WEBVTT - At the Money: Benefits of Quantitative Investing 0:00:00.120 --> 0:00:02.720 For most of the last entry, investing was a lot 0:00:02.759 --> 0:00:06.360 more art than science. People did in whatever was working 0:00:06.480 --> 0:00:11.040 based more on gut feelings than data. Portfolio management was 0:00:11.080 --> 0:00:14.480 a lot less evidence based than it is today, she. 0:00:18.160 --> 0:00:20.160 Noted with silence. 0:00:21.040 --> 0:00:23.759 As it turns out there are ways you can use 0:00:23.840 --> 0:00:27.640 data to your advantage even if you're not a math wizard. 0:00:28.080 --> 0:00:31.440 I'm Barry Ridults, and on today's edition of At the Money, 0:00:31.800 --> 0:00:34.440 we're going to discuss how to use what we've learned 0:00:34.960 --> 0:00:38.800 about quantitative investing to help us unpack all of this 0:00:38.880 --> 0:00:41.959 and what it means for your portfolio. Let's bring in 0:00:42.080 --> 0:00:46.440 Jim O'Shaughnessy. Jim is the former chairman and founder of 0:00:46.479 --> 0:00:50.360 O'Shaughnessy Asset Management, which was sold to Franklin Templeton a 0:00:50.360 --> 0:00:53.800 couple of years ago. He is also the author of 0:00:53.840 --> 0:00:57.160 the New York Times best selling book What Works on 0:00:57.240 --> 0:01:00.880 Wall Street, now in its fourth edition. What Works on 0:01:00.880 --> 0:01:05.600 Wall Street was the first quantitative equity investing work, more 0:01:05.680 --> 0:01:08.759 or less for the late person. Jim, welcome to add 0:01:08.760 --> 0:01:14.679 the money. Let's start very basically define quantitative investing. 0:01:15.200 --> 0:01:21.200 Quantitative investing, Barry is using empirical evidence that you gather 0:01:21.440 --> 0:01:25.559 over looking at how various factors like things like price 0:01:25.600 --> 0:01:30.240 to earnings ratio or earning's growth rate, and testing them 0:01:30.360 --> 0:01:33.840 over as many market cycles as you can. That gives 0:01:33.880 --> 0:01:37.759 you information that you simply couldn't have without such a test. 0:01:37.800 --> 0:01:41.120 For example, you can see what's the biggest draw down, 0:01:41.440 --> 0:01:45.240 how long did it last? How long and how often 0:01:45.319 --> 0:01:49.400 did a strategy beat its benchmark and by what magnitude. 0:01:49.560 --> 0:01:54.200 It's essentially like a very long term study just looking 0:01:54.280 --> 0:01:57.000 at the evidence as opposed to stories. 0:01:57.600 --> 0:02:01.840 So let's compare evidence versus the stories. When we look 0:02:01.880 --> 0:02:08.800 at history, quantitative models outperform professional investors and experts who 0:02:08.840 --> 0:02:14.120 rely on much squishier qualitative judgments. Why is that? 0:02:14.639 --> 0:02:18.280 Primarily the old Pogo cartoon We've met the enemy and 0:02:18.320 --> 0:02:23.519 it's us succinctly points out the reasoning here. Essentially, when 0:02:23.560 --> 0:02:27.880 we model great investors and look at the underlying factors 0:02:27.960 --> 0:02:32.480 of their portfolio, they do do extraordinarily well over time. 0:02:33.000 --> 0:02:38.280 The challenge is that the expert themselves often makes emotional choices, 0:02:38.800 --> 0:02:44.079 especially during times of intense market volatility. For example, during 0:02:44.200 --> 0:02:50.480 the Great Financial Crisis, many even quantitative investors emotionally overrode 0:02:50.480 --> 0:02:56.519 their models, so making decisions consistently according to a process 0:02:56.880 --> 0:02:59.760 that you've tested, sort of saves you from your own 0:03:00.000 --> 0:03:01.560 emotional problems. 