1 00:00:01,520 --> 00:00:05,062 I'm not so worried about missing out on perhaps future gains 2 00:00:05,062 --> 00:00:09,570 by not selling at the absolute top because I know that it had a 3 00:00:09,570 --> 00:00:14,400 really good ride. What I want to avoid is if we are in a bubble, 4 00:00:14,400 --> 00:00:18,264 we are in a big market delusion, and the whole market then 5 00:00:18,264 --> 00:00:22,128 turns south as it did for electric vehicles in sort of 2022, 6 00:00:22,128 --> 00:00:22,450 2023. 7 00:00:24,880 --> 00:00:29,360 Welcome to top Traders Unplugged. In markets, success doesn't 8 00:00:29,360 --> 00:00:33,100 come from predicting what happen, comes from being prepared 9 00:00:33,100 --> 00:00:37,500 for what you can't predict. In each episode, we go deep with some 10 00:00:37,500 --> 00:00:40,980 of the world's most thoughtful minds in investing, economics and 11 00:00:40,980 --> 00:00:44,780 beyond to understand how they think, how they prepare and how they 12 00:00:44,780 --> 00:00:48,540 decide, and the experiences that shaped how they see the world. 13 00:00:49,020 --> 00:00:53,460 No noise, no shortcuts, just real conversations to help you think 14 00:00:53,460 --> 00:00:55,500 better and invest with confidence. 15 00:00:58,250 --> 00:01:06,840 Welcome, everyone, to Top Traders Unplugged. My name is Kevin 16 00:01:06,840 --> 00:01:18,007 Coldiron and I'm host of the Ideas Lab series where we talk with 17 00:01:18,007 --> 00:01:28,315 authors of new books that help us understand and navigate the global 18 00:01:28,315 --> 00:01:37,764 economy. Our guest today is Alex Edmans. He's a professor of 19 00:01:37,764 --> 00:01:47,213 finance at London Business School and a leading expert on market 20 00:01:47,213 --> 00:01:48,072 psychology. You may have read about his research in the FT or the 21 00:01:10,214 --> 00:01:23,175 Wall Street Journal, or you may have read one of his books. He's 22 00:01:23,175 --> 00:01:34,142 the author of May Contain Lies, which was an Amazon number 23 00:01:34,142 --> 00:01:46,106 one bestseller, and also Grow the Pie, a Financial Times book of 24 00:01:46,106 --> 00:01:48,100 the year. Alex is here today to talk about his new book, just 25 00:01:11,307 --> 00:01:24,364 out in September, called The Madness of Markets: Why Smart Investors 26 00:01:24,364 --> 00:01:37,421 Make Crazy Decisions and How to Exploit Them. Alex, thanks so 27 00:01:37,421 --> 00:01:48,104 much for joining us and welcome to the show. 28 00:01:48,520 --> 00:01:50,840 Thanks so much for having me on, Kevin. It's great to be here. 29 00:01:51,940 --> 00:02:00,487 First of all, shout out to London Business School. That's where 30 00:02:00,487 --> 00:02:09,811 I studied finance and got started using a lot of the ideas 31 00:02:09,811 --> 00:02:19,135 you actually talk about in your book. So, yeah, I wanted to 32 00:02:19,135 --> 00:02:29,236 maybe start off with what you say. You say that mood matters, that 33 00:02:29,236 --> 00:02:37,006 sentiment impacts markets in surprising ways. And perhaps we can 34 00:02:37,006 --> 00:02:45,553 start off with a couple of stories from your own research. I 35 00:01:52,695 --> 00:02:01,755 mean, you became well known early in your career for a paper 36 00:02:01,755 --> 00:02:10,060 that looked at the impact of sporting results on markets. And 37 00:02:10,060 --> 00:02:19,875 then later you looked at the impact of music on markets, and you 38 00:02:19,875 --> 00:02:28,935 used that as kind of an illustration to show us that markets 39 00:02:28,935 --> 00:02:36,485 are not just driven by raw fundamentals, sentiment matters a 40 00:02:36,485 --> 00:02:45,545 lot. So, I was wondering, perhaps you could tell us those stories. 41 00:02:46,400 --> 00:03:04,976 Certainly. So, while you learned finance at London Business 42 00:03:04,976 --> 00:03:38,000 School, I learned it at MIT in the US, I did my PhD there. And MIT 43 00:03:38,000 --> 00:04:00,704 being MIT, all the professors and my classmates, they were looking 44 00:04:00,704 --> 00:04:21,344 at rational quantitative factors. So, the price of a stock 45 00:04:21,344 --> 00:04:41,984 depends, on future dividends, on growth rates and inflation, and 46 00:04:41,984 --> 00:04:50,240 all of that stuff. But I had just been in the real world. So, 47 00:03:09,770 --> 00:03:35,477 I was at Morgan Stanley, between undergrad and my PhD, and 48 00:03:35,477 --> 00:04:03,521 while people are really smart and they're able to devise big Excel 49 00:04:03,521 --> 00:04:22,217 spreadsheets, taking those into account, they're also human. 50 00:04:22,217 --> 00:04:45,587 I wanted to show that psychology, and emotions, and sentiment 51 00:04:45,587 --> 00:04:50,261 also matters. Now, showing that sentiment affects the stock 52 00:03:02,759 --> 00:03:23,792 price is quite difficult because many factors that affect 53 00:03:23,792 --> 00:03:47,162 the mood will also affect fundamentals. For example, if there's 54 00:03:47,162 --> 00:04:15,206 a plane crash, yes, people are unhappy, but the stock market might 55 00:04:15,206 --> 00:04:40,913 go down not because people are unhappy, but instead because it's 56 00:04:40,913 --> 00:04:50,261 bad for the economy. So, this is why I looked at crazy things such 57 00:03:12,750 --> 00:03:39,100 as soccer results. Why? Because if England loses, that should 58 00:03:39,100 --> 00:04:08,085 have no effect on profits, or dividends, or inflation, except in 59 00:04:08,085 --> 00:04:37,070 a few sectors like replica merchandise. But it does cause a 60 00:04:37,070 --> 00:04:50,245 huge effect on national mood. And what I found, long story short, 61 00:02:58,965 --> 00:03:22,300 is that when a country gets knocked out of the World Cup, the 62 00:03:22,300 --> 00:03:42,045 market falls significantly the next day. To put some numbers on 63 00:03:42,045 --> 00:04:05,380 it, it's about half a percent, even when you control for the broader 64 00:04:05,380 --> 00:04:25,125 world market movement. And when applied to the UK stock market, 65 00:04:25,125 --> 00:04:48,460 that is over 10 billion pounds wiped off the market on a single 66 00:04:48,460 --> 00:04:50,255 day. And so why is that so profound? Because you might think, 67 00:03:01,712 --> 00:03:17,024 yes, emotions matter. Yes, certain things make me happy or unhappy. 68 00:03:17,024 --> 00:03:33,728 But when I am trading, I become rational. I shake that off. 69 00:03:33,728 --> 00:03:51,824 I'm a professional who puts my emotions to one side. But what my 70 00:03:51,824 --> 00:04:11,312 data shows is that's not the case. You are human and you are affected 71 00:04:11,312 --> 00:04:28,016 by emotions even when you're in your job. And therefore, we need 72 00:04:28,016 --> 00:04:44,720 to be very careful when making investing decisions, not to be swayed 73 00:04:44,720 --> 00:04:50,288 by these extraneous factors. 74 00:04:51,830 --> 00:04:54,670 What happened to the London stock market after they lost to Argentina 75 00:04:54,670 --> 00:04:55,350 this summer? 76 00:04:55,910 --> 00:05:02,200 So that actually opened about 0.5% down. And interestingly, Yahoo 77 00:05:02,200 --> 00:05:09,119 Finance attributed this in part to the England defeat. So, because 78 00:05:09,119 --> 00:05:17,296 my study was published 20 years ago now, it is, as that was 79 00:05:17,296 --> 00:05:24,215 somewhat well known among markets people. And so, it was gratifying. 80 00:05:24,215 --> 00:05:30,505 They said, oh, it's not necessarily just due to fundamentals, 81 00:05:30,505 --> 00:05:34,908 but partly due to this emotional effect. And why is that 82 00:04:59,538 --> 00:05:06,794 interesting? Because the morning after the Argentina defeat, 83 00:05:06,794 --> 00:05:18,585 there was good economic news. I think the UK had the highest growth 84 00:05:18,585 --> 00:05:30,376 rate for the last 13 months, yet it opened lower, in part because 85 00:05:30,376 --> 00:05:34,911 of this negative sentiment indicator. 86 00:05:35,330 --> 00:05:38,210 I wonder if people started building negative futures positions 87 00:05:38,210 --> 00:05:41,730 when they went into that defensive crouch in the second half 88 00:05:41,730 --> 00:05:42,690 of the game. 89 00:05:42,690 --> 00:05:54,162 Everybody I was watching with, and all my friends said, as soon 90 00:05:54,162 --> 00:06:03,722 as they went into that defensive crouch, then people thought, 91 00:06:03,722 --> 00:06:13,282 oh, England were going to lose. More seriously, when Argentina… 92 00:06:13,282 --> 00:06:23,798 sorry… when Germany and Brazil, when there was that 7/1 defeat 93 00:06:23,798 --> 00:06:36,226 a couple of World Cups ago, you could watch in real time, when 94 00:06:36,226 --> 00:06:45,786 the goals went in, actually, the stock prices immediately moved. 95 00:05:42,690 --> 00:05:48,619 Although surprisingly, in that case, that was the one case that 96 00:05:48,619 --> 00:05:54,548 moved against me. Why? Because when Brazil kept conceding more and 97 00:05:54,548 --> 00:06:01,016 more goals, the stock market went up. And you might think, doesn't 98 00:06:01,016 --> 00:06:07,484 that prove me wrong? Well, no, because actually in Brazil, the stock 99 00:06:07,484 --> 00:06:12,874 market was concerned about Dilma Rousseff, who was the Brazilian 100 00:06:12,874 --> 00:06:19,342 president, who is very left wing. And because she had spent loads 101 00:06:19,342 --> 00:06:26,349 of money on stadiums and she had built her identity on the Brazilian 102 00:06:26,349 --> 00:06:32,278 football team's success, there was thought that she would not win 103 00:06:32,278 --> 00:06:38,207 another election if Brazil got knocked out. So, mood mattered, football 104 00:06:38,207 --> 00:06:44,675 mattered in that case, but for other reasons from what it would 105 00:06:44,675 --> 00:06:45,753 do normally. 106 00:06:47,370 --> 00:06:49,610 Did she lose the election? She did, I think. 107 00:06:49,850 --> 00:06:53,230 I think she did, yeah. So, this was the 2014, there was the 108 00:06:53,230 --> 00:06:55,050 very famous 7/1 home defeat to Germany. 109 00:06:56,150 --> 00:07:02,414 Okay, so you did that work with sports results or sports results 110 00:07:02,414 --> 00:07:08,156 impacting markets, why then did you continue to go down that 111 00:07:08,156 --> 00:07:14,420 path? Because later on, I think you did some work on Spotify 112 00:07:14,420 --> 00:07:15,986 playlists and markets. 