WEBVTT - Interview With William Sharpe: Masters in Business (Audio) 0:00:00.320 --> 0:00:02.600 The future may not be clear, but our commitment is 0:00:02.759 --> 0:00:04.600 so when you sit with an advisor at Merrill Lynch, 0:00:04.680 --> 0:00:07.360 we'll put your interests first. Visit mL dot com and 0:00:07.440 --> 0:00:09.799 learn more about Merrill Lynch, an affiliative Bank of America. 0:00:09.920 --> 0:00:12.400 Mery Lynch makes available products and services offered by Merrill Lynch. 0:00:12.440 --> 0:00:14.800 Pierce Veteran Smith Incorporated, a register broker dealer, remember s 0:00:14.800 --> 0:00:20.720 I PC. This is Masters in Business with Barry Ridholts 0:00:20.880 --> 0:00:27.000 on Bloomberg Radio this week on the show, What a delight. 0:00:27.120 --> 0:00:30.080 I don't even know how to begin to explain this. 0:00:30.960 --> 0:00:35.879 Uh Bill Sharp, just a legend in finance, couldn't have 0:00:35.920 --> 0:00:39.920 been more charming and more delightful. There was a lot 0:00:40.040 --> 0:00:43.240 of back and forth in terms of when you guys 0:00:43.320 --> 0:00:45.120 going to be on the West Coast. I'm here, I'm there, 0:00:45.159 --> 0:00:49.280 I'm busy week and he no longer keeps in office 0:00:49.320 --> 0:00:52.720 at Stanford, and he didn't want to come into San Francisco. 0:00:52.840 --> 0:00:56.120 He lives down in Carmel, and the guys in Andrews 0:00:56.120 --> 0:00:59.040 and Horowitz were nice enough to give me a podcasting 0:00:59.120 --> 0:01:03.040 room to sit in and record the show. I can't 0:01:03.080 --> 0:01:06.679 begin to tell you what a delight it was to 0:01:06.880 --> 0:01:12.959 chat with him. He's just beyond knowledgeable. He really helped, 0:01:13.400 --> 0:01:16.320 you know, the original concept of the show is speaking 0:01:16.319 --> 0:01:20.480 to the minds who helped shape finance and investing in business, 0:01:20.560 --> 0:01:24.280 and who who better than Bill Sharp. So many of 0:01:24.319 --> 0:01:28.200 the principles that we just accept as as running the 0:01:28.240 --> 0:01:32.360 middle ordinary things in finance, Bill Sharp helped to create. 0:01:32.880 --> 0:01:35.600 He helped to create the first index fund, He helped 0:01:35.600 --> 0:01:38.959 to create the capital asset pricing model. He helped to 0:01:39.000 --> 0:01:43.520 create our understanding of risk. Just stop and think about 0:01:44.040 --> 0:01:47.639 those achievements. We we talked about a number of really 0:01:47.680 --> 0:01:51.000 fascinating things, and I can't begin to tell you how 0:01:51.120 --> 0:01:54.720 just delightful it is. So my job was to kind 0:01:54.720 --> 0:01:56.840 of get out of his way and just keep nudging 0:01:56.920 --> 0:02:01.040 him in into expounding more and more. Uh, some of 0:02:01.040 --> 0:02:03.880 the stuff is a little wonky. It's not your usual 0:02:04.360 --> 0:02:07.840 market based conversation because he's the guy who built the 0:02:07.920 --> 0:02:13.440 underlying infrastructure. Suffice it to say, it's absolutely fascinating. So, 0:02:13.520 --> 0:02:18.240 with no further ado, my conversation with Nobel Laureate William Sharp, 0:02:22.000 --> 0:02:26.520 Professor Bill Sharp, Welcome to Bloomberg, hosted here at Andres 0:02:26.520 --> 0:02:31.560 and Harrowitz. Your dissertation was called the single index model. 0:02:32.080 --> 0:02:35.040 Tell us what that was about. Okay, it'll be won't 0:02:35.040 --> 0:02:37.520 be as short as possible answer, but it'll be a 0:02:37.520 --> 0:02:41.239 lot shorter than the dissertation the Harry Marco What's His 0:02:41.320 --> 0:02:44.520 work was what we would call normative, in the sense 0:02:44.600 --> 0:02:48.720 that he was asking the question, your portfolio manager, there 0:02:48.720 --> 0:02:51.600 are securities, there's a client. How do you build a 0:02:51.639 --> 0:02:57.240 portfolio that's good for the client. And in his structure 0:02:57.880 --> 0:03:00.960 he allowed for not only as tom to the expected 0:03:01.000 --> 0:03:04.600 return on general motors, let's say, the risk of general 0:03:04.639 --> 0:03:07.720 motors stock, but also the extent to which general motors 0:03:07.720 --> 0:03:11.119 would move with Ford, with general foods, etcetera. A whole 0:03:11.200 --> 0:03:15.640 lot of number precisely, so, a whole lot of numbers. 