0:03:01.919 --> 0:03:04.720 So you've looked at a lot of these strategies and 0:03:04.760 --> 0:03:09.080 strategists going back a century to the nineteen twenties. What 0:03:09.200 --> 0:03:13.640 kinds of approaches have consistently performed the best. 0:03:14.120 --> 0:03:18.920 No big surprise, Barry over long periods of time. Buying 0:03:18.960 --> 0:03:23.480 stocks more cheaply priced than those that are priced into 0:03:23.520 --> 0:03:28.040 the stratosphere generally works over long periods of time. But 0:03:28.120 --> 0:03:30.920 one of the models that we've found that actually performed 0:03:31.080 --> 0:03:34.720 really well over a variety of market cycles was essentially 0:03:34.760 --> 0:03:38.720 buying cheap stocks as measured by things like price to 0:03:38.800 --> 0:03:42.480 cash flow, even to enterprise value, etc. That are on 0:03:42.600 --> 0:03:45.760 the men that have turned a corner and are showing 0:03:45.880 --> 0:03:50.480 some good price momentum. Cheap stocks on the mend is 0:03:50.520 --> 0:03:53.400 a really interesting way to look at the market, because 0:03:53.520 --> 0:03:57.280 essentially the market is saying, yeah, that stock is very, 0:03:57.400 --> 0:04:00.880 very cheap, but we think it's probably too cheap. They're 0:04:00.880 --> 0:04:03.720 putting their money where their mouth is and buying it. 0:04:03.840 --> 0:04:06.040 That's a great strategy overall. 0:04:06.160 --> 0:04:09.400 So let's break that into two half, starting with valuation. 0:04:10.040 --> 0:04:12.280 One of the things that struck me the first time 0:04:12.280 --> 0:04:15.600 I read what works on Wall Street was the price 0:04:15.640 --> 0:04:19.960 to earnings ratio, the PE ratio which everybody seems to 0:04:20.000 --> 0:04:24.000 focus on. It doesn't really produce great results for investors. 0:04:24.440 --> 0:04:27.800 Explain why PE isn't the best way to measure valuation. 0:04:28.400 --> 0:04:32.040 Well, you know, when a measurement becomes a target, it 0:04:32.120 --> 0:04:36.000 often loses its efficacy. And you know, there's the old 0:04:36.080 --> 0:04:39.800 joke about the company hiring a new CFO and they 0:04:39.800 --> 0:04:43.160 only ask them one question, what's two plus two? And 0:04:43.240 --> 0:04:46.640 everyone answers four except for the person they hire, whose 0:04:46.680 --> 0:04:50.159 answer was what number did you have? In mind? Earnings 0:04:50.160 --> 0:04:54.800 are much easier to manipulate than things like revenue and 0:04:55.320 --> 0:04:59.360 other measurements of value, and I think that's one of 0:04:59.400 --> 0:05:03.480 the reasons why. Well, it worked very very well before 0:05:03.800 --> 0:05:08.480 all of our innovations and computer databases, etc. Once it 0:05:08.520 --> 0:05:13.000 became a target for people to pick things on, it 0:05:13.120 --> 0:05:16.039 started getting manipulated at the corporate level. 0:05:16.760 --> 0:05:20.640 So let's talk about some other measures. You talked about, 0:05:20.680 --> 0:05:26.080 price to sales ratio, you talked about EBADA to enterprise value. 0:05:26.640 --> 0:05:30.240 Tell us what actually works as a way of measuring 0:05:30.720 --> 0:05:31.799 corporate value. 0:05:32.600 --> 0:05:36.320 Specifically, we like to look at a composite of various 0:05:36.400 --> 0:05:40.640 value factors, several of which you mentioned. One of my 0:05:40.960 --> 0:05:43.600 rookie mistakes in the first version of the book was 0:05:43.640 --> 0:05:46.760 simply looking at the data and saying, well, price to 0:05:46.839 --> 0:05:50.599 sales has done the best of any single measurement. Well, 0:05:51.040 --> 0:05:55.000 it was a rookie mistake because I was measuring it 0:05:55.040 --> 0:05:58.720 over a specific period of time. As we improved our 0:05:58.800 --> 0:06:03.520 process of testing, we found that using rolling rebalances and 0:06:03.680 --> 0:06:09.760 multiple value factors, it alone was outperformed by a value composite. 