113 00:07:16,790 --> 00:07:36,018 Yes. And so that's because there are many, many things that 114 00:07:36,018 --> 00:07:55,246 could be affecting your emotions. So, yes, I might be upset 115 00:07:55,246 --> 00:08:17,970 because my sports team lost, but I might be upset because I had 116 00:08:17,970 --> 00:08:40,694 an argument with my wife, or because my kids are sick, or because 117 00:08:40,694 --> 00:09:06,914 a bird unloaded on my shirt. But I'm not going to be able to get 118 00:09:06,914 --> 00:09:17,402 data on that as a researcher. So instead, I thought of something 119 00:07:31,262 --> 00:07:55,382 different. Rather than looking at the individual drivers of mood 120 00:07:55,382 --> 00:08:14,678 (sports results, arguments with spouses, kids, being sick), 121 00:08:14,678 --> 00:08:48,446 let me look at the effects of mood. And there is a literature on 122 00:08:48,446 --> 00:09:10,154 what's known as behavioral congruence. We take actions which 123 00:09:10,154 --> 00:09:17,390 reflect our emotions. So, if I'm really happy, I might play Walking 124 00:07:29,867 --> 00:07:47,303 on Sunshine. If I'm really unhappy and I've just broken up with 125 00:07:47,303 --> 00:08:04,739 my partner, I might play Nothing Compares to You. And so why 126 00:08:04,739 --> 00:08:20,722 is music listening something that I wanted to look at? Because 127 00:08:20,722 --> 00:08:36,705 I can get aggregated data across the whole economy. Why? Because 128 00:08:36,705 --> 00:08:54,141 Spotify has a list of the top hundred streamed songs, whereas you 129 00:08:54,141 --> 00:09:11,577 don't have individual data on things such as kids being sick or 130 00:09:11,577 --> 00:09:17,389 arguments with your spouse. And what is interesting is that you 131 00:07:25,148 --> 00:07:40,670 can use an algorithm to look at the happiness of a song. Notice 132 00:07:40,670 --> 00:07:53,804 this is happiness of the music, not the lyrics. And that's 133 00:07:53,804 --> 00:08:06,938 really important because there are some songs with some quite dark 134 00:08:06,938 --> 00:08:21,266 lyrics like Semi Charmed Life or I Don't Like Mondays, which are 135 00:08:21,266 --> 00:08:34,400 still quite positive and upbeat songs. And so, what we did 136 00:08:34,400 --> 00:08:52,310 is we used that algorithm to look at how happy or how sad a country's 137 00:08:52,310 --> 00:09:06,638 playlist was as a reflector of national mood. And what we showed 138 00:09:06,638 --> 00:09:16,190 was that was also significantly correlated with stock 139 00:09:16,190 --> 00:09:17,384 prices. Interestingly, it's also negatively correlated with government 140 00:07:35,144 --> 00:08:01,364 bond prices, which makes sense because when you're more optimistic, 141 00:08:01,364 --> 00:08:32,828 then you're going to put more money into equities, that money needs 142 00:08:32,828 --> 00:09:04,292 to come from somewhere and it will come out of something safer, 143 00:09:04,292 --> 00:09:17,402 such as UK government bonds. 144 00:09:18,790 --> 00:09:22,710 That measure that you developed for kind of aggregate national 145 00:09:22,710 --> 00:09:27,022 mood, is that something that people now publish and track? Is 146 00:09:27,022 --> 00:09:32,510 that easy for me to look at? Is the national mood worse now? And 147 00:09:32,510 --> 00:09:37,214 is that part of the reason that the bond market is struggling? 148 00:09:37,730 --> 00:09:41,210 Yes. So, it's actually a measure that we didn't come up with 149 00:09:41,210 --> 00:09:44,690 ourselves. So, we stood on the shoulder of sort of giants there. 150 00:09:44,690 --> 00:09:48,460 So, the MIT Media Lab came up with something. There was a small 151 00:09:48,460 --> 00:09:52,230 inner lab there called the Echo Nest. And what they did is they 152 00:09:52,230 --> 00:09:55,710 produced a measure of valence, which is the happiness of the song, 153 00:09:55,710 --> 00:09:59,190 and we downloaded it from their website, and we used that as 154 00:09:59,190 --> 00:10:02,670 the measure. So, it is a measure from Echo Nest, not from 155 00:10:02,670 --> 00:10:02,960 us. 156 00:10:03,550 --> 00:10:05,710 So, that measure is out there. People could... 157 00:10:06,110 --> 00:10:09,724 Yes, you can look at it, although what you can't do… So, at 158 00:10:09,724 --> 00:10:13,060 the time, we were able to scrape that data legally and download 159 00:10:13,060 --> 00:10:16,952 the data for a very large time period. Now, I think you can look 160 00:10:16,952 --> 00:10:21,122 at it, I don't think you can scrape it in the same way, but that 161 00:10:21,122 --> 00:10:23,068 data is still at least publicly available. 162 00:10:23,550 --> 00:10:34,890 Gotcha. Okay. So, as I'm reading the book, there's a lot of 163 00:10:34,890 --> 00:10:45,285 very interesting stories about the ways in which markets are, as 164 00:10:45,285 --> 00:10:56,625 you say, impacting my mood and sentiment, etc. And you say that 165 00:10:56,625 --> 00:11:07,020 the goal of your, I guess, trying to educate people shouldn't 166 00:11:07,020 --> 00:11:19,305 be to outwit the market with facts, but to outlast it with judgment. 167 00:11:19,305 --> 00:11:27,810 And. And my question was, do you think that…? I mean, typically 168 00:10:25,959 --> 00:10:33,186 we imagine that professionals have an advantage over individuals, 169 00:10:33,186 --> 00:10:42,822 but I wondered if in this respect, perhaps the reverse is the 170 00:10:42,822 --> 00:10:52,458 case, because individuals have a lot more freedom in the way they 171 00:10:52,458 --> 00:11:00,488 can invest. They're not tied to benchmarks. They're not being 172 00:11:00,488 --> 00:11:09,321 measured every quarter. They aren't going to get fired if they 173 00:11:09,321 --> 00:11:18,154 go to cash. Do individuals, in your mind, have certain advantages 174 00:11:18,154 --> 00:11:25,381 over professionals in exploiting some of these mood and 175 00:11:25,381 --> 00:11:27,790 sentiment related ideas? 176 00:11:28,420 --> 00:12:20,274 They may do. And this is quite paradoxical that actually, maybe, 177 00:12:20,274 --> 00:13:07,414 amateurs may have greater potential to beat the market than 178 00:13:07,414 --> 00:13:54,554 professionals. So, that is a really bold and controversial statement. 179 00:13:54,554 --> 00:14:27,552 So how can I back this up? So, at the end of the book I have a final 180 00:12:02,993 --> 00:12:37,566 chapter where I give practical takeaways to improve your own investing 181 00:12:37,566 --> 00:13:15,282 decisions. And out of the twenty, number two is know your edge. 182 00:13:15,282 --> 00:13:52,998 So, what do I mean by this? Before you think about investing 183 00:13:52,998 --> 00:14:24,428 in the first place, think about what your competitive edge 184 00:14:24,428 --> 00:14:27,571 is. So many people think their edge is knowledge. They might say, 185 00:11:52,741 --> 00:12:23,695 oh, I have knowledge about AI. I've used AI myself for my own planning. 186 00:12:23,695 --> 00:12:54,649 I think AI is going to be the future. But you really don't have 187 00:12:54,649 --> 00:13:18,970 edge in knowing about AI because you're going to be trading 188 00:13:18,970 --> 00:13:43,291 against lots of professionals who study the AI industry every single 189 00:13:43,291 --> 00:14:12,034 day and they have access to management and they are able to understand 190 00:14:12,034 --> 00:14:27,511 the power of these algorithms much better. Similarly, I might think, 191 00:11:36,292 --> 00:12:01,876 oh, I'm a doctor, I understand the power of this beta blocker. But 192 00:12:01,876 --> 00:12:19,588 again, you're going to be trading against pharmaceuticals analysts 193 00:12:19,588 --> 00:12:41,236 who study this every single day. So, as an amateur investor, 194 00:12:41,236 --> 00:13:00,916 maybe your edge is not knowledge, it's not unique information, 195 00:13:00,916 --> 00:13:18,628 because professionals have that. But instead your edge comes 196 00:13:18,628 --> 00:13:38,308 from a different source, which is exactly what you've identified, 197 00:13:38,308 --> 00:14:05,860 Kevin, which may well be the ability to be free, to be able to 198 00:14:05,860 --> 00:14:27,508 play a long game. So, I call this the endurance game. What do 199 00:11:33,328 --> 00:12:07,684 I mean by this? In the depths of the financial crisis, at the start 200 00:12:07,684 --> 00:12:39,586 of 2009, I put a lot of my money into equities. Why? Because 201 00:12:39,586 --> 00:13:04,126 I thought the market was really undervalued because of the 202 00:13:04,126 --> 00:13:26,212 collapse in the financial crisis. Now, would a professional 203 00:13:26,212 --> 00:13:58,114 investor have done the same as me? No, because it was really difficult 204 00:13:58,114 --> 00:14:27,562 to call the bottom. Maybe the market would have crashed even more. 205 00:11:28,420 --> 00:12:00,564 And if I put my money in the equities in January 2009, and it 206 00:12:00,564 --> 00:12:28,116 goes down by April 2009, then I'm bleeding losses. I might plead 207 00:12:28,116 --> 00:12:55,668 with my investors, trust me, it's all going to come better, just 208 00:12:55,668 --> 00:13:20,924 be patient, in five years' time, things will improve. But they 209 00:13:20,924 --> 00:13:48,476 won't wait five years. They might pull their money and say, well, 210 00:13:48,476 --> 00:14:18,324 I'm going to put it in with another fund manager who is going 211 00:14:18,324 --> 00:14:27,508 to be performing better. But for me, investing my own money, I 212 00:11:45,684 --> 00:12:09,422 don't have external clients who I need to be concerned about. 213 00:12:09,422 --> 00:12:33,160 I'm just investing in my future or my family's future. Back 214 00:12:33,160 --> 00:13:03,372 then I didn't have a family, but I thought I might at some point, 215 00:13:03,372 --> 00:13:37,900 and so, my time horizon was 30 or 40 years. And so as long as the 216 00:13:37,900 --> 00:14:05,954 market was cheap and going to recover over the next five or 10 217 00:14:05,954 --> 00:14:27,534 years, it didn't matter that I didn't call the bottom. And this 218 00:11:35,444 --> 00:12:03,540 applies not just to individuals but some institutional 219 00:12:03,540 --> 00:12:35,148 investors. So, Clare College, Cambridge University, they also put 220 00:12:35,148 --> 00:13:17,292 a lot of money into equities in 2009. Why? Because they don't 221 00:13:17,292 --> 00:13:48,900 have external clients. Instead, their mandate was to fund 222 00:13:48,900 --> 00:14:27,532 future research and teaching for the next 30 to 40 years. 223 00:14:29,611 --> 00:14:33,835 Yeah, I actually remember that. I think they issued… Did they 224 00:14:33,835 --> 00:14:37,291 issue…. they issued debt to do that, didn't they? 225 00:14:37,931 --> 00:14:42,353 They exactly did. So, they issued, I think, about £10 million 226 00:14:42,353 --> 00:14:47,177 of debt. And so, that £10 million, over the next 10 years, 227 00:14:47,177 --> 00:14:52,403 I think went up to £15 million because of interest, the equity that 228 00:14:52,403 --> 00:14:57,227 they invested in went to £30 million. So, it tripled while the 229 00:14:57,227 --> 00:14:59,639 interest only increased it by 50%. 