WEBVTT - Cliff Asness on Celebrating 25 Years With AQR 0:00:02.000 --> 0:00:05.960 This is Master's in Business with very Rid Holds on 0:00:06.200 --> 0:00:07.160 Bloomberg Radio. 0:00:09.320 --> 0:00:14.080 This week we have an additional extra special podcast live 0:00:14.160 --> 0:00:17.959 from future Proof, I sit down with AQR's Cliff Astness. 0:00:18.320 --> 0:00:22.760 He is the co founder and chief investment officer at AQR, 0:00:23.120 --> 0:00:28.160 one hundred billion dollar plus hedge fund that specializes in 0:00:28.280 --> 0:00:34.920 a run of strategies ranging from quantitative analytics to value investing, 0:00:35.000 --> 0:00:37.800 and a whole run of different things in between. I 0:00:37.920 --> 0:00:41.840 found this short discussion to be really quite intriguing. Cliff 0:00:41.920 --> 0:00:45.479 talks about all the things he got wrong, which, to 0:00:45.600 --> 0:00:48.599 be fair to him, is much less than all the 0:00:48.600 --> 0:00:52.239 things he got right. AQR is in the middle of 0:00:52.280 --> 0:00:58.760 an incredible performance run, having been positioned correctly for a 0:00:58.840 --> 0:01:02.520 variety of things that took place, not just equities coming 0:01:02.520 --> 0:01:06.840 out of the pandemic, but really approaching the rise and 0:01:06.920 --> 0:01:10.800 interest rates and the increase in inflation in just a 0:01:10.800 --> 0:01:14.560 way that took full advantage of it. He is always 0:01:14.680 --> 0:01:16.920 one of my favorite people to talk to, not just 0:01:17.000 --> 0:01:21.480 because he's so mathy and brilliant on that side, but 0:01:21.560 --> 0:01:24.839 he's one of the few quants that is so articulate 0:01:25.080 --> 0:01:28.479 and funny on the inless side, and so you get 0:01:28.520 --> 0:01:31.160 the best of both words with Cliff, with no further 0:01:31.200 --> 0:01:36.240 ado my future proof Master's in Business Live discussion with 0:01:36.400 --> 0:01:40.760 Cliff astness. You threw away a promising career in academia 0:01:41.240 --> 0:01:44.120 to go into finance. Tell us a little bit about 0:01:44.160 --> 0:01:47.080 what led to AQR and why you thought, Hey, I 0:01:47.640 --> 0:01:49.920 think I can do this better than everybody else. 0:01:50.160 --> 0:01:54.960 Well, I never had. Maybe that feeling was implied, but 0:01:54.960 --> 0:01:56.680 I wouldn't say that even I know not to say 0:01:56.680 --> 0:02:00.920 that out loud. It's funny sitting here in Newport Beach. 0:02:01.840 --> 0:02:03.040 We're in Newport Beach, right. 0:02:03.240 --> 0:02:05.920 Huntington Beach, Newport Beach. Yeah, it's right up the road, 0:02:05.960 --> 0:02:06.280 all right. 0:02:06.360 --> 0:02:08.079 My hotel I think was in Newport Beach. Can I 0:02:08.080 --> 0:02:08.880 say Newport Beach? 0:02:09.400 --> 0:02:09.800 All right? 0:02:09.919 --> 0:02:16.640 Because Newport Beach and Pimco actually features in my personal story. 0:02:17.480 --> 0:02:19.720 I was at the University of Chicago studying for a 0:02:19.760 --> 0:02:26.760 PhD in finance. I first went to Goldman Sachs for 0:02:26.800 --> 0:02:30.320 a summer. I had a professor from undergrad who worked there. 0:02:30.400 --> 0:02:34.359 One of my best friends worked there. Like you can 0:02:34.440 --> 0:02:37.720 name drop, feel free to oh John Biner. He was 0:02:37.840 --> 0:02:40.720 CEO of CIO of Fixed Income for about twenty years 0:02:41.160 --> 0:02:42.640 at g SAM. 0:02:43.720 --> 0:02:47.040 Then I went for a year. I was still writing 0:02:47.040 --> 0:02:47.600 my dissertation. 0:02:47.680 --> 0:02:50.320 I was writing my dissertation at night on value and 0:02:50.360 --> 0:02:54.399 momentum to choose individual stocks while working at Goldman during 0:02:54.400 --> 0:02:57.720 the day. It was It was the probably the second 0:02:57.760 --> 0:03:01.240 hardest time in my life. The heart artist was when 0:03:01.560 --> 0:03:03.920 my wife and I mainly her, but my wife and 0:03:03.960 --> 0:03:10.280 I had two sets of twins eighteen months apart. 