0:03:16.120 --> 0:03:19.320 And he had a procedure to find a so called 0:03:19.520 --> 0:03:24.000 set of efficient portfolios given that set of numbers, and 0:03:24.080 --> 0:03:26.280 it took a big computer, a lot of time, a 0:03:26.280 --> 0:03:29.680 lot of money to do it. Um. He also, in 0:03:29.760 --> 0:03:34.600 his in his work, had suggested you might simplify the 0:03:34.680 --> 0:03:39.960 relationships among securities, and he proposed a number of versions, 0:03:40.480 --> 0:03:44.920 one of which was, well, you could say general motors 0:03:44.960 --> 0:03:48.400 moved with the market to a certain extent, general foods 0:03:48.440 --> 0:03:51.240 moved with the market to a certain extent, each of 0:03:51.280 --> 0:03:55.440 them had movements on their own and leave it at that. 0:03:55.520 --> 0:03:58.920 In other words, a very simple model in which there 0:03:58.960 --> 0:04:02.560 was one single index, let's call it the market for 0:04:02.560 --> 0:04:08.120 for now, which created all the correlation, all the co 0:04:08.320 --> 0:04:14.720 movement among securities. So he essentially before we talked about data, 0:04:14.960 --> 0:04:17.600 he had to come up with the concept of, well, 0:04:17.640 --> 0:04:22.200 here's what the market actually well, and I hope I 0:04:22.279 --> 0:04:25.560 recalling correctly. In his version, this was just, oh, you 0:04:25.640 --> 0:04:28.240 might want to make this simple assumption. And then there 0:04:28.240 --> 0:04:30.880 were a couple of other authors that were writing with 0:04:31.000 --> 0:04:33.840 that same kind of structure, but there was no sense 0:04:34.279 --> 0:04:37.240 this would be the market portfolio was just a thing, 0:04:37.360 --> 0:04:41.320 a single index. But but so I. So what I 0:04:41.360 --> 0:04:46.000 did was I I took that concept and I did 0:04:46.080 --> 0:04:50.080 three things in the dissertation. One was I wrote a 0:04:50.760 --> 0:04:55.159 computer algorithm that could take advantage of that simplified structure 0:04:55.880 --> 0:04:59.960 and find efficient portfolios, you know, for orders of MAGA 0:05:00.000 --> 0:05:03.240 it to less computer time. Then if you expanded it 0:05:03.279 --> 0:05:05.880 to the full structure. So the first part of the 0:05:05.880 --> 0:05:10.279 dissertation was an algorithm and a foretrend program. Probably the 0:05:10.320 --> 0:05:14.320 first dissertation and economics at U c L A. That 0:05:14.400 --> 0:05:19.960 included programs um The second I UH Fred Weston had 0:05:19.960 --> 0:05:22.760 a friend who was an investment advisor, a real human 0:05:22.839 --> 0:05:26.520 investment advisor, and so I worked with him to try 0:05:26.560 --> 0:05:30.640 to get him to make probabilistic forecasts for a group 0:05:30.680 --> 0:05:34.640 of securities, and then we ran them through the algorithm 0:05:34.680 --> 0:05:37.760 to see what it implied. And then the third I 0:05:38.200 --> 0:05:41.400 did what my training as a micro economist, which is 0:05:41.960 --> 0:05:46.320 most of my training, UH, would cause me to do, 0:05:46.760 --> 0:05:49.440 what if everybody did this? What if everybody did what 0:05:49.520 --> 0:05:53.360 Harry said? What would the world look like? So Harry's 0:05:53.400 --> 0:05:58.159 portfolio theorem would basically guide all the investors. That was 0:05:58.240 --> 0:06:01.479 my assumption. So I turned from what should you do 0:06:01.839 --> 0:06:07.240 normative to how might the world work? Positive theory, which 0:06:07.279 --> 0:06:11.520 is what economists at that time in particular generally did, 0:06:11.800 --> 0:06:15.640 and said, well, if everybody did this and markets cleared 0:06:15.839 --> 0:06:20.800 and prices adjusted, what would be the relationship in equilibrium 0:06:21.680 --> 0:06:26.119 between expected returns on securities and some measure of risk? 0:06:26.800 --> 0:06:30.679 And the conclusion was again this was positing, this single 0:06:30.720 --> 0:06:36.000 index model. Conclusion was that the thing, the common factor 0:06:36.080 --> 0:06:41.400 that would matter would be the market portfolio. And that's 0:06:41.440 --> 0:06:46.200 when the term beta came to be, and that expected 0:06:46.240 --> 0:06:52.279 returns would be related only to beta's in a linear manner. 