230 00:15:00,691 --> 00:15:12,054 So, you brought up something when you were explaining the answer 231 00:15:12,054 --> 00:15:20,318 to the last question about professionals having particular, 232 00:15:20,318 --> 00:15:29,615 I guess, domain-based expertise and knowledge. So, I don't 233 00:15:29,615 --> 00:15:39,945 know, a biotech investor, they might have these expert networks, 234 00:15:39,945 --> 00:15:52,341 or they might indeed have people on staff who are medical doctors, 235 00:15:52,341 --> 00:16:03,704 researchers, that type of thing. And it made me think about 236 00:16:03,704 --> 00:16:15,067 when you are an individual investor and you're looking to invest 237 00:16:15,067 --> 00:16:26,430 your own money, should you be looking for specialists versus generalists? 238 00:15:00,691 --> 00:15:10,316 And so, I can see the advantage that specialists have in 239 00:15:10,316 --> 00:15:21,691 that no one knows more about a particular sector than they do. On 240 00:15:21,691 --> 00:15:31,316 the other hand, they are very narrowly constrained in what they 241 00:15:31,316 --> 00:15:43,566 can do. And I was wondering if you could tell us the Turkish hazelnut 242 00:15:43,566 --> 00:15:53,191 story, because to me that's an example of where generalists might 243 00:15:53,191 --> 00:16:03,691 have an advantage of being able to search the entire world for 244 00:16:03,691 --> 00:16:11,566 opportunities. And of course, the disadvantage of generalists is 245 00:16:11,566 --> 00:16:21,191 that there's so much to potentially invest in that it’s very 246 00:16:21,191 --> 00:16:26,441 difficult to filter those ideas down. 247 00:16:27,871 --> 00:16:32,139 Yeah, so, I'm going to give the typical economist's answer of 248 00:16:32,139 --> 00:16:35,243 it depends. So, sometimes specialists outperform and sometimes 249 00:16:35,243 --> 00:16:36,019 generalists outperform. 250 00:16:36,711 --> 00:16:39,545 I’ll allow you to do that one time during the podcast. You can 251 00:16:39,545 --> 00:16:40,417 say it depends once. 252 00:16:41,091 --> 00:17:33,585 Yes, but I'll say what it depends on. So, this is not going 253 00:17:33,585 --> 00:18:26,079 to be complete fence sitting. So, what it depends on is where your 254 00:18:26,079 --> 00:19:10,497 expertise lies. So, there will be some investors who have genuine 255 00:19:10,497 --> 00:19:30,687 expertise in a specific sector. So, it may well be that it's 256 00:17:11,534 --> 00:17:59,373 run by some expositions and they're going to focus on pharmaceuticals 257 00:17:59,373 --> 00:18:38,514 because they really understand that particular sector. You don't 258 00:18:38,514 --> 00:19:30,702 want them to be investing in retail or textiles or something else. 259 00:16:41,091 --> 00:17:01,047 Others might have expertise in a particular theme. So it might be, 260 00:17:01,047 --> 00:17:17,677 I'm able to understand corporate culture. And so, I'm going 261 00:17:17,677 --> 00:17:35,970 to invest in companies with great workplaces, be this in pharmaceuticals, 262 00:17:35,970 --> 00:17:54,263 or financial services, or in retail. And indeed, I invested in 263 00:17:54,263 --> 00:18:17,545 such a fund for many, many years. It was a fund called the Parnassus 264 00:18:17,545 --> 00:18:34,175 Workplace Fund. And when Morningstar did a study, this was 265 00:18:34,175 --> 00:18:57,457 the best performing large cap fund over the last 1 and 3 and 5 266 00:18:57,457 --> 00:19:14,087 and 10 years. So, they were using still specialist expertise, 267 00:19:14,087 --> 00:19:30,717 but knowing that expertise could be applied to every sector. 268 00:16:41,091 --> 00:17:14,501 And then how does this go into the hazelnut example? So, this is 269 00:17:14,501 --> 00:17:40,201 a fund called TerraNova Capital, and their expertise is just 270 00:17:40,201 --> 00:18:13,611 trying to find stuff which is off the beaten path, news that is 271 00:18:13,611 --> 00:18:39,311 clearly very useful, but because it's in the dusty corners 272 00:18:39,311 --> 00:19:10,151 of the market, it might be ignored by other investors. And this 273 00:19:10,151 --> 00:19:30,711 is what takes us to hazelnut and Turkey. So, in April 2025, there 274 00:16:51,826 --> 00:17:19,737 was a huge frost which wiped out a third of the country's hazelnut 275 00:17:19,737 --> 00:17:43,354 crops. And maybe the listeners are not that familiar with Turkey, 276 00:17:43,354 --> 00:18:09,118 but hazelnuts to Turkey are what coffee is to Brazil. So, this 277 00:18:09,118 --> 00:18:34,882 is really big. And it was really big in Turkey, and Turkish 278 00:18:34,882 --> 00:18:56,352 language, local media covered it, but many Americans or Brits, 279 00:18:56,352 --> 00:19:22,116 they don't read Turkish. And so, it was not exactly hitting the 280 00:19:22,116 --> 00:19:30,704 front page of Bloomberg. And so, the wholesale prices began climbing, 281 00:16:57,135 --> 00:17:24,639 but there wasn't that change in the stock price. So, this fund 282 00:17:24,639 --> 00:17:52,143 just noticed it. It flew to Turkey. It found the country's only 283 00:17:52,143 --> 00:18:17,355 listed processor sitting on a huge trove of hazelnuts. And so, 284 00:18:17,355 --> 00:18:44,859 knowing that the price of hazelnuts will go up because of this 285 00:18:44,859 --> 00:19:07,779 drought, they invested $20 million. And when the shortage finally 286 00:19:07,779 --> 00:19:30,699 hit the headlines, the company's value went up by 40%. So, 287 00:16:43,265 --> 00:17:07,179 here their insight was, well, let's try to find relevant news. 288 00:17:07,179 --> 00:17:33,267 News which is ignored by the market, perhaps because it's in a 289 00:17:33,267 --> 00:18:01,529 different language. And I can do this in every sector, but I will 290 00:18:01,529 --> 00:18:23,269 still say that they are specialists in that their trading 291 00:18:23,269 --> 00:18:47,183 strategy is trying to find this obscure news, rather than saying, 292 00:18:47,183 --> 00:19:11,097 well, I believe I can run discounted cash flow models better 293 00:19:11,097 --> 00:19:30,663 than anybody else, which might not be so plausible. 294 00:19:30,961 --> 00:19:36,351 I see. So, you don't see them as generalists, you see them as specialists 295 00:19:36,351 --> 00:19:39,046 in a particular type of story, almost. 296 00:19:39,771 --> 00:19:52,587 Yes. So, specialism can be in theme, not just industry. So, we 297 00:19:52,587 --> 00:20:02,199 often will think about specialist versus generalist as whether 298 00:20:02,199 --> 00:20:13,947 you are sector focused or sector diversified. But I would still 299 00:20:13,947 --> 00:20:21,423 say, well, they have a specialist theme. I think it's very 300 00:19:41,663 --> 00:19:46,866 difficult for a fund to be successful without having that specialism. 301 00:19:46,866 --> 00:19:52,069 Because otherwise, to answer my question, what is your edge? It's 302 00:19:52,069 --> 00:19:57,272 not clear. So even Warren Buffett, even though he invests in 303 00:19:57,272 --> 00:20:03,421 lots of different sectors, he will say his edge is being able to 304 00:20:03,421 --> 00:20:07,678 analyze intangibles, companies with great brands. He helps them 305 00:20:07,678 --> 00:20:13,354 build a moat and he helps protect them from the vagaries of 306 00:20:13,354 --> 00:20:18,557 the short-term stock market by taking a large stake and allowing 307 00:20:18,557 --> 00:20:21,395 them to invest for many years. 308 00:20:22,631 --> 00:20:28,142 Just going, kind of circling back to this question of performance 309 00:20:28,142 --> 00:20:34,154 and benchmarks, and when I think about when I started my career, 310 00:20:34,154 --> 00:20:39,164 I actually started in the investment management world in the 311 00:20:39,164 --> 00:20:45,176 UK in the mid-1990s and that world was dominated by active investors 312 00:20:45,176 --> 00:20:49,685 who really didn't have benchmarks or the benchmarks were 313 00:20:49,685 --> 00:20:55,697 very loose. And that had evolved, I think, from the very old 314 00:20:55,697 --> 00:21:00,206 days of the Scottish investment trusts that were originally 315 00:21:00,206 --> 00:21:06,218 set up to just sort of scour the world for investment opportunities. 316 00:21:08,941 --> 00:21:12,406 And then we sort of (in some sense, I feel a bit bad about this), 317 00:21:12,406 --> 00:21:14,947 we came along and we were like, well, you're not outperforming 318 00:21:14,947 --> 00:21:17,719 the market. And we went to their investors and said, you'd be 319 00:21:17,719 --> 00:21:20,491 better off with an index fund or better off with a risk-controlled 320 00:21:20,491 --> 00:21:23,032 fund that was tightly managed relative to a market benchmark. And 321 00:21:23,032 --> 00:21:26,035 that was novel in the 1990s. Of course, it's not novel at all 322 00:21:26,035 --> 00:21:28,345 today. That's how most funds are run. They're tightly benchmarked. 323 00:21:31,721 --> 00:21:41,441 So, I guess my question is, have things come full circle now 324 00:21:41,441 --> 00:21:50,351 in that is there so much benchmarking that everyone is tightly 325 00:21:50,351 --> 00:21:58,451 constrained and you're better off looking for those that aren't 326 00:21:58,451 --> 00:22:00,071 benchmark constrained? When you think about your own personal 327 00:21:37,916 --> 00:21:45,881 investment, is that a consideration? Do you want people 328 00:21:45,881 --> 00:21:55,616 that are tightly constrained to a benchmark, or people that are 329 00:21:55,616 --> 00:22:00,041 looking for more absolute returns? 330 00:22:00,791 --> 00:22:18,809 Out of every bad and insane practice, there's always a grain, 331 00:22:18,809 --> 00:22:38,465 or actually more than a grain of sensibility behind it. So, often 332 00:22:38,465 --> 00:22:54,845 why there are seemingly incorrect practices is that a good 333 00:22:54,845 --> 00:23:14,501 idea was then just misapplied. So, benchmarking at its core is fundamentally 334 00:23:14,501 --> 00:23:35,795 sound. Why? Because if I'm going to choose to put my money with 335 00:23:35,795 --> 00:23:55,451 an active manager, he or she needs to outperform a passive benchmark. 336 00:23:55,451 --> 00:24:16,745 Why? Because I might be paying them fees of 75 basis points per 337 00:24:16,745 --> 00:24:36,401 year, whereas a passive fund may be charging just 5 basis points 338 00:24:36,401 --> 00:24:39,677 per year. So, I don't want to reward the manager for doing well 339 00:22:33,769 --> 00:23:15,741 in an up market. They need to be doing better than what an index 340 00:23:15,741 --> 00:23:48,719 fund would do. So, it's absolutely important to have a benchmark 341 00:23:48,719 --> 00:24:30,691 which is not zero. You need to be beating what a passive index fund 342 00:24:30,691 --> 00:24:39,685 would have done. But where this becomes a problem is then, if 343 00:22:16,172 --> 00:22:31,553 active managers become so conservative, they're afraid of deviating 344 00:22:31,553 --> 00:22:46,934 from the benchmark because then they will underperform. And 345 00:22:46,934 --> 00:23:09,151 so, they might think, well, as long as I market perform, I will 346 00:23:09,151 --> 00:23:31,368 still keep my job. And as a fund manager, I'm getting paid very 347 00:23:31,368 --> 00:23:57,003 well. And it's a lot of status to be a fund manager. I don't really 348 00:23:57,003 --> 00:24:20,929 need to be in the top 10 percentile as long as I'm sort of 349 00:24:20,929 --> 00:24:39,728 median, that's fine. I just don't want to be really underperforming. 