0:03:10.320 --> 0:03:11.600 Four kids in eighteen months. 0:03:11.720 --> 0:03:14.360 Is she gets mad at me when I refer to 0:03:14.440 --> 0:03:17.760 our family planning as a gross failure of risk control. 0:03:19.960 --> 0:03:21.400 But it was true. 0:03:21.600 --> 0:03:22.880 So I was there about a year and I was 0:03:22.880 --> 0:03:25.280 actually just trying to decide do I want to stay 0:03:25.280 --> 0:03:27.400 at Goldman. I was having a good I was enjoying. 0:03:27.440 --> 0:03:29.440 It didn't feel like what I was supposed to do. 0:03:29.520 --> 0:03:32.799 I was a portfolio manager. I was trading. It was fun, 0:03:33.360 --> 0:03:36.280 but you know, I was studying academic finance. 0:03:36.960 --> 0:03:37.480 Dumb luck. 0:03:37.560 --> 0:03:41.080 I had written one paper in the Journal of Portfolio Management. 0:03:41.960 --> 0:03:45.800 It was called option adjusted Spreads and a Steep yield curve. 0:03:46.600 --> 0:03:48.600 You've seen the movie Chart Least then is in it. 0:03:48.680 --> 0:03:49.920 It's really very good. 0:03:52.040 --> 0:03:55.240 They read it and being fellow geeks, they liked it, 0:03:55.400 --> 0:03:56.240 asked to talk to me. 0:03:56.960 --> 0:03:59.280 Ended up fly me out. 0:03:59.400 --> 0:04:02.520 I met with with them. I remember one thing. I'm 0:04:02.520 --> 0:04:05.320 telling you too many stories. This is the first time 0:04:05.360 --> 0:04:08.680 I ever ate sushi. The guy from uh Pimpkoh I 0:04:08.680 --> 0:04:10.760 didn't eat fish at that point. I like fish now. 0:04:11.600 --> 0:04:14.720 But the guy from Pimco guy Nam, he's passed away now. 0:04:14.720 --> 0:04:16.800 But a guy named Frank Furbinovich is a great guy. 0:04:17.440 --> 0:04:20.800 Said to me, Hey, sushi okay for lunch? And this 0:04:20.960 --> 0:04:23.120 is something if there are any twenty somethings in the 0:04:23.160 --> 0:04:27.240 audience you think you're you gotta say fine to that. 0:04:27.680 --> 0:04:30.960 You don't realize the person asking you will be totally. 0:04:30.640 --> 0:04:33.120 Okay with you going. Actually, I don't love sushi. You 0:04:33.160 --> 0:04:35.560 feel like you're gonna blow the whole thing. So I'm like, yeah, no, 0:04:35.680 --> 0:04:39.880 sushi's wonderful. So on this interview, I chewed one piece 0:04:39.920 --> 0:04:42.120 of sushi for about fifteen minutes, and I did spit 0:04:42.160 --> 0:04:44.359 it into a napkin when he wasn't looking. And they 0:04:44.480 --> 0:04:46.800 still offered me a job to go out there and 0:04:46.960 --> 0:04:51.000 start a research group. It kind of light bulb went off. 0:04:51.040 --> 0:04:54.120 I'm like, wait, doing academic research, but an applied way. 0:04:55.360 --> 0:04:57.280 I get to study the same things. I get to 0:04:57.279 --> 0:04:59.360 see if they work for real, and I think Pimpko 0:04:59.480 --> 0:05:02.560 pays better academia, so good idea. So I told Goldman, 0:05:02.560 --> 0:05:05.040 I think I'm doing this, and Goldman said, you know, 0:05:05.120 --> 0:05:07.920 we're thinking of doing that. To this day, I don't 0:05:07.960 --> 0:05:12.040 know if that's true, but I decided Manhattan is much 0:05:12.080 --> 0:05:13.760 more attractive than Newport Beach. 0:05:14.200 --> 0:05:17.960 And needs this weather look but cloudy overcast. Who wants 0:05:18.000 --> 0:05:18.839 to be in California? 0:05:18.960 --> 0:05:21.479 Anyone who's happy where they ended up, who doesn't admit 0:05:21.520 --> 0:05:24.640 that there was a fair amount of luck. Now, you 0:05:24.720 --> 0:05:27.560 can make your own luck, and effort counts ability accounts, 0:05:27.640 --> 0:05:29.520 but luck counts too. 0:05:29.560 --> 0:05:31.400 I got lucky at various stages. 