0:06:10.080 --> 0:06:12.839 And let's talk a bit about price momentum. That has 0:06:12.880 --> 0:06:17.880 been a robust factor for strong performance, especially as you mentioned, 0:06:17.880 --> 0:06:22.520 when you combine momentum with value metrics give us an 0:06:22.520 --> 0:06:26.280 explanation for how we should be looking at momentum. 0:06:26.640 --> 0:06:31.000 So momentum is really interesting because academics hate it because 0:06:31.040 --> 0:06:35.440 there is no reason underlying economic reason why it should 0:06:35.480 --> 0:06:37.960 make sense. But it does when you test it all 0:06:38.000 --> 0:06:41.560 the way back to the twenties. The rolling batting averages 0:06:41.680 --> 0:06:45.000 i e. The number of periods over one, three, five, 0:06:45.040 --> 0:06:49.000 and ten years where it beats its benchmark is extremely high. 0:06:49.520 --> 0:06:52.719 And that's sort of the wisdom of crowds working there. 0:06:52.760 --> 0:06:57.840 I believe when people have very differing opinions on a stock. 0:06:58.160 --> 0:07:02.440 They have heterogeneous opinions, right, as long as those opinions 0:07:02.480 --> 0:07:07.479 remain heterogeneous, the price movement is an excellent indicator of 0:07:07.600 --> 0:07:11.880 the net net net sentiment of investors. When it's going 0:07:12.000 --> 0:07:15.680 much much higher, obviously, that's positive. When it's going negative, 0:07:16.120 --> 0:07:19.760 that's very negative. If you invert momentum and look at 0:07:19.800 --> 0:07:22.840 buying the stocks with the worst six month or twelve 0:07:22.880 --> 0:07:29.120 month price momentum, the results are a true disaster. So essentially, it's, 0:07:29.560 --> 0:07:32.920 as Ben Graham would call it, it's listening to mister market, 0:07:33.000 --> 0:07:35.520 and they're putting their money where their mouth is. And 0:07:35.520 --> 0:07:38.920 that's why I think it's such a strong and robust 0:07:39.040 --> 0:07:42.960 indicator over a huge number of market cycles. 0:07:43.400 --> 0:07:45.840 You know, it's interesting you say that. I always just 0:07:45.960 --> 0:07:49.800 assumed that if you're a big fund manager and you're 0:07:49.840 --> 0:07:53.200 buying fill in the blank, Microsoft and Vidia, Apple, it 0:07:53.240 --> 0:07:59.080 doesn't matter. You're not saying, hey, Tuesday, March nineteenth, I'm 0:07:59.080 --> 0:08:04.000 buying my five year allowance of Nvidia. You're buying that 0:08:04.240 --> 0:08:08.920 as cash flows into your funds, you're consistently buying your 0:08:08.960 --> 0:08:13.400 favorite names kind of relentlessly over time. Is that to 0:08:13.720 --> 0:08:18.040 pop psychology? Of an explanation for momentum or is there 0:08:18.080 --> 0:08:22.960 something too names that institutions like they tend to buy 0:08:23.160 --> 0:08:24.800 and continue to buy over time. 0:08:25.920 --> 0:08:29.920 Yeah, that's the persistent underlying bid theory, and I'm sure 0:08:29.960 --> 0:08:34.240 that there is an effect when institutions continue to pour 0:08:34.320 --> 0:08:37.120 money into their favorites on a buy list. But I 0:08:37.200 --> 0:08:43.080 think that the reason momentum really works is those names 0:08:43.080 --> 0:08:46.840 that you just mentioned. They do have positive momentum most 0:08:46.840 --> 0:08:50.439 of the time, but the fact is they probably aren't 0:08:50.520 --> 0:08:53.599 qualifying for the list of the stocks with the biggest 0:08:53.720 --> 0:08:57.559 change in prices. Those names tend to be very, very 0:08:57.600 --> 0:09:02.679 different than institutional favorites. So having an underlying persistent bid 0:09:02.679 --> 0:09:06.640 from institutions yeah helpful. But a lot of those names 0:09:06.760 --> 0:09:09.839 don't actually make the cut when you're sorting on your 0:09:09.840 --> 0:09:11.520 final factor being momentum. 