350 00:22:00,791 --> 00:22:31,172 So, even if there's an AI bubble right now, and I know there's 351 00:22:31,172 --> 00:23:03,890 a bubble, I don't want to sell my AI stocks because if the bubble 352 00:23:03,890 --> 00:23:31,934 continues to inflate, I don't want to be the muppet who's still 353 00:23:31,934 --> 00:24:06,989 out of AI and missing out on this big boom. And so, if you do 354 00:24:06,989 --> 00:24:32,696 that, then you're actually not doing what your clients are paying 355 00:24:32,696 --> 00:24:39,707 you to do. Your clients have chosen to buy an active fund because 356 00:22:33,221 --> 00:23:05,651 they don't want to just passively hold the market portfolio. 357 00:23:05,651 --> 00:23:41,324 So, you need to have conviction and to move away from 358 00:23:41,324 --> 00:24:10,511 the benchmark. Otherwise, you're just giving them the same 359 00:24:10,511 --> 00:24:39,698 as an index fund for a much higher fee. So, there is a measure 360 00:22:07,581 --> 00:22:21,161 called active share, which was invented by some academics nearly 361 00:22:21,161 --> 00:22:37,457 20 years ago. And what they found was that this active share 362 00:22:37,457 --> 00:22:49,679 was generally correlated with outperformance. So, funds which were 363 00:22:49,679 --> 00:23:01,901 having conviction to deviate from the benchmark. Why? Because 364 00:23:01,901 --> 00:23:14,123 they're using their specialist knowledge. Even if that specialist 365 00:23:14,123 --> 00:23:26,345 knowledge is across different industries, they are still deviating 366 00:23:26,345 --> 00:23:41,283 from that benchmark by buying high culture firms or firms with 367 00:23:41,283 --> 00:23:53,505 strong intangibles. If you're using that expertise, you're going 368 00:23:53,505 --> 00:24:11,159 to be moving away from that benchmark. And so, the ones who are 369 00:24:11,159 --> 00:24:27,455 inactive, who are hugging the benchmark are the ones who are not 370 00:24:27,455 --> 00:24:39,677 using any expertise at all. They're just overcharging you. 371 00:24:40,371 --> 00:24:44,518 So that active share is potentially a proxy that people could 372 00:24:44,518 --> 00:24:48,288 use when they're evaluating funds, like how much conviction these 373 00:24:48,288 --> 00:24:52,435 managers have, or to the extent to which they're just kind 374 00:24:52,435 --> 00:24:56,205 of managing their professional risk by hugging the benchmark as 375 00:24:56,205 --> 00:25:00,352 opposed to delivering what you're looking for as an active investor. 376 00:25:00,601 --> 00:25:05,437 Absolutely. So, it highlights, well, who are the ones who are actually 377 00:25:05,437 --> 00:25:09,870 managing for their clients, which does involve, if you have unique 378 00:25:09,870 --> 00:25:14,303 insights, acting on the insight, and who are the ones who 379 00:25:14,303 --> 00:25:18,736 are overly conservative and just trying to manage their own career 380 00:25:18,736 --> 00:25:24,378 risk. Now, of course, we don't want to take this too far the other 381 00:25:24,378 --> 00:25:29,214 way because we know Goodhart's law. When a measure becomes a target, 382 00:25:29,214 --> 00:25:34,453 it ceases to be a good measure. If you have no unique ideas, 383 00:25:34,453 --> 00:25:38,886 don't just be active and deviate from the benchmark just to 384 00:25:38,886 --> 00:25:40,498 improve your active share. And so, this is why what's really important 385 00:25:07,401 --> 00:25:16,751 when evaluating funds is to see, well, what is their expertise? 386 00:25:16,751 --> 00:25:25,251 Do their active positions reflect what they claim their expertise 387 00:25:25,251 --> 00:25:36,301 to be? And if so, great. But if not, maybe they're just trading 388 00:25:36,301 --> 00:25:40,551 for the sake of trading. 389 00:25:41,441 --> 00:25:45,929 What do you think about a benchmark that's kind of more old-fashioned, 390 00:25:45,929 --> 00:25:50,417 that says, well, the equity risk premium is, let's call it 5% 391 00:25:50,417 --> 00:25:55,653 above cash. So, cash is 3%, I'm looking for 8%. And it's an absolute 392 00:25:55,653 --> 00:26:00,141 return fund, I'm scouring the world and don't evaluate me on performance 393 00:26:00,141 --> 00:26:04,255 this year or next year. Evaluate me over long horizons on 394 00:26:04,255 --> 00:26:08,743 whether I can do better than the market. And is that something 395 00:26:08,743 --> 00:26:13,231 that you find compelling or is that something that you think people 396 00:26:13,231 --> 00:26:17,345 should avoid? Because that's more like what a Warren Buffett would 397 00:26:17,345 --> 00:26:18,093 say. Right? 398 00:26:19,191 --> 00:26:38,619 It depends on how you're doing better than the market. Because we 399 00:26:38,619 --> 00:26:58,047 know from basic finance theory you can do better than the market 400 00:26:58,047 --> 00:27:17,475 by just taking on more systematic risk. So, if I am having 401 00:27:17,475 --> 00:27:38,522 a beta of 1.5, I should be delivering a 1.5 times greater excess 402 00:27:38,522 --> 00:27:56,331 return. So again, this is why benchmarking does have some sensibility 403 00:27:56,331 --> 00:28:17,378 behind it. We want to look at your alpha, which is your outperformance 404 00:28:17,378 --> 00:28:33,568 after taking risk into account, rather than just your raw 405 00:28:33,568 --> 00:28:40,044 unadjusted for risk outperformance. But if, indeed, what 406 00:26:29,435 --> 00:27:00,167 you're doing is, you are able to capitalize on markets in which 407 00:27:00,167 --> 00:27:28,338 you think there is an equity premium, and that equity premium 408 00:27:28,338 --> 00:28:01,631 is more than what it should be, given the level of risk, then 409 00:28:01,631 --> 00:28:37,485 there might be a way, there might be a reason to reward you for 410 00:28:37,485 --> 00:28:40,046 that. So, what do I mean by this? So, the equity premium puzzle 411 00:26:42,999 --> 00:27:00,855 is the following idea. So, stocks have outperformed treasury 412 00:27:00,855 --> 00:27:22,679 bills over any reasonable length of time. For example, if we 413 00:27:22,679 --> 00:27:50,455 look at the last 10 years up to December 2025, I think the T-bill 414 00:27:50,455 --> 00:28:18,231 average return was 2% and the S&P 500 was about 15%. So, that difference 415 00:28:18,231 --> 00:28:40,055 is 13%. Maybe on other time periods it might be 7%. Now, some 416 00:26:22,627 --> 00:26:44,961 people will say, well, that is just due to risk, that there's no 417 00:26:44,961 --> 00:27:03,859 free money on the table. People don't like risk and that's 418 00:27:03,859 --> 00:27:26,193 why the S&P 500 needs to give you a higher return. But actually, 419 00:27:26,193 --> 00:27:46,809 when people run the numbers, and they put into them any reasonable 420 00:27:46,809 --> 00:28:03,989 number for risk aversion and risk appetite, the equity premium 421 00:28:03,989 --> 00:28:26,323 should be something like 3% or 4%, not 7%. So, the stock market 422 00:28:26,323 --> 00:28:40,067 over any reasonable period does offer good value. And so, one 423 00:26:22,929 --> 00:26:39,127 argument is that, well, a fund, if it just goes for risk by 424 00:26:39,127 --> 00:26:54,079 buying into markets with high equity premia, that is a way that 425 00:26:54,079 --> 00:27:07,785 they can legitimately add value for your clients. Now, if your 426 00:27:07,785 --> 00:27:21,491 client was able to replicate it themselves just by buying index 427 00:27:21,491 --> 00:27:35,197 funds in those individual markets, then you would not be adding 428 00:27:35,197 --> 00:27:48,903 value. But if your client was sort of very unsophisticated and 429 00:27:48,903 --> 00:28:03,855 not able to pick out foreign markets for some reason, there is 430 00:28:03,855 --> 00:28:16,315 an argument for maybe if you're buying a globally actively 431 00:28:16,315 --> 00:28:28,775 managed fund, that active fund could go into particular markets 432 00:28:28,775 --> 00:28:39,989 where it thinks the equity premium is particularly rich. 433 00:28:46,091 --> 00:28:59,219 Let's talk about some of the kind of specific effects that you 434 00:28:59,219 --> 00:29:12,347 go through in your book. Momentum is one that you spend a 435 00:29:12,347 --> 00:29:23,287 lot of time with. Momentum is essentially trend following. And 436 00:29:23,287 --> 00:29:35,321 that's something that we see working through, you know, we see 437 00:29:35,321 --> 00:29:51,731 it in stocks, we see it in bonds, we see it kind of cross sectionally 438 00:29:51,731 --> 00:30:07,047 between stocks, we see it on a time series basis. It's a very strong 439 00:30:07,047 --> 00:30:08,141 effect. The tricky thing, of course, with momentum is that it 440 00:28:58,521 --> 00:29:13,437 has periodic crashes that are very violent. And if you're a professional 441 00:29:13,437 --> 00:29:25,867 investor like I was, and you were running momentum strategies, 442 00:29:25,867 --> 00:29:42,026 you can have multiple years of performance wiped out in a very short 443 00:29:42,026 --> 00:29:58,185 space of time. So, the kind of holy grail with trend following strategies 444 00:29:58,185 --> 00:30:08,129 is can you get out before the crash? And there's been some interesting 445 00:28:51,951 --> 00:29:07,187 research done over time on that. And you talk a little bit about 446 00:29:07,187 --> 00:29:21,251 that in your book. But I'm curious what you think the state 447 00:29:21,251 --> 00:29:37,659 of the art is with that and is it even possible? In other words, 448 00:29:37,659 --> 00:29:51,723 is the crash risk part of the reason that momentum makes money 449 00:29:51,723 --> 00:30:05,787 in the first place? So, you really aren't able to separate out 450 00:30:05,787 --> 00:30:08,131 the two. 451 00:30:08,561 --> 00:30:48,773 Yeah, so this is a great question. So, what is momentum and 452 00:30:48,773 --> 00:31:25,634 what psychological bias does this try to exploit? The idea of 453 00:31:25,634 --> 00:32:05,846 under reaction. So, even if we see some news, we might discount 454 00:32:05,846 --> 00:32:39,356 it. Why? Because of confirmation bias. We have our view 455 00:32:39,356 --> 00:33:22,919 of the world and any news that challenges that view, we just down 456 00:33:22,919 --> 00:33:46,376 weight it because we don't like it. So, let's just give an example. 