0:05:31.040 --> 0:05:34.240 In this I hear that on a regular basis from 0:05:34.320 --> 0:05:35.720 guests on Masters in Business. 0:05:37.120 --> 0:05:39.360 It's terrible advice to give. By the way, what first 0:05:39.400 --> 0:05:41.440 get lucky? Well, not useful? 0:05:41.920 --> 0:05:42.960 Well the table. 0:05:43.480 --> 0:05:45.240 There are lots of people who are very smart and 0:05:45.360 --> 0:05:49.800 very hard working who don't achieve outstanding, amazing success, a 0:05:49.960 --> 0:05:51.679 top point one percent success. 0:05:52.240 --> 0:05:55.240 What separates some of them? You've mentioned this repeatedly. 0:05:55.400 --> 0:05:59.920 There's a little serendipity in everything. So, speaking of serendipity, 0:06:00.160 --> 0:06:03.400 you're at Goldman, you meet some folks tell us how 0:06:03.520 --> 0:06:04.919 AQR came about. 0:06:06.120 --> 0:06:08.480 So I started this group. This was nineteen ninety four. 0:06:09.800 --> 0:06:12.000 I know, I look around, I look for the people 0:06:12.040 --> 0:06:14.320 who were born in ninety before nineteen ninety four, and 0:06:14.360 --> 0:06:19.240 I want to say hi, shout out to them. So 0:06:19.320 --> 0:06:22.120 we started this group was initially four people with a 0:06:22.720 --> 0:06:26.560 very general, scary general mandate. They had no idea why 0:06:26.600 --> 0:06:28.760 they wanted this group. They just knew other people were 0:06:28.760 --> 0:06:33.200 starting quant groups on Wall Street. Ironically, the then success 0:06:33.560 --> 0:06:35.640 of long term capital management. 0:06:37.400 --> 0:06:38.279 Made get us. 0:06:38.279 --> 0:06:39.880 Some of that made a lot of the firms think 0:06:39.960 --> 0:06:43.080 we need applied academics here. I don't think we've ever 0:06:43.120 --> 0:06:45.920 done anything particularly similar to long term capital management. 0:06:45.960 --> 0:06:49.159 But they actually helped me out too. So we started 0:06:49.200 --> 0:06:52.360 this group with no mandate, just see if you can help. 0:06:53.040 --> 0:06:55.760 We kind of went around like door to door, quants 0:06:56.320 --> 0:06:59.960 around g SAM saying does anyone need anything quantitative? 0:07:00.960 --> 0:07:04.480 And the first job they asked us to do was 0:07:04.600 --> 0:07:04.960 to help. 0:07:05.040 --> 0:07:09.520 They had an active global equity product, you know, concentrated 0:07:09.520 --> 0:07:13.880 stock picking, that had kind of a SOSO record, and 0:07:14.000 --> 0:07:16.360 if you analyze their record, it was really strong within 0:07:16.400 --> 0:07:19.160 each country, but they were always in the wrong countries. 0:07:19.880 --> 0:07:22.240 So they'd have client meetings that would be well. The 0:07:22.280 --> 0:07:25.120 good news is are Norwegian stocks crushed the Norwegian index. 0:07:25.680 --> 0:07:27.640 The bad news is there was a kup in Norway. 0:07:29.320 --> 0:07:31.600 That doesn't mean in a kup in Norway since nineteen forty, 0:07:32.360 --> 0:07:36.480 but they approximated it. So they turned us and they said, 0:07:36.720 --> 0:07:40.400 can you guys help pick countries? And here's where you know, 0:07:40.800 --> 0:07:46.280 luck meets hutzba. At this point you don't go, well, 0:07:46.320 --> 0:07:48.440 I don't know, I've never really tested picking countries. 0:07:48.720 --> 0:07:51.120 You go, hell, yes, we can help pick countries. 0:07:52.960 --> 0:07:55.400 We went into a very tiny room, four of us 0:07:55.520 --> 0:07:58.080 and kind of it's embarrassing how simple it was. We 0:07:58.120 --> 0:08:00.800 spent about three days before we hit on the obvious solution, 0:08:01.360 --> 0:08:05.040 which was to pretend countries were individual stocks and treat 0:08:05.080 --> 0:08:07.200 them like what we had studied in academia, the early 0:08:07.280 --> 0:08:11.040 work on value, momentum, even size for predicting individual stocks. 