42:04,614 leverageable strategies like you can in futures by changing the 606 00:42:04,614 --> 00:42:14,668 margin equity dynamics, they are sort of saying like, hey, what 607 00:42:14,668 --> 00:42:25,636 margin equity should I run this thing at? It's a question of 608 00:42:25,636 --> 00:42:35,690 my own risk appetite. And that's fine. And that makes sense. 609 00:42:35,690 --> 00:42:38,432 Totally makes sense. But I think an interesting and related 610 00:41:11,296 --> 00:41:26,440 but different question is, how far could I push the envelope prudently 611 00:41:26,440 --> 00:41:40,322 in search of compounded returns? We all know that you can't 612 00:41:40,322 --> 00:41:55,466 bet a hundred percent of Kelly because Kelly's brittle. And if you 613 00:41:55,466 --> 00:42:08,086 have estimation error, you will hurt yourself. So, greater than 614 00:42:08,086 --> 00:42:24,492 Kelly for sure is bad. And then even at Kelly, full Kelly, it's 615 00:42:24,492 --> 00:42:37,112 like, you better be really confident in your estimations of 616 00:42:37,112 --> 00:42:38,374 distributions. 617 00:42:39,162 --> 00:42:46,884 Yeah. So just stepping in there… So, what you're talking about 618 00:42:46,884 --> 00:42:53,904 is fractional Kelly bet sizing, and like the Kelly criterion, 619 00:42:53,904 --> 00:42:59,520 just for listeners who haven't heard that term. Yeah, go on. I mean, 620 00:42:41,707 --> 00:42:48,833 so you probably… But the issue in our business is we do not have 621 00:42:48,833 --> 00:42:54,941 defined odds. So, even coming up with a fractional Kelly is kind 622 00:42:54,941 --> 00:42:59,522 of like yeah, well, it's not a precise exercise. 623 00:43:05,622 --> 00:43:18,305 Not at all. And so it gets pretty philosophical pretty fast. 624 00:43:18,305 --> 00:43:30,988 But if the question you're asking yourself is, for some pool 625 00:43:30,988 --> 00:43:45,977 of capital, how do I compound returns as solidly as I can while 626 00:43:45,977 --> 00:43:59,813 being indifferent to the risk? All I'm trying to do is prudently 627 00:43:59,813 --> 00:44:13,649 maximize my expected return, then how far can I push the envelope? 628 00:44:13,649 --> 00:44:28,638 And if you knew what the odds were, the optimal leverage, you have 629 00:44:28,638 --> 00:44:34,403 to be smaller than that. But picking that fraction is really a 630 00:43:10,550 --> 00:43:18,998 game of estimation error and thinking about, how do we know what 631 00:43:18,998 --> 00:43:28,854 we know and how can we be humble about it while at the same 632 00:43:28,854 --> 00:43:35,190 time doing the fundamentally unhumble thing of really leaning 633 00:43:35,190 --> 00:43:43,638 into it? And I guess the observation I would put out there, 634 00:43:43,638 --> 00:43:50,678 which I think is really interesting, is that the incremental 635 00:43:50,678 --> 00:43:58,422 returns for being able to stomach a little bit more drawdown, 636 00:43:58,422 --> 00:44:05,462 when you start talking about compounding through time, is really, 637 00:44:05,462 --> 00:44:12,502 really big. And so, the difference between somebody who can 638 00:44:12,502 --> 00:44:20,950 live with a 10 point drawdown versus somebody who can live with 639 00:44:20,950 --> 00:44:30,102 a 20 point drawdown is not like 10% more. It's way bigger than 640 00:44:30,102 --> 00:44:34,326 that, depending on your time horizon. There are all these arbitrage 641 00:43:12,902 --> 00:43:31,830 alphas in the world and people spend a lot of time thinking about 642 00:43:31,830 --> 00:43:47,846 predicting returns on the basis of those things. But the sizing 643 00:43:47,846 --> 00:44:03,862 is just such an important piece of the puzzle and thinking 644 00:44:03,862 --> 00:44:21,334 more about that, I think it's just the place that we're spending 645 00:44:21,334 --> 00:44:34,438 a lot of time thinking about and working on. 646 00:44:35,002 --> 00:44:44,561 That is an excellent wrap. Thank