0:06:52.279 --> 0:06:54.680 For that matter, were you aware at the time how 0:06:55.080 --> 0:06:59.919 innovative and groundbreaking and influential this idea would be going forward? 0:07:00.200 --> 0:07:02.200 Was it? Oh, I think I have a pretty nice 0:07:02.200 --> 0:07:06.800 dissertation here the ladder, the ladder. And then I finished 0:07:06.839 --> 0:07:10.120 the dissertation in June and then started teaching the University 0:07:10.160 --> 0:07:13.800 of Washington in September, and you know, I'd written up 0:07:14.200 --> 0:07:17.480 the algorithm for a paper, and uh now I was saying, 0:07:17.520 --> 0:07:20.880 this is really a nifty result. This they'd you know, 0:07:21.320 --> 0:07:24.800 x equilibrium result. But it's sort of like, Okay, I 0:07:24.840 --> 0:07:26.960 pulled a rabbit out of a hat. But but I 0:07:27.080 --> 0:07:30.680 put it in beforehand with this single index model assumption, 0:07:31.240 --> 0:07:33.560 and I said, boy, it would be nice if I 0:07:33.560 --> 0:07:37.560 could get that beautiful answer without making this assumption which 0:07:37.600 --> 0:07:42.480 almost directly created the answer. So I I noodled around 0:07:42.520 --> 0:07:46.240 and talked to colleagues and um thought of it, and 0:07:46.240 --> 0:07:48.160 then all of a sudden, without you know, within two 0:07:48.200 --> 0:07:50.360 or three or four months, said wait a minute, I 0:07:50.400 --> 0:07:53.040 don't have to make that assumption. I can get that 0:07:53.160 --> 0:07:58.200 result in a general situation. And at that point that 0:07:58.280 --> 0:08:01.480 was the capital has at pricing model m HM and uh. 0:08:01.680 --> 0:08:04.400 I first wrote it up and submitted it for publication 0:08:04.480 --> 0:08:08.800 in sixty two, and it was rejected by a referee. 0:08:08.840 --> 0:08:13.000 And then how fantastic is that? Yeah? And and I 0:08:13.120 --> 0:08:16.600 finally published it in sixty four. It took you three 0:08:16.680 --> 0:08:21.600 years to get your paper on capital, and that ultimately 0:08:21.840 --> 0:08:25.040 is what leads That was what was cited in the 0:08:25.040 --> 0:08:29.320 Nobel Prize. Yeah. Isn't that fascinating that such an interesting 0:08:29.360 --> 0:08:33.280 innovative idea rejected for a few years. Well, you know, 0:08:33.320 --> 0:08:35.800 I happen to know. I found out who the referee 0:08:35.880 --> 0:08:39.320 was and his argument was, well, that's an unrealistic assumption. 0:08:39.400 --> 0:08:43.079 And and I was taught by Milton Friedman indirectly in 0:08:43.200 --> 0:08:47.959 others that you don't evaluate a theory of that sort 0:08:48.080 --> 0:08:52.480 on the assumption. You evaluated on the conformance of the 0:08:52.559 --> 0:08:56.200 results with the real world, because there is always making something. 0:08:56.520 --> 0:08:59.720 The thesis is, we're gonna start with these assumptions. Where 0:08:59.760 --> 0:09:04.000 does it lead to? It leads us here? Isn't that interesting? Exactly? So? 0:09:04.640 --> 0:09:08.200 Um So? In any event, and as a matter of fact, 0:09:08.280 --> 0:09:11.680 when it was finally published in the journal Finance, you know, 0:09:11.720 --> 0:09:14.160 I asked as far as the refereeing. I asked it, 0:09:14.200 --> 0:09:17.040 what could we have another referee? Please? None? They changed 0:09:17.200 --> 0:09:20.520 editors and such. But when it was published, I thought, well, 0:09:21.360 --> 0:09:23.000 this is the best thing I'm ever going to do, 0:09:23.040 --> 0:09:27.560 and in that I was correct. And and I sat 0:09:27.640 --> 0:09:30.440 by the phone. We didn't have computers in those days, 0:09:30.600 --> 0:09:34.000 no no text or emails, waiting for the phone to ring. 0:09:34.080 --> 0:09:36.160 And it didn't ring, and it didn't ring, and I thought, 0:09:36.840 --> 0:09:38.960 you know, months passed and that, man, I've just written 0:09:39.000 --> 0:09:42.079 the best paper I'm ever going to write, and nobody cares. 