0:08:11.200 --> 0:08:12.800 We ended up writing a paper on and it held 0:08:12.880 --> 0:08:18.000 up for countries, and we just kept growing it individual stocks, bonds, currencies, 0:08:18.280 --> 0:08:21.120 applying the same philosophy over and over again. And then 0:08:21.200 --> 0:08:26.680 after about four or five years of that, we had 0:08:26.880 --> 0:08:30.200 one of my co founders, David Cabillar, started a full 0:08:30.280 --> 0:08:33.000 year of working on me saying, you know, you could 0:08:33.000 --> 0:08:35.600 do this on your own, and he was a little 0:08:35.640 --> 0:08:40.080 bit of the Mephistopheles, so I was very. 0:08:40.000 --> 0:08:43.000 Scared of doing that. It's scary to leave Goldman Sachs. 0:08:43.280 --> 0:08:45.440 Goldman Sacks is really good at terrifying you too, by 0:08:45.480 --> 0:08:49.200 the way, when you say you're thinking about leaving, if 0:08:49.200 --> 0:08:50.920 you're honest, if you don't just give him a fade 0:08:50.920 --> 0:08:53.079 a company and say I'm considering this, which I did. 0:08:53.679 --> 0:08:56.319 You sit down with very senior people and they kind 0:08:56.360 --> 0:08:57.800 of say, you know, you only get to leave Goldman 0:08:57.880 --> 0:08:58.600 Sachs once. 0:08:58.480 --> 0:09:01.960 Young man, and then you shrivel and you go, I'll stay. 0:09:03.280 --> 0:09:06.920 So I did that a couple rounds till finally end 0:09:06.960 --> 0:09:10.920 of ninety seven, early ninety eight took the plunge. 0:09:11.200 --> 0:09:14.400 And to Goldman's credit, they're very good at keeping hooks 0:09:14.480 --> 0:09:18.560 and people. They become your prime broker. They're selling you transactions, trades. 0:09:19.840 --> 0:09:22.680 This is this is getting really into some nitty gritty details. 0:09:23.120 --> 0:09:26.520 But when you leave Goldman, the part you left is 0:09:26.600 --> 0:09:29.600 quite mad at you. The rest of Goldman is quite excited. 0:09:30.280 --> 0:09:33.120 Because the money we were managing at Goldman, I did 0:09:33.200 --> 0:09:35.280 not expect to get into this kind of stuff. But 0:09:35.360 --> 0:09:37.160 the money we were managing at Goldman, we could not 0:09:37.240 --> 0:09:40.520 trade with Goldman for rather obvious reasons. 0:09:42.280 --> 0:09:43.720 Once we leave, suddenly we're. 0:09:43.600 --> 0:09:46.800 A potential and that did help smooth the relationship over 0:09:47.200 --> 0:09:51.360 and Goldman is still certainly one of the alternates. 0:09:51.440 --> 0:09:53.040 It bounces around a bit and one of the top 0:09:53.240 --> 0:09:55.080 two or three people we trade with these days. 0:09:55.559 --> 0:09:58.480 So you leave Goldman, you set up your own own shop. 0:09:59.360 --> 0:10:01.760 At what point do you get a sense, hey, this 0:10:01.920 --> 0:10:04.520 can work. We're having clients come in. The numbers look 0:10:04.559 --> 0:10:07.760 pretty good. A performance doesn't stink. When does it dawn 0:10:07.800 --> 0:10:10.920 on you? This was a good decision to hang our 0:10:10.920 --> 0:10:11.240 own shin. 0:10:11.360 --> 0:10:13.160 I don't know. I'm hoping it kicks in soon, one 0:10:13.240 --> 0:10:18.480 of these days. Soon. We had a very strong start 0:10:18.760 --> 0:10:21.600 prior to trading. We left. 0:10:21.640 --> 0:10:23.480 It took us about nine months to We didn't take 0:10:23.520 --> 0:10:25.920 a thing out of Goldman. We were building and rebuilding. 