you. Charlie, thanks for joining 647 00:44:44,561 --> 00:44:55,858 me on the podcast today. It's been great and I hope our listeners 648 00:44:55,858 --> 00:45:02,810 will find it interesting and informative as well. A quick note 649 00:44:36,490 --> 00:44:41,450 to everyone. As usual, I'll put the most important takeaways 650 00:44:41,450 --> 00:44:47,402 of my chat with Charlie into our show notes, and please remember 651 00:44:47,402 --> 00:44:53,354 that you're always welcome to submit questions to us in case you 652 00:44:53,354 --> 00:44:57,322 have any. Our email address is info@toptradersunplugged.com. So, 653 00:44:57,322 --> 00:45:02,778 thanks for listening and until next time on Top Traders Unplugged. 654 00:45:04,982 --> 00:46:37,970 Thanks for listening to Top Traders Unplugged. If you feel you 655 00:45:10,294 --> 00:45:23,574 learned something of value from today's episode, the best way 656 00:45:23,574 --> 00:45:40,838 to stay updated is to go on over to your favorite podcast platform 657 00:45:40,838 --> 00:45:59,430 and follow the show so that you'll be sure to get all the new 658 00:45:59,430 --> 00:46:12,710 episodes as they're released. We have some amazing guests lined 659 00:46:12,710 --> 00:46:29,974 up for you and to ensure our show continues to grow, please leave 660 00:46:29,974 --> 00:46:37,942 us an honest rating and review. It only takes a minute and 661 00:45:26,438 --> 00:46:16,502 it's the best way to show us you love the podcast. We'll see you 662 00:46:16,502 --> 00:46:37,958 next time on Top Traders Unplugged. This podcast expresses 663 00:45:10,916 --> 00:45:36,630 the views of its hosts and the guests appearing on the podcast as 664 00:45:36,630 --> 00:46:02,344 of the date of its recording, and such views are subject to change 665 00:46:02,344 --> 00:46:24,102 without notice. Top Traders Unplugged do not have any duty or 666 00:46:24,102 --> 00:46:37,948 obligation to update the information contained herein. Furthermore, 667 00:45:07,250 --> 00:45:29,930 Top Traders Unplugged make no representation to its accuracy and 668 00:45:29,930 --> 00:45:54,878 it shall not be assumed that past investment performance is an 669 00:45:54,878 --> 00:46:17,558 indication of future results. Moreover, wherever there is a potential 670 00:46:17,558 --> 00:46:37,970 for profit, there is also the possibility of loss. This content 671 00:45:15,314 --> 00:46:01,808 is made available for educational purposes only and should 672 00:46:01,808 --> 00:46:37,970 not be used for any other purpose. The information contained 673 00:45:12,731 --> 00:45:41,144 in this podcast does not constitute and should not be construed 674 00:45:41,144 --> 00:46:12,140 as investment advice or an offer to sell or a solicitation to 675 00:46:12,140 --> 00:46:37,970 buy any securities or related financial instruments in any jurisdiction. 676 00:45:04,982 --> 00:45:22,573 Certain information contained herein concerning economic trends 677 00:45:22,573 --> 00:45:47,703 and performance is based on or derived from information provided 678 00:45:47,703 --> 00:46:07,807 by independent third-party sources. Top Traders Unplugged may 679 00:46:07,807 --> 00:46:32,937 believe that the sources from which such information are obtained 680 00:46:32,937 --> 00:46:37,963 are reliable. However, Top Traders Unplugged cannot guarantee 681 00:45:13,694 --> 00:45:26,762 the accuracy of such information and has not independently 682 00:45:26,762 --> 00:45:38,378 verified the accuracy or completeness of such information 683 00:45:38,378 --> 00:45:54,350 or the assumptions on which such information is based. This podcast, 684 00:45:54,350 --> 00:46:07,418 including the information contained herein, may not be reproduced, 685 00:46:07,418 --> 00:46:24,842 copied, republished or posted in whole or in part in any form, 686 00:46:24,842 --> 00:46:37,910 without the prior written consent of Top Traders Unplugged.