Busting 50 Years of Investing Myths | Rob Arnott

[00:00:00] This is going to sound very weird. 20 years is not a long horizon. Five years later, I was talking with a fellow at a conference and just said, how's life? And we were chatting and he said, by the way, I don't hate you anymore. The value side of the U.S. stock market is cheap relative to the market. Historically, just about as cheap as it can get. Small cap is cheap.

[00:00:26] So I see this as an opportunity-rich environment. Right now, the top 10 names in the S&P comprise just under 40% of the index. When has it ever been that concentrated? Never. Never in history. Even back in the 18th, 19th century.

[00:00:51] Welcome to Excess Returns, where we focus on what works over the long term in the markets. Join us as we talk about the strategies and tactics that can help you become a better long-term investor. Jack Forehand is a principal at Validia Capital Management. No information on this podcast should be construed as investment advice. Securities discussed in the podcast may be holdings of clients of Validia Capital. In this episode of Excess Returns, Jack and special guest host Perth Toll sit down with research affiliates founder Rob Arnott.

[00:01:16] Rob wrote an excellent article titled 50 Years of Innovation, Myth Making, and Myth-Busting for the 50th anniversary of the Journal of Portfolio Management. And they dig into the key myths Rob identified from his career in investing. They discussed the truth about long-term stock returns, why Smart Beta has lost its way, dividends versus VIVAX, and a lot more. Rob has had a front row seat throughout his career to the evolution of investing. And this discussion brings together many of the most important lessons he has learned. As always, thank you for listening. Please enjoy this discussion with research affiliates Rob Arnott. Rob, thank you for joining us.

[00:01:46] It's a privilege. Thank you for the invitation. I believe you're your fifth appearance now on Excess Returns, and I can only assume after five appearances or four previous appearances, you're incredibly bored. Well, I'm hoping that someday I'll actually get something right. We've decided, due to your boredom, we've decided to call in the reinforcements. For this interview. So instead of just dealing with Justin and I, we've asked our good friend Perth Toll, the founder of Life and Liberty Indexes, to help me with the questioning.

[00:02:14] Perth, thank you very much. Appreciate you joining us. Thanks for having me. It's my pleasure. So we're going to cover an article, a great article you wrote recently. I believe it was November. Is that right? It was. It came out in November. I think it didn't have a specific date on the publication. It was the 50th anniversary edition of Journal of Portfolio Management. So you wrote a great article called 50 Years of Innovation, Myth-Making, and Myth-Busting, as you mentioned, for their 50th anniversary.

[00:02:40] And it was a really good article, and it was great because it was all-encompassing of a lot of different things in finance. So you mentioned we were talking before that we often in finance write these specific articles about specific topics. But you were able to cover a lot, and you were able to cover a lot about what you've learned in your career. So we're excited to dive into it today. I'm thrilled. I want to first take a step back and talk to you about why you decided to write the article you did, because the focus on myths and in myth-busting.

[00:03:07] I assume you probably could have written a variety of articles for them for the 50th anniversary. Why did you decide to go this direction? You know, there's a saying, I forget who first said it, that science advances one obituary at a time, which basically means that there's resistance to new ideas. And people get locked into a worldview.

[00:03:32] And the disruptions that take place in science broadly, and certainly also the social sciences like economics and finance, these disruptions come in out of left field and come in in a context of busting myths.

[00:03:55] And so sometimes an idea comes along that is actually creating a new myth. So it has to get broken a second time. But that was the basis for choosing that as the title and the theme. The science of investing, the science of finance, more broadly, the science of economics, is a little dubious to call it a science.

[00:04:25] Because it's about human behavior, and human behavior is very difficult to quantify and predict. And so that was the basis for the approach. I did like that you acknowledged at the beginning that we at least have made some progress. You said during the first century of the Journal of Portfolio Management, finance theory has evolved from a world of heuristics, rules of thumb, and narrative into something resembling a serious science. So at least we're resembling now. We've made some movements in the right direction. Haley. Yes.

[00:04:55] Yeah. Whether it's a pseudoscience or a soft science is up for endless debate. But one thing it is not is as precise as physics. So you outlined three approaches to quantitative finance at the beginning of the article. You talked about the idea of a data-first approach, a theory-first approach, and Bayesian methods. Can you just talk about what those are? Sure. Well, data-first is very simple.

[00:05:21] You just mine the data, look for patterns, and say, okay, that's how the world works. That's the way much of academic finance and much quantitative finance in the practitioner community is done. Basically, in the factor world. Basically, in the factor world. Cam Harvey did a study of 400 factors that had been identified over the last 30 years in just the top three finance journals.

[00:05:50] 400 factors. 400 factors.

[00:06:20] So you're going to throw in a hurdle of it's got to have a T statistic of at least three. And now you're down to a couple dozen out of 400. So basically, a lot of what goes on masquerading as serious study is really just data mining. Data-first means you look at the data and you find a pattern that seems to be statistically significant and you publish it or you start running money on it. And think of the incentives in academia.