0:10:26.400 --> 0:10:29.599 We had a road show, we made it. What I 0:10:29.640 --> 0:10:32.439 think is I've been saying for years was an era, 0:10:32.559 --> 0:10:34.440 and it was an era. When we were at Goldman, 0:10:34.520 --> 0:10:37.760 we ran all kinds of different mandates. We saw quantitative 0:10:37.800 --> 0:10:40.240 tools is a very general thing, and you could run 0:10:40.280 --> 0:10:44.040 an aggressive hedge fund style long and short market neutral 0:10:44.120 --> 0:10:47.679 a lever to be to try to make a lot 0:10:47.720 --> 0:10:50.560 of money, or you could be long only and try 0:10:50.600 --> 0:10:52.440 to beat a benchmark by one and a half percent 0:10:52.480 --> 0:10:54.600 a year. You could use the tools either way. When 0:10:54.640 --> 0:10:57.400 we launched our own firm, we said we're only going 0:10:57.480 --> 0:11:00.280 to start. We intend to do it all eventually, but 0:11:00.400 --> 0:11:03.040 we're going to start with the very aggressive version of 0:11:03.080 --> 0:11:04.160 our hedge fund product. 0:11:05.160 --> 0:11:07.199 Partly, I think it was an era. Partly I think 0:11:07.440 --> 0:11:09.240 we didn't have a tremendous amount of choice. 0:11:10.000 --> 0:11:14.079 When you're thirty and starting your own firm and you 0:11:14.160 --> 0:11:17.360 say you want to run traditional long only assets, people, 0:11:18.280 --> 0:11:21.080 your potential clients kind of look and go come back 0:11:21.120 --> 0:11:22.800 when you've been doing it for five years and you 0:11:22.920 --> 0:11:25.719 look more like I look now than I looked. Then 0:11:26.559 --> 0:11:29.439 when you say we're launching an aggressive hedge fund then 0:11:29.640 --> 0:11:34.360 we're closing, they go, oh cool, we're in. So apparently 0:11:34.440 --> 0:11:37.840 if you charge two and twenty it's a lot easier 0:11:37.880 --> 0:11:39.640 to get new clients at that point in your career 0:11:39.679 --> 0:11:41.560 than if youre and we don't charge two and twenty anymore. 0:11:41.559 --> 0:11:43.640 But that's that's the old days. So we launched a 0:11:43.760 --> 0:11:47.640 very aggressive version. We took it for the fellow geeks 0:11:47.679 --> 0:11:49.880 in the audience. I assume there were a couple in 0:11:49.960 --> 0:11:54.120 the load to mid twenties. In terms of targeted volatility, 0:11:54.600 --> 0:11:58.959 that's north of a naked equity market exposure. We raised 0:11:59.000 --> 0:12:02.320 a billion dollars I do believe was the largest standing 0:12:02.400 --> 0:12:07.120 starret hedge fund launch to that point. So in that sense, 0:12:07.160 --> 0:12:09.760 we were successful immediately. It is not. 0:12:10.000 --> 0:12:12.120 It's been way bigger one since then. We've been eclipsed 0:12:12.160 --> 0:12:12.360 by that. 0:12:13.559 --> 0:12:17.679 The next eighteen months of actually running money were very, 0:12:17.800 --> 0:12:23.280 very bad. We started running money in August of ninety eight. 0:12:24.640 --> 0:12:26.559 I'm gonna mention them again again. I didn't mean to 0:12:26.640 --> 0:12:29.120 mention them once, let alone twice. But August of ninety 0:12:29.120 --> 0:12:34.000 eight is when the LTCM kind of death spiral began. 0:12:34.400 --> 0:12:40.040 Russia defaulted, and it's always Russia and LTCM began a 0:12:40.120 --> 0:12:43.520 few months our initial you know, the stock market was 0:12:43.559 --> 0:12:46.599 down about twenty percent in August of ninety eight. I 0:12:46.640 --> 0:12:49.880 always think of this as like the crash nobody remembers, 0:12:50.520 --> 0:12:52.600 and I think no one remembers it for two reasons. 0:12:53.360 --> 0:12:57.840 It wasn't one day. It was kind of fairly steady 0:12:57.880 --> 0:13:00.439 all month, and it came back pretty quickly. After it, 0:13:00.559 --> 0:13:02.880 so there's no scarring moment, and there was there's no 0:13:02.920 --> 0:13:04.079 accompanying pandemic. 0:13:05.720 --> 0:13:07.160 We did well during that crash. 0:13:08.760 --> 0:13:13.160 Then something called the dot com bubble took off, and 0:13:14.440 --> 0:13:15.640 I have written a lot about this. 0:13:17.240 --> 0:13:18.360 I think in the last five. 0:13:18.320 --> 0:13:21.120 Years people think of us maybe too much as value investors. 