457 00:30:25,985 --> 00:30:57,929 So, in 2006, Amazon quietly launched what ended up being Amazon 458 00:30:57,929 --> 00:31:29,873 Web Services. But back then, people never saw Amazon as being 459 00:31:29,873 --> 00:31:58,913 this cloud computing company. They thought, well, you buy books 460 00:31:58,913 --> 00:32:33,761 and CDs there. What is it doing playing in web services? So, 461 00:32:33,761 --> 00:33:08,609 it was only when it hit 2015, and Amazon Web Services actually 462 00:33:08,609 --> 00:33:31,841 reported their income separately, that people finally cottoned 463 00:33:31,841 --> 00:33:46,361 onto it being really valuable. And so, if we indeed ignore information, 464 00:30:32,018 --> 00:31:12,230 there is a simple strategy of buying stocks that have done well 465 00:31:12,230 --> 00:31:52,442 over the past six months and shorting stocks that have done badly 466 00:31:52,442 --> 00:32:36,005 over the past six months. Why? Because a stock that has done well 467 00:32:36,005 --> 00:33:12,866 recently had some very good news, but the market only thought 468 00:33:12,866 --> 00:33:46,376 it was moderately good, and so it underreacted to it. And over any 469 00:30:15,512 --> 00:30:36,365 reasonable time period, people have found that momentum typically 470 00:30:36,365 --> 00:30:54,901 outperforms. And interestingly, this is not just in 471 00:30:54,901 --> 00:31:25,022 the US, it's around the world. It's not just in equities, but it's 472 00:31:25,022 --> 00:31:48,192 in bonds, and currencies, and virtually any asset class. Which 473 00:31:48,192 --> 00:32:11,362 points to human psychology being a common driver. Because it 474 00:32:11,362 --> 00:32:32,215 doesn't matter whether you're a Latin American commodities trader 475 00:32:32,215 --> 00:32:55,385 or an American equities trader, you're all affected by confirmation 476 00:32:55,385 --> 00:33:25,506 bias and under reaction. But as you say, there is a danger, which 477 00:33:25,506 --> 00:33:46,359 is when momentum loses, it loses really, really badly. Now, 478 00:30:11,912 --> 00:30:45,422 fortunately, you can foresee these momentum crashes. And as I 479 00:30:45,422 --> 00:31:25,634 describe in the book, there's a really good study by two professors 480 00:31:25,634 --> 00:32:02,495 who also double up as investors for AQR and Goldman Sachs. 481 00:32:02,495 --> 00:32:42,707 What they found is that when VIX, which is the volatility index, 482 00:32:42,707 --> 00:33:16,217 is high, momentum is typically going to be underperforming. So, 483 00:33:16,217 --> 00:33:46,376 this tells you when to get out of momentum. Now, it is not a perfect 484 00:30:29,303 --> 00:31:03,873 predictor. And sometimes it may well be that momentum crashes 485 00:31:03,873 --> 00:31:41,900 unexpectedly. And that may well be the secret to answering your 486 00:31:41,900 --> 00:32:13,013 question. Why, even though momentum was first discovered in 487 00:32:13,013 --> 00:32:57,954 1993, is it still alive and well 33 years later? Why don't people 488 00:32:57,954 --> 00:33:35,981 exploit the free money on the table? Because people are worried 489 00:33:35,981 --> 00:33:46,352 about these crashes. And if I go back to our early discussion on 490 00:30:25,841 --> 00:30:44,849 career concerns, even if it's a strategy which will make money 491 00:30:44,849 --> 00:31:05,585 99% of the time, you're just weirdly worried about that 1% of 492 00:31:05,585 --> 00:31:29,777 the time where if it crashes you're going to be out of a job, 493 00:31:29,777 --> 00:31:48,785 even though it would be in your client's interest to invest 494 00:31:48,785 --> 00:32:07,793 in momentum. Why? Because your client probably is not just invested 495 00:32:07,793 --> 00:32:25,073 in your fund, they're investing in lots of other funds. 496 00:32:25,073 --> 00:32:44,081 So, they don't worry so much about your fund crashing because 497 00:32:44,081 --> 00:33:01,361 they're diversified. You as a fund manager, you're not diversified. 498 00:33:01,361 --> 00:33:23,825 Your entire reputation is tied to this one fund. And so, you might 499 00:33:23,825 --> 00:33:42,833 be excessively conservative and not pick up the free money on 500 00:33:42,833 --> 00:33:46,289 the table. And again, this is why, going back to one of our earlier 501 00:30:48,161 --> 00:31:24,461 conversations, why I think I can make the bold and controversial 502 00:31:24,461 --> 00:32:00,761 claim that perhaps sometimes amateurs might be able to beat the 503 00:32:00,761 --> 00:32:30,461 market more easily than professionals because if you're unconstrained 504 00:32:30,461 --> 00:33:06,761 and you're not worried about external clients, you might be able 505 00:33:06,761 --> 00:33:46,361 to exploit momentum because you don't worry so much about a crash. 506 00:33:47,191 --> 00:34:00,802 Yeah, I think I agree with that. And I guess, just talking a 507 00:34:00,802 --> 00:34:11,272 little bit more about momentum, there was one area where 508 00:34:11,272 --> 00:34:25,930 I think I didn't agree with you, and that was when you talk about 509 00:34:25,930 --> 00:34:35,353 the notion of corner strategies, which is actually, funnily 510 00:34:35,353 --> 00:34:48,964 enough, what we used to call them too. But the idea that you've 511 00:34:48,964 --> 00:35:01,528 got momentum that works across the market. But let's say it works 512 00:35:01,528 --> 00:35:11,998 particularly well if you filter down just the small caps. 513 00:35:11,998 --> 00:35:24,562 And then within small caps, it might work particularly well if you 514 00:35:24,562 --> 00:35:36,079 filter further down to, say, small caps that have negative earnings. 515 00:33:47,191 --> 00:33:59,635 Let's just say that's the case. And so, the more you apply 516 00:33:59,635 --> 00:34:11,042 these layers, you get to, “corners of the market” where momentum 517 00:34:11,042 --> 00:34:20,375 is potentially very strong. And, as a professional investor, 518 00:34:20,375 --> 00:34:31,782 we were very well aware of these corner strategies, but they're 519 00:34:31,782 --> 00:34:43,189 not really exploitable in any kind of scale. It's very difficult 520 00:34:43,189 --> 00:34:57,707 to put a lot of money to work if you're layering on three, four, 521 00:34:57,707 --> 00:35:10,151 or five different types of filters because you get a small number 522 00:35:10,151 --> 00:35:21,558 of stocks, that's number one. Number two is you get increasingly 523 00:35:21,558 --> 00:35:32,965 less statistical confidence in the result the more you narrow down 524 00:35:32,965 --> 00:35:36,076 the sample space. So, my thinking was, well, maybe these corners 525 00:33:59,127 --> 00:34:17,031 are an area that are more suited to individuals who don't need 526 00:34:17,031 --> 00:34:37,919 to put a lot of money to work. Now, the individuals might not have 527 00:34:37,919 --> 00:34:54,331 the sophistication to identify the corners, but it seemed to me 528 00:34:54,331 --> 00:35:09,251 that might be another area where, I don't know, sophisticated 529 00:35:09,251 --> 00:35:24,171 individuals or small funds could play, that big investors couldn't. 530 00:35:24,171 --> 00:35:36,107 I don't know, how you think about that? 531 00:35:36,441 --> 00:35:52,617 Yes, you're absolutely right. So, momentum is what I call a level 532 00:35:52,617 --> 00:36:11,489 one strategy. So, what do I mean by this is, we find a behavioral 533 00:36:11,489 --> 00:36:31,709 bias and we see how to exploit it. So, if the bias is that people 534 00:36:31,709 --> 00:36:45,189 underreact, then we want to buy recent winners, because those 535 00:36:45,189 --> 00:36:55,973 are stocks to which the market has underreacted. But then we can 536 00:35:39,529 --> 00:35:48,793 think about what I call a level two strategy, which is, where 537 00:35:48,793 --> 00:35:58,057 is that bias more prevalent? So, you think about, well, why is 538 00:35:58,057 --> 00:36:05,777 it that people underreact to information? Well, maybe they just 539 00:36:05,777 --> 00:36:13,497 don't even notice the information to begin with. Why? Because 540 00:36:13,497 --> 00:36:23,533 what do we see on Bloomberg, on the Wall Street Journal, we see 541 00:36:23,533 --> 00:36:32,025 coverage of the Magnificent Seven, or with AI stocks, or large 542 00:36:32,025 --> 00:36:39,745 companies like Amex and Goldman Sachs. And so, if there's 543 00:36:39,745 --> 00:36:49,009 a small company which has had some great news, we're less likely 544 00:36:49,009 --> 00:36:55,957 to notice it because it's not hitting the headlines. And so, the 545 00:35:39,711 --> 00:35:52,791 level two strategy is that we play the momentum game, but particularly 546 00:35:52,791 --> 00:36:03,691 in small stocks and stocks with little analyst coverage. And 547 00:36:03,691 --> 00:36:14,591 if, indeed, it is the case that large professional investors 548 00:36:14,591 --> 00:36:26,581 are unable to play that level two strategy, why, because small 549 00:36:26,581 --> 00:36:39,661 stocks are less liquid, then there is another reason for why individual 550 00:36:39,661 --> 00:36:51,651 investors, who don't have so much of a constraint, they might 551 00:36:51,651 --> 00:36:56,011 be able to outperform. 552 00:36:57,941 --> 00:37:04,031 Okay. Okay. So, it sounds like we're philosophically in agreement 553 00:37:04,031 --> 00:37:12,557 on that. I want to maybe talk about another area where I had a 554 00:37:12,557 --> 00:37:18,647 question of how maybe individuals could use one of your 555 00:37:18,647 --> 00:37:24,737 research findings. So, you talk about this fixation with round 556 00:37:24,737 --> 00:37:30,827 numbers, that we get unusual price movements when stocks move 557 00:37:30,827 --> 00:37:38,744 above or move below a round number, like US$50 a share or US$100 558 00:37:38,744 --> 00:37:46,052 a share. So, I was wondering if you could explain what happens 559 00:37:46,052 --> 00:37:48,488 around these large numbers. And then I want to ask a question 560 00:37:06,189 --> 00:37:15,468 about whether there's an implication for people setting limit 561 00:37:15,468 --> 00:37:28,871 orders when they want to buy or sell a stock for another reason. 