0:09:43.240 --> 0:09:46.959 But eventually people started paying attention to That's a that's 0:09:47.000 --> 0:09:51.840 a fascinating story. It's very it would be very different now. Well, 0:09:52.160 --> 0:09:55.600 things seemed to ricochet around the world nearly much more rapidly, 0:09:55.880 --> 0:10:00.199 although I don't think people recognize the depth with a 0:10:00.280 --> 0:10:04.120 gravitas of certain ideas right away. But what are you 0:10:04.200 --> 0:10:08.440 thinking during that period where you know you've found something 0:10:08.559 --> 0:10:16.800 unique and valuable and nobody else's is recognizing that yet? Well, um, 0:10:16.960 --> 0:10:20.360 needless to say, you have some Maybe it wasn't I 0:10:20.440 --> 0:10:24.040 knew it was unique. Well, there's some dispute about that too. 0:10:24.400 --> 0:10:29.000 Others were going down similar paths and in various ways. 0:10:29.080 --> 0:10:32.040 But but I thought it was valuable, and I was, 0:10:32.240 --> 0:10:35.560 and I thought, well, if this isn't valuable, let me 0:10:35.600 --> 0:10:40.559 give you a little broader context. In those days, economics 0:10:40.920 --> 0:10:44.720 was theory and equilibrium and all that sort of stuff. 0:10:45.200 --> 0:10:50.960 Finance was very old timey by any modern standards. And 0:10:51.040 --> 0:10:55.440 so I was one of the first economists that went 0:10:55.600 --> 0:10:59.720 into the field of finance Fred Weston before me, to 0:10:59.760 --> 0:11:03.040 try to bring some of the rigor of economics to finance. 0:11:03.800 --> 0:11:06.720 And so there was a kind of a cultural issue 0:11:07.400 --> 0:11:13.040 as well, that's interesting and so, but the journal Finance 0:11:13.120 --> 0:11:16.200 had a number of I would call let's call them 0:11:16.200 --> 0:11:20.240 scientific articles. It was not unusual, but it was. It 0:11:20.360 --> 0:11:24.120 was a change for the field of finance, not only 0:11:24.160 --> 0:11:27.280 in practice, but also in academics, which is, you know, 0:11:27.360 --> 0:11:30.960 there was no field call financial economics. Now there is. 0:11:31.880 --> 0:11:35.200 How how influence was Harry Markoitz and your work with 0:11:35.280 --> 0:11:38.960 him it rans on you eventually creating the capital asset 0:11:39.000 --> 0:11:44.440 pricing model. Well, obviously, if I hadn't read Harry's work, 0:11:44.480 --> 0:11:46.920 if I hadn't done the work in the first part 0:11:46.920 --> 0:11:51.120 of the dissertation, I wouldn't have moved to go to 0:11:51.160 --> 0:11:54.120 the stage of asking the question what if everybody did this, 0:11:54.800 --> 0:11:59.120 so in that sense crucial um it was interesting. Harry. 0:11:59.120 --> 0:12:02.160 Two or three years later he said, oh, I just 0:12:02.280 --> 0:12:04.160 re read your paper, and now I see you didn't 0:12:04.160 --> 0:12:07.880 assume that you actually derived it. So so you know, 0:12:08.080 --> 0:12:10.400 I will say that that part I think was pretty 0:12:10.480 --> 0:12:14.680 much my work. Fascinating, although, as I say others, we're 0:12:14.760 --> 0:12:18.120 beginning down a similar path. Jack Traynor, you know, he 0:12:18.160 --> 0:12:21.440 didn't publish, but he was going down and he came 0:12:21.480 --> 0:12:24.680 at it differently. Where was Jack Traynor? He was well, 0:12:24.679 --> 0:12:27.720 he wasn't at heart. He he was working I think 0:12:27.760 --> 0:12:30.679 as a student at Harvard. I believe he was at 0:12:30.760 --> 0:12:34.960 Arthur D. Little when he was doing that work. He 0:12:35.040 --> 0:12:39.200 came at it a different way. And I became aware 0:12:39.200 --> 0:12:41.960 of his work a year or two after I had 0:12:42.000 --> 0:12:46.360 submitted Mind for publication. So I put a footnote in saying, 0:12:46.480 --> 0:12:48.840 you know, here's this other work you should be aware of. 0:12:49.080 --> 0:12:52.960 So you become a professor emeritus and decided to spend 0:12:53.000 --> 0:12:57.480 some time consulting at William F. Sharp Associates. What sort 0:12:57.520 --> 0:13:00.560 of consulting work were you doing back? Actually was when 0:13:00.559 --> 0:13:03.280 I finally bit the bullet and became an emeritus professor. 