0:13:21.160 --> 0:13:22.480 It's a big part of what we do, but it's 0:13:22.520 --> 0:13:25.760 far from all. We go through decade long periods where 0:13:25.760 --> 0:13:27.559 we hardly talk about it because many of the other 0:13:27.640 --> 0:13:32.240 things are dominating. In giant bubbles, the value portion of 0:13:32.280 --> 0:13:36.439 what we were doing suffers. And that happened in ninety nine, 0:13:36.480 --> 0:13:40.719 two thousand, and it happened in nineteen and twenty. So 0:13:41.000 --> 0:13:44.680 we were down in this account about thirty five percent 0:13:45.640 --> 0:13:49.079 sole product. You won one product in your first eighteen months, 0:13:49.200 --> 0:13:50.440 you're down thirty five percent. 0:13:51.280 --> 0:13:51.680 Now, if the. 0:13:51.720 --> 0:13:54.400 Fellow quants in the room are saying, well, you were 0:13:54.440 --> 0:13:59.920 taking twenty three percent volatility, that's barely maybe two standard deviations. 0:14:01.040 --> 0:14:05.000 Whoever's thinking that as a moron, I was thinking it 0:14:05.280 --> 0:14:07.560 at the time because I'm thinking the world thinks about 0:14:07.600 --> 0:14:10.880 risk adjusted returns. Right we were We had up three 0:14:11.000 --> 0:14:12.120 triple digit years in this. 0:14:12.720 --> 0:14:14.719 Kind of account at Goldman not a good way to 0:14:14.760 --> 0:14:15.360 start a business. 0:14:16.520 --> 0:14:19.840 So let's talk about you mentioned you started with a 0:14:19.920 --> 0:14:23.960 long short. You've become known as value guys, but really 0:14:24.640 --> 0:14:29.080 you'll run a variety of different strategies and different different silos. 0:14:29.560 --> 0:14:31.920 Let's talk a little bit about how that developed. How 0:14:31.960 --> 0:14:34.000 did you go from hey, we're going to run and 0:14:34.080 --> 0:14:37.360 gun a long short and take an occasional thirty percent 0:14:37.480 --> 0:14:43.400 hit to adding diversified approaches to not just risk management, 0:14:43.520 --> 0:14:45.600 but where are you going to generate your alpha from? 0:14:47.360 --> 0:14:49.880 Partly, that's just been always what we've been done, even 0:14:49.960 --> 0:14:51.360 in our time at Goldman Sachs. 0:14:52.200 --> 0:14:53.880 As I just told you guys, we started out. 0:14:54.040 --> 0:14:57.960 First models we built were about what equity countries looked 0:14:58.000 --> 0:15:02.040 more attractive. We quickly poured those two bond markets. 0:15:03.320 --> 0:15:03.880 At that time. 0:15:04.000 --> 0:15:07.360 The world has gotten deeper, the factor world has gotten deeper, 0:15:07.560 --> 0:15:09.920 but value and momentum were the two main factors. 0:15:10.800 --> 0:15:13.080 When you go to a bond market, you go what's value. 0:15:14.280 --> 0:15:16.920 Everybody who does this will claim, and I will too, 0:15:17.000 --> 0:15:19.880 that their models are much better and more subtleness. But 0:15:20.000 --> 0:15:23.960 if you just looked at prospective real bond yield across 0:15:24.040 --> 0:15:27.560 twenty countries, the ones would higher yields tend to be 0:15:27.640 --> 0:15:30.760 scarier like value things often are and tend to on 0:15:30.840 --> 0:15:34.360 average do better. Momentum is ridiculously easy. Everywhere you go, 0:15:34.840 --> 0:15:37.400 things tend to keep trending in the same direction. So 0:15:37.520 --> 0:15:40.440 by the time we left Goldman, we already traded countries, 0:15:41.360 --> 0:15:46.320 stock markets, bond markets, currencies, and individual stocks, which is 0:15:46.360 --> 0:15:48.440 really what a few of us had written our dissertations 0:15:48.480 --> 0:15:53.680 on Globally. At AQR we've expanded that. Again, this is 0:15:53.720 --> 0:15:55.760 twenty five years, so it's not like we just walked 0:15:55.800 --> 0:15:59.960 in and did this into many different factors, not one hundreds. 0:16:00.520 --> 0:16:02.800 Sometimes people people refer to the factor zoo. 0:16:02.640 --> 0:16:03.280 I get annoyed. 