562 00:37:28,871 --> 00:37:43,305 Can they use that round number effect to kind of help them be a 563 00:37:43,305 --> 00:37:48,460 little smarter about doing that? 564 00:37:49,141 --> 00:37:58,073 Yeah, absolutely. So, let's do this with a numerical example. So, 565 00:37:58,073 --> 00:38:08,629 let's say I buy a stock and it's currently US$80. I might think, 566 00:38:08,629 --> 00:38:17,561 well, my price target is US$100. Let's sell it at US$100, 567 00:38:17,561 --> 00:38:29,741 and I'll put a limit order to do this. But it may well be because 568 00:38:29,741 --> 00:38:37,049 other people like these round numbers. It's actually difficult 569 00:38:37,049 --> 00:38:46,793 for it to hit this US$100 because if other people have put 570 00:38:46,793 --> 00:38:56,537 these limit sell orders there. It might be that those sell orders 571 00:38:56,537 --> 00:39:05,469 get satisfied first and that drives the stock price down. Because 572 00:39:05,469 --> 00:39:12,777 of the selling pressure, it never really hits there. So, what 573 00:37:51,885 --> 00:38:12,465 I do with my own limit order selling is I might put the sell at 574 00:38:12,465 --> 00:38:27,557 say US$99.90. Why? Because it doesn't really matter to me much 575 00:38:27,557 --> 00:38:42,649 emotionally if I sell for US$99.90 versus US$100, once I realize 576 00:38:42,649 --> 00:38:59,113 that the round number is just some false barrier, and it makes 577 00:38:59,113 --> 00:39:12,833 it more likely that I will get my order filled. It's also the same 578 00:37:52,989 --> 00:38:06,457 on the buy side. So, I've actually put a buy order in for a 579 00:38:06,457 --> 00:38:18,001 fund where I put it in at US$12,510, not US$12,500. Why? Because 580 00:38:18,001 --> 00:38:31,469 that's again a round number. So, it may well be that if the fund 581 00:38:31,469 --> 00:38:42,051 price goes down, it might not hit US$12,500 because other people 582 00:38:42,051 --> 00:38:56,481 have put the buy order. So, what I want to do is front run. I 583 00:38:56,481 --> 00:39:09,949 want to get in before other people get in or get out before other 584 00:39:09,949 --> 00:39:12,835 people get out. 585 00:39:13,561 --> 00:39:17,201 Gotcha. Okay, so you do use this effect in your own...? 586 00:39:17,201 --> 00:39:23,281 Yes, I do it myself. And there was a fund which I had sold at US$14,000. 587 00:39:23,281 --> 00:39:27,841 Well, actually at US$13,990, and I think it's going to go down 588 00:39:27,841 --> 00:39:32,021 again. So, that's why I put a buy order at US$12,510. 589 00:39:32,501 --> 00:39:37,145 Okay. Maybe one last thing to talk about is the disposition effect. 590 00:39:37,145 --> 00:39:41,402 And you had a… So, you explained what the disposition effect 591 00:39:41,402 --> 00:39:47,207 is, and then you had a fix that I really liked. So, I was wondering 592 00:39:47,207 --> 00:39:51,077 maybe you could walk us through that, what the disposition 593 00:39:51,077 --> 00:39:54,947 effect is and then what your “fix” for it is. 594 00:39:55,391 --> 00:40:26,594 Absolutely. So, the disposition effect is such a common 595 00:40:26,594 --> 00:41:11,665 phenomenon in finance and it's such a big mistake, but it's so human 596 00:41:11,665 --> 00:41:42,868 and it's something where people understand why it happens. 597 00:41:42,868 --> 00:42:21,005 So, what is this? It's the idea that we're really reluctant 598 00:42:21,005 --> 00:42:59,142 to take losses and we really want to lock in winners. So, if I've 599 00:40:03,047 --> 00:40:36,223 bought a stock for US$80 and it bumbles around and it's now US$75, 600 00:40:36,223 --> 00:41:11,951 I really don't want to sell it. Why? Because then I have to admit 601 00:41:11,951 --> 00:41:47,679 to myself that I've made a loss. And even if I think it's going 602 00:41:47,679 --> 00:42:23,407 to go down to US$60, I hold on, just hoping it's going to get 603 00:42:23,407 --> 00:42:56,583 above US$80. And then when it hits US$80, then I'm going to be 604 00:42:56,583 --> 00:42:59,135 selling. And this is exactly the same that you might do with a 605 00:40:13,919 --> 00:40:32,447 casino if I've lost US$50, I will keep playing, and playing, and 606 00:40:32,447 --> 00:40:50,975 playing, until I recover that US$50, because I want to come out 607 00:40:50,975 --> 00:41:09,503 at least break even if it means throwing good money after bad. 608 00:41:09,503 --> 00:41:28,031 Why is this such bad behavior? Because of momentum, which we discussed 609 00:41:28,031 --> 00:41:45,015 earlier. If something has done poorly recently, if it's gone down 610 00:41:45,015 --> 00:42:03,543 from US$80 to US$75, it's likely to continue to go down in 611 00:42:03,543 --> 00:42:20,527 the future. And similarly, if something has gone up recently, it's 612 00:42:20,527 --> 00:42:42,143 likely to rise. But because I want to lock in again, I might sell 613 00:42:42,143 --> 00:42:59,127 the stock at US$90, even if it's heading up to US$100. And that's 614 00:40:00,043 --> 00:40:27,955 bad not only because of momentum, it's also bad for tax reasons, 615 00:40:27,955 --> 00:40:55,867 because if I sell a winner, I'm immediately locking in a tax, 616 00:40:55,867 --> 00:41:28,431 capital gains tax that I need to pay today. So, what is the solution 617 00:41:28,431 --> 00:41:56,343 to this? Well, one solution would be to say, oh, just ignore 618 00:41:56,343 --> 00:42:21,929 the purchase price, ignore the disposition effect. And if you were 619 00:42:21,929 --> 00:42:49,841 really strong willed, you might be able to do this, but not 620 00:42:49,841 --> 00:42:59,145 everybody is strong willed. It's just like to say only one Pringle, 621 00:40:24,791 --> 00:41:08,891 that is really hard to put into practice, even though it makes 622 00:41:08,891 --> 00:42:00,341 sense in theory. So, what I explore in the book is a more realistic 623 00:42:00,341 --> 00:42:44,441 solution, which is the idea of rolling your stocks over. What do 624 00:42:44,441 --> 00:42:59,141 I mean by this? Because if I was to sell my stock in Tesco for 625 00:40:16,225 --> 00:40:35,165 US$75 after buying it at US$80, and immediately reinvest the 626 00:40:35,165 --> 00:40:54,105 proceeds in Vodafone, I actually don't view myself as exiting 627 00:40:54,105 --> 00:41:14,939 that position because the trade is still live. I've just rolled 628 00:41:14,939 --> 00:41:33,879 it into a different investment. And indeed, in my Excel 629 00:41:33,879 --> 00:41:50,925 spreadsheet, I would have invested, say, maybe ten thousand 630 00:41:50,925 --> 00:42:09,865 pounds initially. It's gone down to nine thousand pounds. But 631 00:42:09,865 --> 00:42:30,699 I still know that my nine thousand pound investment is still 632 00:42:30,699 --> 00:42:53,427 carrying on and I'm not going to close out that position until 633 00:42:53,427 --> 00:42:59,109 I've sold Vodafone. So, by rolling on a stock into a new stock, 634 00:40:23,231 --> 00:40:59,423 I actually am able to take losses because I don't need to admit 635 00:40:59,423 --> 00:41:30,047 defeat to myself, because I've still kept that position going on. 636 00:41:30,047 --> 00:42:00,671 And indeed, there's a really clever study by some people which 637 00:42:00,671 --> 00:42:34,079 found that rolling on your position is a way of overcoming the 638 00:42:34,079 --> 00:42:59,135 disposition effect that you might have otherwise suffered from. 639 00:43:00,732 --> 00:43:14,676 Yeah, I thought that was a very clear explanation both in the 640 00:43:14,676 --> 00:43:32,106 book and right now. And also, a fix that I think we can all put 641 00:43:32,106 --> 00:43:46,050 into practice. So, while we're on that notion of fixes or things 642 00:43:46,050 --> 00:43:59,994 that individuals can do to take advantage of some of these ideas, 643 00:43:59,994 --> 00:44:13,938 I thought maybe we could focus a little bit more on that. You have 644 00:43:02,242 --> 00:43:11,302 a really nice summary at the end of the second chapter, which 645 00:43:11,302 --> 00:43:20,362 I thought maybe you could walk through. And it's really kind of 646 00:43:20,362 --> 00:43:29,422 three things that individuals can do to avoid some of these, avoid 647 00:43:29,422 --> 00:43:37,727 trading on emotion. And one is, don't check your portfolio too 648 00:43:37,727 --> 00:43:46,787 much. Two is, write down the reasons you're planning a trade and 649 00:43:46,787 --> 00:43:56,602 write down not only what the world looks like if it works, but 650 00:43:56,602 --> 00:44:06,417 also what the world would look like if it doesn't. And then three, 651 00:44:06,417 --> 00:44:13,967 pay attention to your mood when you make a decision. So, I was 652 00:43:07,389 --> 00:43:34,017 wondering maybe if you just give us the rationale for those three 653 00:43:34,017 --> 00:44:00,645 things. I think we actually covered some of those ideas and yeah, 654 00:44:00,645 --> 00:44:13,959 talk us through those three things. 655 00:44:14,292 --> 00:44:31,782 Absolutely. So, let's start with don't check your portfolio too 656 00:44:31,782 --> 00:44:51,021 much. And going back to my earlier comment that behind every 657 00:44:51,021 --> 00:45:12,009 bad behavior there is a grain of sensibility behind. So, we often 658 00:45:12,009 --> 00:45:32,997 say, oh, you snooze, you lose; the markets are fast moving, you 659 00:45:32,997 --> 00:45:52,236 should always be responding to information. And so, this will encourage 660 00:45:52,236 --> 00:46:11,475 us to always check our portfolio and always try to read 661 00:46:11,475 --> 00:46:32,463 the news to see should I be increasing this position or cutting 662 00:46:32,463 --> 00:46:37,710 back that position. But what we don't realize is that, well, if 663 00:44:30,033 --> 00:44:47,523 there's news out there, such as an earnings announcement, maybe 664 00:44:47,523 --> 00:45:05,013 somebody else has already responded to that and it's already 665 00:45:05,013 --> 00:45:24,252 more than priced in. So, if we're trading on information, it 666 00:45:24,252 --> 00:45:46,989 may well be that we're trading on old news. We're buying high because 667 00:45:46,989 --> 00:46:06,228 we're buying after the stock has already risen and we're selling 668 00:46:06,228 --> 00:46:27,216 low, which is the opposite of what an investor should be doing, 669 00:46:27,216 --> 00:46:37,710 which is buy low, sell high. So, if we are to not check our portfolio 670 00:44:29,304 --> 00:44:45,984 too often, which is actually advice that Warren Buffett gave, 671 00:44:45,984 --> 00:45:02,664 we're more likely to be investing on long-term factors such 672 00:45:02,664 --> 00:45:24,348 as the market being in a downturn in 2009 and I'm going to 673 00:45:24,348 --> 00:45:47,700 be willing to wait five or 10 years for it to recover rather than 674 00:45:47,700 --> 00:46:04,380 short-term things which might already be priced in. The second 675 00:46:04,380 --> 00:46:26,064 bit of advice is to always try to take the opposite position to 676 00:46:26,064 --> 00:46:37,740 the trade that I want to make. So, if right now I wanted to say, 677 00:44:25,652 --> 00:44:44,112 well, let's buy into an AI fund. Why? Because AI is the future. 678 00:44:44,112 --> 00:45:03,992 I've seen just how powerful AI is in my own life. Well, let's try 679 00:45:03,992 --> 00:45:21,032 to find reasons for not investing in AI. I could be looking 680 00:45:21,032 --> 00:45:36,652 at the valuation ratios and saying, well actually the price earnings 681 00:45:36,652 --> 00:45:56,532 ratio here is extremely high. And if I'm not able to come up with 682 00:45:56,532 --> 00:46:16,412 that myself, I could use AI to help me. I could ask ChatGPT, give 683 00:46:16,412 --> 00:46:36,292 me reasons for why I should not be investing in an AI fund right 684 00:46:36,292 --> 00:46:37,712 now. And this doesn't mean that we should always be inert and 685 00:44:31,628 --> 00:44:52,116 not taking any decision, but it does mean that when we take a 686 00:44:52,116 --> 00:45:11,028 decision, we will have seen both sides to it. It's just like 687 00:45:11,028 --> 00:45:36,244 if I want to buy a house, I want to make sure a survey is done 688 00:45:36,244 --> 00:45:50,428 to see whether it's structurally sound. This is just 689 00:45:50,428 --> 00:46:07,764 good due diligence. And I think we should treat any investment 690 00:46:07,764 --> 00:46:25,100 decision with the same seriousness as you would do any decision 691 00:46:25,100 --> 00:46:37,708 like buying a house or buying a car. 692 00:46:38,502 --> 00:46:42,961 How important is it to write down both sides of it as opposed 693 00:46:42,961 --> 00:46:47,420 to just like, oh, I'll do a Google search, I'll do ChatGPT search. 