0:13:04.240 --> 0:13:08.000 My wife and I started a research slash consulting firm 0:13:08.040 --> 0:13:12.840 in eighty six which went through an eight nine and 0:13:12.920 --> 0:13:14.880 I took leave. Then I went back and then I 0:13:14.880 --> 0:13:17.760 thought I need to do it full time. UM and 0:13:17.800 --> 0:13:23.600 so that lasted six years in different manifestations up. What 0:13:23.640 --> 0:13:27.439 we were trying to do is bring I hate to 0:13:27.480 --> 0:13:29.960 call it modern finance theory, I hate that term, but 0:13:30.320 --> 0:13:36.680 financial economics theory, you know, empirical work to bear on 0:13:36.720 --> 0:13:40.439 the problems faced by the manager of the General Motors 0:13:40.440 --> 0:13:45.000 Pension Fund, the manager of the Stanford University endowment, so 0:13:45.600 --> 0:13:49.840 professionals who were managing large pools of money in an 0:13:49.920 --> 0:13:54.720 institutional setting. And so the idea was bring to bear 0:13:54.800 --> 0:13:58.720 the research that existed, and do new research and bring 0:13:58.720 --> 0:14:01.920 in new things that could help those folks. So so 0:14:02.000 --> 0:14:05.800 that that was the target in terms of the problems. 0:14:06.720 --> 0:14:10.200 And so we set up this firm and again in 0:14:10.320 --> 0:14:16.240 various manifestations, and we worked with General Motors, pension Fund, endowment, etcetera. 0:14:16.640 --> 0:14:19.800 So a lot of really substantial institutions with a lot 0:14:19.840 --> 0:14:22.600 of money. What were the sort of problems that they 0:14:22.600 --> 0:14:29.920 were encountering in the real world that your theory helped resolve? UM. Well, first, uh, 0:14:30.680 --> 0:14:35.560 what were the risks, Where were the risks? What was 0:14:35.600 --> 0:14:38.880 their performance? How did it you know, was it good 0:14:38.960 --> 0:14:41.280 or a bad relative to the risks that we're taking. 0:14:41.480 --> 0:14:44.960 People really were not all that clued into risk adjusted 0:14:45.000 --> 0:14:48.240 performance or had at that point. It was better understood 0:14:48.480 --> 0:14:50.920 at one level, but not. For example, one of the 0:14:51.000 --> 0:14:53.840 things that came out of that was what's called returns 0:14:53.840 --> 0:14:57.440 based style analysis. How do you get your hand You've 0:14:57.520 --> 0:15:01.080 got this portfolio've got a hundred different money managers out there, 0:15:01.680 --> 0:15:04.280 how do you get your arms around it. Who's doing what? 0:15:04.840 --> 0:15:08.240 How does this piece fit in with that piece? Are 0:15:08.240 --> 0:15:12.320 you getting enough average returns out of this manager to 0:15:12.480 --> 0:15:16.040 justify being in the portfolio and being in at that 0:15:16.160 --> 0:15:19.920 level or less or more? Um. We take that for 0:15:20.000 --> 0:15:23.120 granted that you could today run a spreadsheet, crunch froom 0:15:23.200 --> 0:15:25.800 numbers and bang you can figure that out. This was 0:15:25.880 --> 0:15:30.400 not simple to do best so so for example, returns 0:15:30.400 --> 0:15:33.640 based anile analysis. And again the idea was to get 0:15:33.640 --> 0:15:38.360 your arms around the whole portfolio, time managed portfolio and 0:15:39.760 --> 0:15:42.640 evaluated as a whole. Try to figure out whether or 0:15:42.640 --> 0:15:45.160 not you've got the right pieces and you've got them 0:15:45.160 --> 0:15:48.600 in the right magnitudes and uh, you know, at the 0:15:48.680 --> 0:15:51.480 end of the day, are you adding value? So so, 0:15:51.560 --> 0:15:55.160 there are a number of problems and uh, we got 0:15:55.200 --> 0:16:00.480 to deal with the real world very sophisticated clients and um, 0:16:00.600 --> 0:16:03.480 try out some new ideas, developed some new ideas. It was, 0:16:03.520 --> 