0:09:12.040 --> 0:09:16.720 So let's talk about a fascinating piece of research you did. 0:09:16.720 --> 0:09:20.840 I believe is also referenced in the book. People like 0:09:20.920 --> 0:09:24.000 things like private equity and venture capital, but they're not 0:09:24.520 --> 0:09:28.160 thrilled with being locked up for five years or seven years, 0:09:28.200 --> 0:09:34.199 or sometimes even ten years. You identified that the microcaps 0:09:34.440 --> 0:09:39.760 screened for quality seem to reproduce venture capital and private 0:09:39.800 --> 0:09:42.800 equity returns but without the lock up period. Tell us 0:09:42.800 --> 0:09:43.280 about that. 0:09:44.880 --> 0:09:48.920 Yeah, we have several papers at Shawn c Asset Management 0:09:49.000 --> 0:09:54.240 on that effect. It's really fascinating because the microcap universe 0:09:54.400 --> 0:09:57.680 is kind of this undiscovered country. Half of the names 0:09:57.800 --> 0:10:01.280 in it aren't even covered by a single analysts and 0:10:01.360 --> 0:10:05.760 when you use quality, momentum, etc. To sort it out, 0:10:05.840 --> 0:10:12.120 because warning, the universe itself is pretty not a great, 0:10:12.720 --> 0:10:13.560 not a great universe. 0:10:13.679 --> 0:10:15.960 You can call it garbage, Jim, It's okay. 0:10:15.720 --> 0:10:19.679 Yeah, okay, all right, So the universe itself is garbage, 0:10:20.040 --> 0:10:23.800 but there are a lot of hidden gems there and 0:10:23.960 --> 0:10:27.280 the ability to sort out those hidden gems that are 0:10:27.440 --> 0:10:31.560 little covered or not covered at all. Basically, what we 0:10:31.640 --> 0:10:35.000 found in a paper that we published several years ago 0:10:35.360 --> 0:10:39.160 was the returns sort of are great proxy for private 0:10:39.200 --> 0:10:43.360 equity in particular. And so if you're looking for a 0:10:43.640 --> 0:10:47.960 far less expensive way to get private equity, like returns 0:10:48.320 --> 0:10:51.280 at lower fees with no lock up, you'll want to 0:10:51.320 --> 0:10:55.800 take a look at the microcap universe sort it by 0:10:56.080 --> 0:10:57.319 these various metrics. 0:10:57.960 --> 0:11:00.439 So in the book What Works on Wall Street you 0:11:00.640 --> 0:11:08.160 emphasize the importance of having a systematic, disciplined approach. Explain 0:11:08.280 --> 0:11:13.959 to listeners what goes into taking what is kind of 0:11:15.200 --> 0:11:19.040 used to be sort of a loose, an undisciplined approach 0:11:19.120 --> 0:11:23.319 to stock selection and turning it into something much more disciplined. 0:11:25.040 --> 0:11:29.440 Well, I think that essentially, I'd say, would you go 0:11:29.520 --> 0:11:32.280 to a doctor who looked at you and said, hey, Barry, 0:11:32.640 --> 0:11:35.040 I just got these little yellow pills and they look 0:11:35.080 --> 0:11:37.720 appealing to me, and I think they might work for 0:11:37.800 --> 0:11:40.600 what's wrong with you? I don't think you would, right, 0:11:40.679 --> 0:11:44.080 I think you'd say, well, where are the studies, where's 0:11:44.120 --> 0:11:49.120 the evidence? Where is the long longitudinal studies to prove 0:11:49.160 --> 0:11:52.600 the efficacy of this little yellow pill? Right? That's really 0:11:52.679 --> 0:11:56.520 what we're doing with factor or quantitative investing. We are 0:11:56.679 --> 0:12:01.679 looking historically at ideas that make economic sense, right, don't 0:12:01.679 --> 0:12:05.720 pay the moon by momentum, et cetera. But then this 0:12:05.880 --> 0:12:09.400 is the key important part. We're turning it into a 0:12:09.520 --> 0:12:13.880 process that we run time and again and don't override. 0:12:14.400 --> 0:12:19.920 You know, the in basketball to investing, the process is 0:12:20.080 --> 0:12:24.160 much more important than the either intuitive ooh I should 0:12:24.200 --> 0:12:27.319 jump on this name or the terror Oh my god, 0:12:27.720 --> 0:12:28.720 the name is collapsing. 