694 00:46:47,420 --> 00:46:52,222 We had Annie Duke on the show, who I know gave an endorsement to 695 00:46:52,222 --> 00:46:56,681 your book and she talked about this idea. And I believe, if I'm 696 00:46:56,681 --> 00:47:00,797 correct, she said, you really, you want to go through the exercise, 697 00:47:00,797 --> 00:47:04,570 but you've actually got to physically kind of write down what 698 00:47:04,570 --> 00:47:08,686 the world looks like both when you're right and when you're wrong, 699 00:47:08,686 --> 00:47:13,488 and then go back and kind of keep checking: is the world it getting 700 00:47:13,488 --> 00:47:17,261 closer to one of those two scenarios? Is that bit important 701 00:47:17,261 --> 00:47:21,720 in your mind, or is it more just, hey, think it through both 702 00:47:21,720 --> 00:47:22,406 sides carefully? 703 00:47:23,072 --> 00:47:33,576 I think it is very important. I think it's important on a number 704 00:47:33,576 --> 00:47:43,272 of levels. First, at the time that you make the investment decision, 705 00:47:43,272 --> 00:47:52,968 writing it down is really useful to make sure that you've done 706 00:47:52,968 --> 00:48:01,856 that process really carefully. So rather than just thinking in my 707 00:48:01,856 --> 00:48:12,360 head, oh yeah, I know that a danger might be that there's overvaluation. 708 00:48:12,360 --> 00:48:21,248 If I've written down, I'm concerned that AI could be overvalued. 709 00:48:21,248 --> 00:48:29,328 Why? Because the price earnings ratio is 50 and typically 710 00:48:29,328 --> 00:48:39,832 you might think it should be 30 or something else. This just makes 711 00:48:39,832 --> 00:48:49,528 sure that you are taking it seriously. Also writing it down makes 712 00:48:49,528 --> 00:48:56,800 sure that you are actually internalizing the information and 713 00:48:56,800 --> 00:49:02,456 it's going to be affecting your decision. This is why when you're 714 00:47:34,367 --> 00:47:56,957 going to a lecture at university, see, taking notes is 715 00:47:56,957 --> 00:48:24,065 really useful. Even if those notes end up catching on fire, the 716 00:48:24,065 --> 00:48:48,914 actual act of taking notes makes sure that you're more likely 717 00:48:48,914 --> 00:49:02,468 to be understanding and internalizing information. But also 718 00:47:26,970 --> 00:47:50,358 writing down is useful after the fact. Why? Because if the investment 719 00:47:50,358 --> 00:48:07,899 decision went south afterwards, then you will be able 720 00:48:07,899 --> 00:48:33,236 to go back to it and say, well, was this because my investment 721 00:48:33,236 --> 00:48:58,573 thesis was wrong? Because often what we do is we want to blame 722 00:48:58,573 --> 00:49:02,471 external factors. You might say, oh, my thesis was sound and 723 00:47:35,159 --> 00:47:51,275 it's just that the markets went crazy because Trump made a trade 724 00:47:51,275 --> 00:48:07,391 announcement. But actually, if you said, oh, my view is that, well, 725 00:48:07,391 --> 00:48:22,164 this stock, these valuations were high, but I think there could 726 00:48:22,164 --> 00:48:36,937 be further AI growth and, actually, that ended up not transpiring, 727 00:48:36,937 --> 00:48:54,396 then you can look at the outcome and compare it with the original 728 00:48:54,396 --> 00:49:02,454 rationale that you made for it. 729 00:49:03,862 --> 00:49:12,486 You know, after I had her on the show, I started doing, I actually 730 00:49:12,486 --> 00:49:18,646 started doing what she recommended, which is, I rarely follow 731 00:49:18,646 --> 00:49:26,654 through on things like that, but I did. And what I found was, 732 00:49:26,654 --> 00:49:33,430 because the process of writing down your positive thesis, your negative 733 00:49:33,430 --> 00:49:39,590 thesis, she also recommends you put down kill criteria, which 734 00:49:39,590 --> 00:49:46,982 is specific things that happen that are going to make you close 735 00:49:46,982 --> 00:49:51,910 that trade. That was like an involved process. And because it 736 00:49:06,580 --> 00:49:17,452 was so involved, I found I didn't check my portfolio as much 737 00:49:17,452 --> 00:49:28,324 because I didn't want to go through the whole process of thinking 738 00:49:28,324 --> 00:49:40,102 about a new trade in that same level of detail. So maybe that's 739 00:49:40,102 --> 00:49:50,068 just my personal laziness, but I'm wondering if those two things 740 00:49:50,068 --> 00:49:51,880 are related. 741 00:49:51,912 --> 00:50:07,476 Yeah, and I think it dovetails with advice that people give for 742 00:50:07,476 --> 00:50:20,446 just general good behavior and good habits outside of investing. 743 00:50:20,446 --> 00:50:34,713 So, while my book is about investing, investing is just the 744 00:50:34,713 --> 00:50:47,683 setting. It's about good decision making in general. So, when 745 00:50:47,683 --> 00:51:03,247 people say you want to develop a good habit, actually keeping score 746 00:51:03,247 --> 00:51:07,138 is really valuable. So, if my habit is to be, I'm going to check 747 00:49:58,743 --> 00:50:07,437 my phone less, what I used to have was a scorecard where, at the 748 00:50:07,437 --> 00:50:15,510 end of each day, I had to write down the number of minutes 749 00:50:15,510 --> 00:50:24,825 that I had been on my phone, because you can look at this on Apple 750 00:50:24,825 --> 00:50:32,277 screen time. And every day I would write down that number. And 751 00:50:32,277 --> 00:50:40,971 if a particular day the number was 105, I think, what was I doing 752 00:50:40,971 --> 00:50:50,286 in all that time? And yes, I could look at the end of the day 753 00:50:50,286 --> 00:50:58,980 and see it was 105. But I just feel more embarrassed when I write 754 00:50:58,980 --> 00:51:07,053 down it was 105. And I say, well, yesterday it was only 72. Also, 755 00:49:52,829 --> 00:50:04,750 you might think, oh, I want to do 30 minutes of cardiovascular exercise 756 00:50:04,750 --> 00:50:20,339 a day and I'm going to put a tick when I do it. But if I don't 757 00:50:20,339 --> 00:50:33,177 do it, I don't just leave it blank, I'm going to put a cross. 758 00:50:33,177 --> 00:50:45,098 Just to say that I have failed to myself. And that's something that 759 00:50:45,098 --> 00:50:55,185 people teach you, in terms of developing habits, is whenever you 760 00:50:55,185 --> 00:51:06,189 can, write it down and put feedback, you're more likely to internalize 761 00:51:06,189 --> 00:51:07,106 it. 762 00:51:08,372 --> 00:51:14,638 Let me ask you one more question. And all this, by the way, 763 00:51:14,638 --> 00:51:20,422 is coming from this last section of the book, which I really 764 00:51:20,422 --> 00:51:26,206 like, which you call Turning Madness into Money. And I think that's 765 00:51:26,206 --> 00:51:31,990 one of the, I think, really valuable aspects of your book, is 766 00:51:31,990 --> 00:51:38,738 that you end with these, I think it's 20 kind of the tools that 767 00:51:38,738 --> 00:51:44,522 I think individuals can take away. But one of them, which also 768 00:51:44,522 --> 00:51:49,824 goes back to what Annie Duke would recommend, is called Outsource 769 00:51:49,824 --> 00:51:55,126 the Ugly Decisions. And I think in her framing, it's like, 770 00:51:55,126 --> 00:52:01,392 hey, people tend to be better at the buy decision than the sell 771 00:52:01,392 --> 00:52:06,212 decision. So, you need someone external to yourself to basically 772 00:52:06,212 --> 00:52:10,068 pull that trigger and to do the sell. You tell them when you 773 00:51:13,432 --> 00:51:26,588 want, you know, why you put the position on, and when it's going 774 00:51:26,588 --> 00:51:40,756 to go wrong, you set the kill criteria, and they kill it for you. 775 00:51:40,756 --> 00:51:52,900 Is that what you had in mind with Outsource the Ugly Decisions? 776 00:51:52,900 --> 00:52:06,056 And do you do that yourself? Do you have someone who makes the 777 00:52:06,056 --> 00:52:10,104 ugly decisions for you? 778 00:52:11,012 --> 00:52:21,152 Yeah. So, this is to try to overcome the idea of the disposition 779 00:52:21,152 --> 00:52:29,732 effect, which is, it's difficult for you to reverse a losing 780 00:52:29,732 --> 00:52:38,312 position. Why? Because you are. You don't want to admit that 781 00:52:38,312 --> 00:52:50,012 you made a mistake. And so, it may well be that I'm going to invest 782 00:52:50,012 --> 00:52:59,372 in a single stock, let's say Goldman Sachs, right now, at US$100, 783 00:52:59,372 --> 00:53:11,852 and I will ask maybe my spouse to review it for me. And if she thinks, 784 00:53:11,852 --> 00:53:21,992 oh, actually it's gone down too much and it's going to go down 785 00:53:21,992 --> 00:53:31,352 even further, I might not delegate it fully to her by allowing 786 00:53:31,352 --> 00:53:43,052 her to sell, but she is able to tell me, Alex, you need to get 787 00:53:43,052 --> 00:53:43,832 out. Similarly, within a company, it may well be the risk 788 00:52:20,782 --> 00:52:29,575 manager or the chief investment officer, who is saying, 789 00:52:29,575 --> 00:52:43,253 well, I'm going to be telling you when to get out of this. So, 790 00:52:43,253 --> 00:52:57,908 the big question is, do I do it myself? And I might say no. And 791 00:52:57,908 --> 00:53:11,586 you might think, oh, this is crazy. How can I read the book of 792 00:53:11,586 --> 00:53:22,333 somebody who doesn't follow his own advice? But the reason why 793 00:53:22,333 --> 00:53:34,057 it is that I invest very little in single stocks to begin 794 00:53:34,057 --> 00:53:43,827 with, why? Because I believe in the power of diversification. 795 00:52:11,012 --> 00:52:28,029 I invest most of my money within either passive index funds 796 00:52:28,029 --> 00:52:43,499 or actively managed funds, where I believe the fund manager 797 00:52:43,499 --> 00:53:05,157 has a lot of skills, such as talent over capital. But I do implement 798 00:53:05,157 --> 00:53:23,721 the principle in other areas of life, which is similar to the 799 00:53:23,721 --> 00:53:39,191 idea that financial markets are only a laboratory for decision 800 00:53:39,191 --> 00:53:43,832 making more generally. 