0:16:06.440 it was, it was pretty heavy. So you joined the 0:16:06.520 --> 0:16:10.120 Rand Corporation in nineteen fifty six and you meet Harry 0:16:10.200 --> 0:16:15.080 Markowitz who ultimately helps you with your dissertation. Tell us 0:16:15.080 --> 0:16:18.520 about what it was worth like working with the markt 0:16:18.520 --> 0:16:22.000 Witz on your PhD project. Well, let me if I 0:16:22.080 --> 0:16:24.320 may do a little more of the backstory. Make sure 0:16:24.720 --> 0:16:26.960 when I first went to RAND out, I came out 0:16:26.960 --> 0:16:30.360 of the service with a master's degree to RAND. And 0:16:30.360 --> 0:16:32.960 when I first went, Harry was not there at that point, 0:16:33.720 --> 0:16:37.960 um and um. So I was working on logistics issues, 0:16:39.160 --> 0:16:43.840 big models and computer programs and what have you. And 0:16:43.880 --> 0:16:47.680 I decided I wanted to teach. So I the path 0:16:47.720 --> 0:16:51.080 of least resistance was to take some education courses get 0:16:51.080 --> 0:16:55.440 a junior college credential. I took one education course at 0:16:55.520 --> 0:16:58.120 night and decided no, I'd rather get a PhD and 0:16:58.400 --> 0:17:00.840 teach at the university level. So I was able to 0:17:00.840 --> 0:17:02.440 get a pH d at u c l A while 0:17:02.440 --> 0:17:06.600 working full time supporting my family at rand because they 0:17:06.640 --> 0:17:10.000 were very generous in that regard. Uh I started a 0:17:10.080 --> 0:17:14.480 dissertation a totally different subject, transfer pricing, using a lot 0:17:14.520 --> 0:17:22.080 of operations research methodology, and when Jack Hurst Lifer, whose 0:17:22.119 --> 0:17:25.360 work I was building on, came to U c l A, 0:17:26.359 --> 0:17:28.840 my adviser said, once you go talk to Jack, and 0:17:28.880 --> 0:17:32.400 I did, and Jack read my half dissertation and said, 0:17:32.400 --> 0:17:35.199 I don't I don't think there's a dissertation here that 0:17:35.280 --> 0:17:37.600 has to be a little frustrating you. Half that was, 0:17:37.760 --> 0:17:41.040 but I've remember Jack's dead now, but telling him more 0:17:41.080 --> 0:17:43.879 than once that he did me a great favor because 0:17:43.880 --> 0:17:46.919 then I went to Fred Weston, financial economist at U 0:17:46.920 --> 0:17:48.480 c l A, and said, what am I gonna do? 0:17:49.040 --> 0:17:51.600 He said, well, you really liked this work Harry Marco, 0:17:51.640 --> 0:17:54.360 what's did? Harry has just come to rand On. Donce 0:17:54.359 --> 0:17:57.600 you go talk to Harry. So I did, and I 0:17:57.680 --> 0:18:00.240 worked out an arrangement between Fred and our i'm an 0:18:00.240 --> 0:18:03.040 Alchin and at U c l A, and Harry, who 0:18:03.080 --> 0:18:05.040 was not at U c l A, that I'd work 0:18:05.080 --> 0:18:07.720 with Harry. So it was a little more. He wasn't 0:18:07.720 --> 0:18:11.200 a professor at US, but he effectively acts as your 0:18:11.240 --> 0:18:15.600 PhD advisor. Well, it was a little more collegial because 0:18:15.800 --> 0:18:19.359 you know, we're both working around together. Um, and he 0:18:19.400 --> 0:18:24.199 didn't have any authority, but but yes, basically he and 0:18:24.280 --> 0:18:26.800 I chatted about this and that and wanting to try 0:18:26.880 --> 0:18:30.080 that and and the rest and and so forth. So 0:18:30.520 --> 0:18:36.960 there are worse PhD thesis advisors than Harry markets every day. 0:18:36.960 --> 0:18:40.440 I'm thankful that that worked out. Let's let's talk a 0:18:40.480 --> 0:18:44.400 little bit about capham and and how how that has 0:18:44.440 --> 0:18:49.080 evolved over time. First, when you first introduced it, has 0:18:49.119 --> 0:18:53.760 your thinking evolved on that since or is it still 0:18:53.880 --> 0:18:56.639 what it was when you first thought it up? Now 0:18:56.720 --> 0:19:00.240 it has evolved and people are sometimes surprised of this. 0:19:00.880 --> 0:19:02.800 And let me if I may sort a little bit 0:19:02.800 --> 0:19:06.760 of time. The capitalistic pricing model builds on Harry's view 0:19:06.800 --> 0:19:10.679 of the world, which