0:16:03.800 --> 0:16:08.320 They're about five to ten major things that most and 0:16:08.400 --> 0:16:11.760 I'm speaking in general of the exceptions that most quants 0:16:11.800 --> 0:16:16.440 probably believe in. Cheap tends to be expensive, good momentum 0:16:16.520 --> 0:16:20.720 both priced and then later on importantly, fundamentals, things getting 0:16:20.760 --> 0:16:23.760 better tend to keep getting better. Lower risk things measured 0:16:23.760 --> 0:16:26.720 both fundamentally and in terms of the quant measures bita 0:16:26.760 --> 0:16:29.560 volatility tend to do better than you would think. Higher 0:16:29.680 --> 0:16:33.480 quality assets, more profitable, higher return on whatever you'd like to. 0:16:33.760 --> 0:16:34.200 Measure it on. 0:16:36.000 --> 0:16:38.960 Expanding the set of things beyond value and momentum, and 0:16:39.000 --> 0:16:42.000 then expanding the places we do it. We didn't do 0:16:42.080 --> 0:16:47.240 emerging markets at Goldman. When you add that, you get 0:16:47.280 --> 0:16:50.320 two wonderful things at once. Actually you could fail utterly. 0:16:50.400 --> 0:16:52.960 If it works, you get two wonderful things at once. 0:16:53.520 --> 0:16:56.960 You get another strategy to add, which is correlated. 0:16:56.960 --> 0:16:58.000 You shouldn't pretend it's not. 0:16:58.560 --> 0:17:01.960 But it's not perfectly correlated, so it's somewhat diversifying, and 0:17:02.160 --> 0:17:07.160 you get another little out of sample tests. Statisticians quants, 0:17:08.359 --> 0:17:12.960 we have very strange dreams. We don't dream about cars 0:17:13.080 --> 0:17:16.960 and houses and significant others. We dream about out of 0:17:17.040 --> 0:17:21.240 sample tests. It's kind of the gold standard. You often 0:17:21.280 --> 0:17:23.119 don't have enough. Sometimes you gotta wait thirty years to 0:17:23.160 --> 0:17:25.200 get a good out of sample test or something. But 0:17:25.320 --> 0:17:27.480 when you go to a new market you haven't looked 0:17:27.480 --> 0:17:32.120 at yet, and it holds up, you go, maybe we're 0:17:32.160 --> 0:17:35.280 actually onto something here. By the way, we think, if 0:17:35.320 --> 0:17:38.440 we have good out of sample tests, we will get 0:17:38.960 --> 0:17:42.040 nice cars and houses and significant others. We're not indifferent 0:17:42.080 --> 0:17:45.120 to that, we just take a different path. So we've 0:17:45.160 --> 0:17:48.119 been expanding both how we measure things and where we 0:17:48.359 --> 0:17:48.600 do it. 0:17:49.880 --> 0:17:55.280 I'm glad you reference that not geographically. But let's talk 0:17:55.320 --> 0:17:58.320 a little bit about one of the issues that seems 0:17:58.359 --> 0:18:01.640 to really have come into its own over the past decade, 0:18:01.720 --> 0:18:03.560 which is tax aware investing. 0:18:04.480 --> 0:18:06.320 It's people talk about. 0:18:06.520 --> 0:18:10.240 Asset location, and hey, I'm going to put my highest turnover, 0:18:10.480 --> 0:18:15.399 most active account into my tax deferred portfolio and the 0:18:15.520 --> 0:18:19.920 long term boring index stuff I'll keep in my taxable account. 0:18:20.560 --> 0:18:24.040 That seems to be the conventional wisdom. You seems to 0:18:24.600 --> 0:18:28.440 have moved in a direction opposite that tell us about that. 0:18:28.560 --> 0:18:32.119 Well, first, the conventional wisdom isn't wrong if you're owning 0:18:32.280 --> 0:18:36.760 a traditional, say, actively managed long only stock portfolio with 0:18:36.920 --> 0:18:41.800 some decent turnover that doesn't have very attractive tax properties. 