801 00:53:44,482 --> 00:53:53,032 So, whenever I write a book, I will try to send it to other people 802 00:53:53,032 --> 00:54:00,442 and say, just please tear it apart as much as possible. Why? Because 803 00:54:00,442 --> 00:54:07,282 I will review it myself many, many times, but I'm very unlikely 804 00:54:07,282 --> 00:54:12,982 to completely scrap a section because I've thought, well, I've 805 00:54:12,982 --> 00:54:20,392 spent a day of my life writing this section or this chapter. Even 806 00:54:20,392 --> 00:54:26,662 though I, yes, I edit something, I might just change this 807 00:54:26,662 --> 00:54:34,642 word or the sentence. I'm not going to be cutting out a whole story. 808 00:54:34,642 --> 00:54:40,912 So, I'll send it to some people with explicit requests. Can 809 00:54:40,912 --> 00:54:47,752 you tell me which of these sections are just not pulling their 810 00:54:47,752 --> 00:54:51,172 weight completely? Get rid of this. And so those ugly decisions 811 00:53:46,862 --> 00:53:53,526 I will send to other people. For my last book, (which is the first 812 00:53:53,526 --> 00:53:59,238 book I went with an agent because it was published with Penguin 813 00:53:59,238 --> 00:54:04,474 Random House, a mainstream publisher), out of all of the agents 814 00:54:04,474 --> 00:54:08,758 that offered me representation, I went with the most 815 00:54:08,758 --> 00:54:14,470 negative agent, the one who was the most critical of my proposal. 816 00:54:14,470 --> 00:54:19,230 There were other agents who said, oh, your proposal's great, 817 00:54:19,230 --> 00:54:24,942 we can send it off to publishers right now. But I thought, 818 00:54:24,942 --> 00:54:31,606 well, this is the guy who will be able to improve my proposal the 819 00:54:31,606 --> 00:54:38,270 most. And so, he did the ugly decisions for me and said, oh, this 820 00:54:38,270 --> 00:54:43,982 really is not earning its weight and this needs to be improved. 821 00:54:43,982 --> 00:54:51,122 And so, I will try to apply this in many other areas of my life. 822 00:54:51,652 --> 00:54:59,075 Nice, I appreciate that. Maybe we can wrap up by just talking very 823 00:54:59,075 --> 00:55:05,356 briefly about the kind of current situation. There was a number 824 00:55:05,356 --> 00:55:12,208 of things in your book that I think are relevant or potentially 825 00:55:12,208 --> 00:55:19,631 relevant for what's going on today. One of them is what you call 826 00:55:19,631 --> 00:55:26,483 the big market delusion. And I was wondering if you could explain 827 00:55:26,483 --> 00:55:32,193 to us what that is and then what's your view? Are we seeing this 828 00:54:56,568 --> 00:55:11,316 big market delusion in a form with the AI hyperscalers? Are they 829 00:55:11,316 --> 00:55:24,835 falling victim to the big market delusion or investors in pricing 830 00:55:24,835 --> 00:55:32,209 the hyperscalers falling victim to that? 831 00:55:33,112 --> 00:55:51,762 Certainly. So, I'll first explain what the big market delusion 832 00:55:51,762 --> 00:56:16,007 is and then apply it to AI hyperscalers right now. So, the big 833 00:56:16,007 --> 00:56:36,522 market delusion is the idea that when there is excitement about 834 00:56:36,522 --> 00:56:57,037 a new technology, often the prices of every company within that 835 00:56:57,037 --> 00:57:15,687 industry will rise up significantly. So much so that if 836 00:57:15,687 --> 00:57:36,202 you added all of those companies together, you'd get a valuation 837 00:57:36,202 --> 00:57:56,717 of the industry, which is just implausible. So, the numbers just 838 00:57:56,717 --> 00:58:02,312 don't add up. So, in the early 2020s, if you were to take the value 839 00:56:06,268 --> 00:56:33,898 of the electric vehicle industry, adding up Nio, and Tesla, 840 00:56:33,898 --> 00:57:09,817 and Xpeng, and so on, you got a combined valuation which was even 841 00:57:09,817 --> 00:57:40,210 higher than the traditional car market. And that was about five 842 00:57:40,210 --> 00:58:02,314 years ago when electric vehicles were still nascent. Now 843 00:55:35,684 --> 00:56:06,548 it is indeed true that one of those companies may well succeed 844 00:56:06,548 --> 00:56:34,840 and take over the world and become extremely valuable. But the 845 00:56:34,840 --> 00:57:00,560 market was pricing the industry as if every single company 846 00:57:00,560 --> 00:57:31,424 would be a winner. And there's an adding up constraint. Not every 847 00:57:31,424 --> 00:57:59,716 single company can become a winner. Not everybody can get gold 848 00:57:59,716 --> 00:58:02,288 medals. And so, the whole industry was being overpriced. And 849 00:55:43,201 --> 00:55:58,895 why might this be? It might be the fear of missing out. You don't 850 00:55:58,895 --> 00:56:14,589 want to just own Nio and Tesla because what if BYD ends up taking 851 00:56:14,589 --> 00:56:28,041 the lion's share of the market? And so, then if you apply 852 00:56:28,041 --> 00:56:41,493 this to AI right now, even though, yes, AI will absolutely be 853 00:56:41,493 --> 00:56:53,824 the future, not every AI hyperscaler will win. It may well 854 00:56:53,824 --> 00:57:08,397 be that some of them end up over investing or investing in the 855 00:57:08,397 --> 00:57:21,849 wrong stuff. So, this is why I, myself, reduced my position in 856 00:57:21,849 --> 00:57:35,301 AI and tech more generally about six to nine months ago. And 857 00:57:35,301 --> 00:57:50,995 so, I don't know whether this will be a good decision in the long 858 00:57:50,995 --> 00:58:02,205 term. Certainly, I'm not unhappy that I've exited from that. 859 00:55:33,112 --> 00:55:49,840 Also I know that my immediate investment objective is that I have 860 00:55:49,840 --> 00:56:06,568 a large mortgage coming up for refinancing about one year from now. 861 00:56:06,568 --> 00:56:24,690 And because the AI in my portfolio has done so well, I wanted 862 00:56:24,690 --> 00:56:41,418 to lock in those gains. So, I'm not so worried about missing 863 00:56:41,418 --> 00:56:59,540 out on perhaps future gains by not selling at the absolute top because 864 00:56:59,540 --> 00:57:20,450 I know that it had a really good ride. What I want to avoid is 865 00:57:20,450 --> 00:57:41,360 if we are in a bubble, if we are in a big market delusion, and 866 00:57:41,360 --> 00:57:58,088 the whole market then turns south, as it did for electric vehicles 867 00:57:58,088 --> 00:58:02,270 in 2022, 2023. 868 00:58:02,892 --> 00:58:06,313 Cool. Well, I appreciate that and I appreciate all the stories. 869 00:58:06,313 --> 00:58:09,423 This is a really fun conversation. We could easily keep 870 00:58:09,423 --> 00:58:14,088 it going, or at least I could. You probably have a day job to get 871 00:58:14,088 --> 00:58:18,442 to, but that's a good place to wrap up, Alex. Thanks so much for 872 00:58:18,442 --> 00:58:22,796 taking the time to join us and to share your research and we wish 873 00:58:22,796 --> 00:58:24,351 you the best of luck. 874 00:58:24,612 --> 00:58:26,904 You're very welcome. Kevin, thank you so much for having me on. 875 00:58:26,904 --> 00:58:28,050 I really enjoy the conversation myself. 876 00:58:29,252 --> 00:58:37,196 Okay. The book is called The Madness of Markets: Why Smart Investors 877 00:58:37,196 --> 00:58:45,140 Make Crazy Decisions and How to Exploit Them. It's out now, so 878 00:58:45,140 --> 00:58:53,084 please make sure to get a copy and to follow Alex's work. As always, 879 00:58:30,386 --> 00:58:36,056 what we've discussed today really is just scratching the surface 880 00:58:36,056 --> 00:58:43,427 of what you'll learn from the book. So that's it for today. For 881 00:58:43,427 --> 00:58:50,231 all of us here at Top Traders Unplugged. Thanks for listening and 882 00:58:50,231 --> 00:58:53,066 we'll see you next time. 883 00:58:55,092 --> 00:58:58,412 Thanks for listening to Top Traders Unplugged. If you feel you 884 00:58:58,412 --> 00:59:01,212 learned something of value from today's episode, the best way 885 00:59:01,212 --> 00:59:04,332 to stay updated is to go on over to your favorite podcast platform 886 00:59:04,332 --> 00:59:07,332 and follow the show so that you'll be sure to get all the new 887 00:59:07,332 --> 00:59:10,772 episodes as they're released. We have some amazing guests lined 888 00:59:10,772 --> 00:59:13,662 up for you and to ensure a show continues to grow, please leave 889 00:59:13,662 --> 00:59:16,662 us an honest rating and review. It only takes a minute and 890 00:59:16,662 --> 00:59:20,062 it's the best way to show us you love the podcast. We'll see you 891 00:59:20,062 --> 00:59:24,102 next time on Top Traders Unplugged. This podcast expresses 892 00:59:24,102 --> 00:59:26,742 the views of its hosts and the guests appearing on the podcast as 893 00:59:26,742 --> 00:59:29,022 of the date of its recording, and such views are subject to change 894 00:59:29,022 --> 00:59:31,502 without notice. Top Traders Unplugged do not have any duty or 895 00:59:31,502 --> 00:59:34,262 obligation to update the information contained herein. Furthermore, 896 00:59:34,262 --> 00:59:37,102 Top Traders Unplugged make no representation to its accuracy and 897 00:59:37,102 --> 00:59:39,872 it shall not be assumed that past investment performance is an 898 00:59:39,872 --> 00:59:42,752 indication of future results. Moreover, wherever there is a potential 899 00:59:42,752 --> 00:59:45,392 for profit, there is also the possibility of loss. This content 900 00:59:45,392 --> 00:59:47,712 is made available for educational purposes only and should 901 00:59:47,712 --> 00:59:50,032 not be used for any other purpose. The information contained 902 00:59:50,032 --> 00:59:52,592 in this podcast does not constitute and should not be construed 903 00:59:52,592 --> 00:59:55,472 as investment advice or an offer to sell or a solicitation to 904 00:59:55,472 --> 00:59:58,792 buy any securities or related financial instruments in any jurisdiction. 905 00:59:58,792 --> 01:00:01,392 Certain information contained herein concerning economic trends 906 01:00:01,392 --> 01:00:04,402 and performance is based on or derived from information provided 907 01:00:04,402 --> 01:00:07,002 by independent third party sources. Top Traders Unplugged may 908 01:00:07,002 --> 01:00:09,282 believe that the sources from which such information are obtained 909 01:00:09,282 --> 01:00:11,682 are reliable. However, Top Traders Unplugged cannot guarantee 910 01:00:11,682 --> 01:00:14,042 the accuracy of such information and has not independently 911 01:00:14,042 --> 01:00:16,402 verified the accuracy or completeness of such information 912 01:00:16,482 --> 01:00:19,562 or the assumptions on which such information is based. This podcast, 913 01:00:19,562 --> 01:00:22,402 including the information contained herein, may not be reproduced, 914 01:00:22,402 --> 01:00:25,442 copied, republished or posted in whole or in part in any form, 915 01:00:25,442 --> 01:00:28,082 without the prior written consent of Top Traders Unplugged.