is that you think about the 0:19:10.720 --> 0:19:13.880 world in terms of mean and variance, expected return and 0:19:14.359 --> 0:19:19.240 risk variations mean variance. So here here's what we here's 0:19:19.240 --> 0:19:21.800 what's average, here's how much you can how much volatility 0:19:21.840 --> 0:19:24.880 around that is exactly what you should expect and and 0:19:25.160 --> 0:19:28.920 what you might not get despite your expectations. And then 0:19:28.960 --> 0:19:31.840 what I did in my equilibrium model is say, well, 0:19:32.160 --> 0:19:36.520 what if everybody thinks about the world that way, and 0:19:36.680 --> 0:19:39.400 they come to market and they trade with each other 0:19:39.440 --> 0:19:42.840 and prices at just what would one expect to find, 0:19:43.280 --> 0:19:47.800 and not surprisingly you find that securities in that world 0:19:47.880 --> 0:19:51.719 would be priced so higher expected return goes with higher beta, 0:19:52.200 --> 0:19:55.639 which is a measure of how things move together, and 0:19:55.680 --> 0:20:03.359 it's related to this variance mean variance structure. And you know, 0:20:03.400 --> 0:20:07.240 the economical line is it's market risk that matters. That's 0:20:07.240 --> 0:20:10.320 what you get rewarded for. Other risks you don't get 0:20:10.359 --> 0:20:13.960 paid for. That's sort of the bottom line. About the 0:20:14.000 --> 0:20:18.679 time Harry was first working, can Arrow and Gerard Debreu 0:20:18.960 --> 0:20:23.679 independently developed what came to be known as state preference theory, 0:20:24.440 --> 0:20:30.240 which basically is a model of prices in an equilibrium framework. 0:20:31.119 --> 0:20:35.720 And the basic idea there, to just make it overly simplified, 0:20:36.359 --> 0:20:39.600 is that you know, how much would it cost you 0:20:39.680 --> 0:20:42.960 to buy a security that pays you a dollar three 0:20:43.040 --> 0:20:45.680 years from now, if at that point the market is up, 0:20:47.080 --> 0:20:49.720 you know, give or take what would that cost, and 0:20:50.440 --> 0:20:53.280 there's some number present value of that, and then you 0:20:53.359 --> 0:20:54.760 just think of the world. There's a whole bunch of 0:20:54.760 --> 0:21:01.719 those and when you put that into a security market context, 0:21:02.400 --> 0:21:05.960 you again get the result that it's market risk that matters, 0:21:06.000 --> 0:21:09.080 but it may matter in a different way instead of 0:21:09.600 --> 0:21:12.680 in a certain kind of diagram. Instead of a straight line, 0:21:12.680 --> 0:21:16.480 it may be a curve. Just to widely oversimplify. And 0:21:16.600 --> 0:21:20.680 the great thing about that view is that it extends 0:21:20.800 --> 0:21:24.360 very beautifully to multi periods. It's it's much more general. 0:21:25.240 --> 0:21:29.399 And so that approach, which now in PhD programs and 0:21:29.440 --> 0:21:33.399 finance often is called pricing kernel k E R N 0:21:33.440 --> 0:21:37.600 E l UH. It has the same character, it has 0:21:37.680 --> 0:21:41.640 many of the same pragmatic results, but it's more general, 0:21:42.160 --> 0:21:44.280 and so that's what I use. I do not use 0:21:44.320 --> 0:21:47.920 the capitalized pricing model in my models in my work, 0:21:48.359 --> 0:21:52.320 which surprises some people. So you're you're using what is 0:21:52.600 --> 0:21:56.040 effectively the natural evolution of that. Well, they happen to come. 0:21:56.640 --> 0:22:00.720 They sort of the paths combined in some sense with 0:22:00.840 --> 0:22:03.160 Jack kerchlife for at U C l A, and Mark 0:22:03.200 --> 0:22:07.560 Rubinstein is his student there um. But so in some 0:22:07.640 --> 0:22:11.000 sense they're not sequential. But yes, I do, and I 0:22:11.040 --> 0:22:13.800 try to because I this is taught, as I say, 0:22:13.840 --> 0:22:16.080 at the PhD level, I think it ought to be 0:22:16.119 --> 0:22:19.439 taught at the NBA level and the undergraduate level. So 0:22:19.520 --> 0:22:22.040 in two thousand and seven I published a book the 0:22:22.119 --> 0:22:25.640 name of which I can't quite remember, that came out 0:22:25.640 --> 0:22:29.440 of some lectures I gave at Princeton, trying to make 0:22:29.520 --> 0:22:33.560 the case that yes, you can teach undergraduates and NBA 0:22:33.640 --> 0:22:40.720 students this approach rather than mean variance slash C A 0:22:40.800 --> 0:22:45.560 p M. Although again, as they say qualitatively, they're not 0:22:45.720 --> 0:22:49.040 wildly different. So the two thousand and seven book is 0:22:49.080 --> 0:22:54.560 called Investors and Markets Portfolio Choices, Asset Prices, and Investment Advice. 