0:18:42.760 --> 0:18:46.040 Where we got into this starting the research about ten 0:18:46.119 --> 0:18:49.679 years ago, writing our first paper, we've been fairly public 0:18:49.720 --> 0:18:52.680 about this stuff. A first paper I think on this 0:18:52.920 --> 0:18:56.960 was in twenty fifteen. Were a few different sets, and 0:18:57.080 --> 0:18:59.879 here again we got lucky. We got lucky in that 0:19:00.080 --> 0:19:02.280 we are ready were in both the traditional long only 0:19:02.359 --> 0:19:06.240 world and the long short world. And it turns out 0:19:07.160 --> 0:19:11.760 that first the very act imagine you have an active 0:19:12.200 --> 0:19:16.080 beat the benchmark traditional stock portfolio. Now imagine you separate 0:19:16.160 --> 0:19:20.119 that into an index fund and a long short portfolio. 0:19:20.160 --> 0:19:22.679 And just for the sake of this argument, imagine if 0:19:22.720 --> 0:19:24.800 you add those two together, you get back to the original. 0:19:25.520 --> 0:19:28.000 The long short is the over and the underweights that 0:19:28.119 --> 0:19:28.800 you had before. 0:19:29.560 --> 0:19:33.200 Just that active separation makes something far more tax efficient. 0:19:33.280 --> 0:19:36.520 The index fund a cruise, like all index funds are 0:19:36.560 --> 0:19:40.400 fairly tax efficient, and the active part you only pay 0:19:41.080 --> 0:19:42.200 tax if you make money. 0:19:43.000 --> 0:19:45.720 You have a bad year, you don't pay a tax 0:19:45.840 --> 0:19:46.000 on it. 0:19:46.840 --> 0:19:50.720 In a traditional long only portfolio, imagine markets they do 0:19:50.920 --> 0:19:53.240 go up on average over time, and you want to 0:19:53.280 --> 0:19:55.640 sell something you actually haven't produced alfin you just don't 0:19:55.680 --> 0:19:57.960 like the stock anymore. You get a tax hit from that. 0:19:58.680 --> 0:20:04.439 So simply the separation gets far more efficient. Then you say, well, 0:20:04.480 --> 0:20:07.760 can I do any better about this? And here it's 0:20:07.800 --> 0:20:09.640 the last time I'll try to mention luck. Not again, 0:20:10.040 --> 0:20:12.040 this will be the last time I'll mention luck. But 0:20:12.240 --> 0:20:16.080 the average turnover of our stock selection models, and this 0:20:16.240 --> 0:20:17.000 was not by design. 0:20:17.080 --> 0:20:19.920 This is why it's luck. It's about a year. You know, 0:20:20.359 --> 0:20:21.360 you know what averages means. 0:20:21.400 --> 0:20:23.040 Some things are fairly quick, some things be old for 0:20:23.160 --> 0:20:25.240 five years, but average is about a year. 0:20:26.080 --> 0:20:28.800 A year is a magic number. In stock a year 0:20:28.840 --> 0:20:29.200 and a day. 0:20:29.720 --> 0:20:33.520 Right, you know you have a big winner at eleven 0:20:33.520 --> 0:20:36.600 and a half months, you're kind of an idiot if 0:20:36.640 --> 0:20:39.760 you sell it, right and lets unless you have illegal 0:20:39.840 --> 0:20:41.120 inside information. 0:20:41.320 --> 0:20:42.639 Wait two weeks in a day. 0:20:43.600 --> 0:20:47.520 It turns out that in a long short portfolio with proxy, 0:20:47.600 --> 0:20:50.840 with not tremendous turnover but decent turnover and an average 0:20:51.240 --> 0:20:53.800 kind of one year holding period, there's a tremendous amount 0:20:53.840 --> 0:20:55.480 of optimization around that you can do. 0:20:55.920 --> 0:21:00.240 So, so the optimization on holding something beyond the to 0:21:00.359 --> 0:21:03.399 get to the lower long term capital gains tax is 0:21:03.480 --> 0:21:04.320 pretty obvious. 0:21:05.119 --> 0:21:06.840 What about the flip side of that? 0:21:07.000 --> 0:21:10.399 What about tax lost harvesting to offset those some of 0:21:10.480 --> 0:21:11.000 those gains. 0:21:11.400 --> 0:21:13.760 We're big fans of that. How do you approach this? 0:21:13.960 --> 0:21:19.520 Yeah, essentially, what we do you can think of as 0:21:20.440 --> 0:21:23.680 a I hope this sounds arrogant, a more advanced form 0:21:23.760 --> 0:21:25.280 of tax loss harvesting. 0:21:25.720 --> 0:21:28.920