0:12:28.760 --> 0:12:29.200 I've got to. 0:12:29.240 --> 0:12:32.760 Jump out of it. It really brings a rigor and 0:12:33.080 --> 0:12:37.559 a discipline to approaching the market that is really hard 0:12:37.600 --> 0:12:43.559 to duplicate without that process underlying the quantitative methodology not impossible, 0:12:44.080 --> 0:12:50.120 but willpower dissipates very very quickly, especially in times of 0:12:50.240 --> 0:12:55.160 either exuberance right during a bubble or despair during a 0:12:55.160 --> 0:12:59.640 bear market. Following the process through thick and thin, which 0:12:59.679 --> 0:13:02.680 you're all always trying to improve, by the way, but 0:13:02.960 --> 0:13:08.319 following that process without making any additional emotional overrides, has 0:13:08.320 --> 0:13:12.640 proven itself to be quite effective at getting rid of, 0:13:13.280 --> 0:13:17.280 or at least neutralizing some of the very famous behavioral 0:13:17.320 --> 0:13:20.360 biases that we all have as humans. Right, we're all 0:13:20.679 --> 0:13:26.400 running human operating system and helping us avoid the pitfalls 0:13:26.559 --> 0:13:30.160 is really what the underlying process does, and does very 0:13:30.280 --> 0:13:30.840 very well. 0:13:30.920 --> 0:13:35.079 So let's address that for a final question. One of 0:13:35.120 --> 0:13:39.040 the things you have discussed previously is some of the 0:13:39.040 --> 0:13:45.440 biggest challenges investors face is avoiding emotional decision making. What 0:13:45.640 --> 0:13:50.520 are the tools you recommend for making sure that the 0:13:50.559 --> 0:13:55.960 average mom and pop investor doesn't succumb to their own 0:13:55.960 --> 0:14:02.480 emotional limbic system and making choice from the wrong place, 0:14:02.559 --> 0:14:06.280 making choices from emotional panic or greed. 0:14:07.320 --> 0:14:09.800 Well, you know, I've often said that the four horsemen 0:14:09.840 --> 0:14:13.400 of the investment of apocalypse are fear, greed, hope, and ignorance, 0:14:13.720 --> 0:14:16.680 and ignorance is the only one that is really correctable 0:14:16.800 --> 0:14:22.800 by studying. It's very, very difficult, especially as you note, 0:14:22.920 --> 0:14:26.880 for retail investors who look they have other interests, they 0:14:26.920 --> 0:14:29.960 have other things that they're going to spend their time on. 0:14:30.440 --> 0:14:34.560 So what I concluded was probably the best thing that 0:14:34.600 --> 0:14:38.640 you can do is find yourself a good financial advisor 0:14:39.040 --> 0:14:42.720 who could sort of serve as your wingman. The thing 0:14:42.760 --> 0:14:45.320 that advisors are able to do because of a lot 0:14:45.320 --> 0:14:47.400 of reasons, right it's not their money. They can be 0:14:47.520 --> 0:14:50.360 much more dispassionate about it, they can be much more 0:14:50.400 --> 0:14:53.880 professional about it, and then they can help their client 0:14:54.840 --> 0:14:57.520 during those tough times. It's like the old joke about 0:14:57.560 --> 0:15:02.120 anesthesiologists ninety five percent of the time they're board silly. 0:15:02.440 --> 0:15:05.240 Five percent of the time that is where they earn 0:15:05.360 --> 0:15:06.720 all their money. 0:15:07.400 --> 0:15:11.280 Really interesting. Thanks Jim for all these insights. So to 0:15:11.360 --> 0:15:17.680 wrap up, quantitative investing provides an enormous advantage to investors. 0:15:18.480 --> 0:15:23.080 It's specific, it's evidence based, it uses data, and it 0:15:23.160 --> 0:15:28.000 avoids the emotional decision making that leads investors to stray. 0:15:28.800 --> 0:15:33.320 If you want to apply some quantitative strategies to your portfolio, 0:15:33.880 --> 0:15:38.280 consider looking at the combination of momentum and low price 0:15:38.360 --> 0:15:44.400 stocks or microcaps that have been screened for quality and value. 0:15:44.760 --> 0:16:00.960 I'm Barry Ridoults. You're listening to Bloomberg's At the Money.