0:22:54.920 --> 0:22:58.679 Assuming Amazon is giving me the correct information. Amazon never failed, 0:22:59.480 --> 0:23:02.760 So so let's talk a little bit about that. Um, 0:23:03.119 --> 0:23:05.919 you know, we used to have memories and now we 0:23:06.000 --> 0:23:09.360 have these devices with us, and it's like I've outsourced. 0:23:09.680 --> 0:23:11.159 I knew the name of that book, and I couldn't, 0:23:11.160 --> 0:23:14.159 for the life of me recall it because it's just 0:23:14.200 --> 0:23:19.280 so easy to pull it up. Um paper, you you 0:23:19.359 --> 0:23:24.200 discussed earlier capital asset prices, the theory of market equilibrium 0:23:24.320 --> 0:23:27.560 under conditions of risk. I'm gonna pull a quote from 0:23:27.600 --> 0:23:32.879 that that I think is really powerful. Diversification enables the 0:23:32.960 --> 0:23:37.680 investor to escape all but the risk resulting from swings 0:23:37.680 --> 0:23:43.399 and economic activity. This type of risk remains even inefficient combinations. 0:23:44.240 --> 0:23:47.840 That's very powerful. How did you come upon that because 0:23:47.880 --> 0:23:51.760 it's not obvious or intuitive. Well, as I say that, 0:23:51.880 --> 0:23:56.680 you know that statement really needs another sence. Okay, as 0:23:56.800 --> 0:24:02.119 wonderful and I appreciate your finding it. The um the 0:24:02.200 --> 0:24:06.000 basic idea is that is the risk for which you're 0:24:06.000 --> 0:24:10.560 going to be rewarded if you expose you If you 0:24:10.640 --> 0:24:15.840 expose yourself more dramatically to down markets, then in a 0:24:15.920 --> 0:24:18.360 long run you should do better. In the short run, 0:24:18.400 --> 0:24:23.840 you can be wiped out at least widely injured. Um So, 0:24:23.840 --> 0:24:27.720 so that's the basic notion. And in the single index 0:24:27.800 --> 0:24:31.960 model that's assumed in the more general capitalizet pricing model, 0:24:33.000 --> 0:24:36.919 that's a conclusion. So so the assumptions, and you you 0:24:37.160 --> 0:24:41.440 discussed this earlier about how it was potentially problematic um 0:24:41.520 --> 0:24:44.120 for some of the people who were refereeing your your 0:24:44.160 --> 0:24:50.800 initial papers. Were the assumptions fairly straightforward in order to 0:24:50.840 --> 0:24:53.359 test the thesis or did you have to go out 0:24:53.400 --> 0:24:55.359 on a ledge a little bit with with some of 0:24:55.400 --> 0:25:01.200 your assumptions? Well, you know, I mean the the models, 0:25:01.240 --> 0:25:04.760 either the dissertation or the subsequent one. You know our models, 0:25:04.760 --> 0:25:07.399 You make some assumptions, and then you do some some 0:25:07.480 --> 0:25:11.880 calculations and some operations. You get a conclusion. I did, ah, 0:25:12.560 --> 0:25:15.359 a test that is so crude. I don't even we 0:25:15.400 --> 0:25:19.440 don't remember how crude our data sets were. I mean, 0:25:19.480 --> 0:25:22.480 I did a test in my dissertation. I used annual 0:25:22.560 --> 0:25:27.520 returns on thirty mutual funds. That was my database, and 0:25:27.560 --> 0:25:30.399 I had to put it on index cards and go 0:25:30.480 --> 0:25:33.040 to the library and write down all the numbers and 0:25:33.160 --> 0:25:37.200 use a hand, you know, a desk calculator. Um. So, 0:25:37.359 --> 0:25:40.520 over the years, of course, we did more sophisticated test 0:25:40.960 --> 0:25:46.320 had better databases. And that said, even today with all 0:25:46.359 --> 0:25:49.720 that we have, um it's hard. You know, there's a 0:25:49.720 --> 0:25:53.240