[00:06:49] You want tenure, so you want to publish. Now, are you going to look for the flaws in your thesis and blow your own idea out of the water if you have an alternative of not looking for the flaws and just going ahead and publishing and having a shot at tenure? Of course, you're going to be reluctant to look for the flaws. But science advances by finding flaws.

[00:07:14] Scientific method involves, there's a term in science, falsifiability. And it's a damning thing to say to somebody that their idea is unfalsifiable because that means that you can't prove it wrong. Efficient market hypothesis is unfalsifiable. All right. That means that it's an inherently flawed idea.

[00:07:40] Theory-first simply says, here's my theory of how things work. And I'm going to believe it regardless of what the data says. Both of those ideas are flawed. Bayesian methods are quite simply start with a hypothesis of the way the world works and then test it. But don't test it again and again and again and again to try to tweak it and make it better.

[00:08:08] Because if you do that, you're just going to wind up migrating towards whatever has the best past returns, which might not be useful in the future. And the amount of data matters. If you have thousands of samples of data, data-first is useless. If you have billions of samples of data, data-first can actually work pretty damn well.

[00:08:34] I want to ask you about this idea of data-first because it's something that's, I think, rising a little bit in our field. I mean, it's always been something people have done, but it's something I'm seeing more and more. And you've been referencing the paper, like, you can use that in an extremely wrong way and you can just data-mine every factor. And I think there was a paper you referenced where they came up with common shares outstanding minus common shares used to calculate EPS divided by rental commitments four years hence as the best-performing factor that exists. And obviously, that's not going to work out, but there have been some better works.

[00:09:03] Well, you don't know that. Maybe it will. It might, exactly. Trying to think about, like, launching an ETF based on that. You're probably not going to get too many investors who are too excited to invest based on that. But there have been some better works done recently. The paper we've talked about, and we've had them on the podcast by Andrew Chen and Alejandro Lopez Lira, they looked at a bunch of factors historically. They tested factors to 1990. They looked at ones that had behavioral explanations, ones that had risk-based explanations, and ones that had no explanations.

[00:09:31] And then they tested them all again out of sample. And what you would expect to happen happened, which is they all deteriorated to some extent out of sample. Right. But the ones that had no explanation did just a little bit better than the others. And so the question is, do factors really need to be intuitive? Can we use, like, a data-first approach? Or does it matter why these factors work? Well, I like the Chen Lopez-Lira paper.

[00:09:55] I think it's a good paper, but I think it overlooks one very important thing, and that is that factors can produce a good return by dint of the factor having merit or by dint of the factor getting popular and more expensive. So let's take low volatility.

[00:10:19] Low volatility was introduced in the 2000s, was becoming a little more popular in the early 2010s. And by 2016, money was pouring into low volatility strategies at the pace of billions of dollars a month. There was one Invesco ETF that had almost a billion a month pouring into just the single ETF. Well, that's pretty impressive.

[00:10:46] And it's based on the fact that low volatility was actually beating high volatility strategies for the prior 15 years. But what was missed was the fact that low volatility strategies, peak of the dot-com bubble, were trading at less than half the valuation multiples of high volatility strategies. By 2016, it had almost reversed. They were trading at almost twice the valuation multiples.

[00:11:16] So if you take low volatility strategies and have their valuations eclipse high volatility strategies, what's going to happen? The performance of this will look brilliant, but it's a non-recurring revaluation. So one of the things I focus a lot on, and I've been pounding this drum for well over a decade, is we need in the science of finance to ask the question, where did our performance come from? Instead of just saying, this works.

[00:11:47] Ask the question, did it work because it just got expensive and popular in crowded space? If so, that's a non-recurring alpha, and if there's any meme reversion, watch out. You're going to have great past returns lead to lousy future returns. So Chen and Lopez Lira did not include that test. And that's, I think, an important critique of an otherwise excellent paper.

[00:12:13] Yeah, so Rob, why do you think we're so resistant to new ideas in finance, and why do MIS continue to endure? Um, myths continue to endure because people are building careers on them. Uh, I wrote a paper in 2000 called Death of the Risk Premium, which basically said, if you look at tips, the yield is 4%. If you look at stocks, the yield is 1%.

[00:12:40] That means stocks have to enjoy 3% real growth in earnings and dividends, uh, just to match the performance of tips, let alone to deliver a risk premium. Uh, over the last century, the, the real growth in earnings and dividends has been 2%, not 3%. And so I described the risk premium as being dead.

[00:13:03] And, um, five years later, I was talking with a fellow at a conference and just said, how's life? And we were chatting and, and he said, by the way, I don't hate you anymore. And I said, that's interesting. Why did you hate me? And he said, because you wrote that paper, Death of the Risk Premium. I'd built my career on counting on a risk premium.

[00:13:31] And I hated that paper. And be partly because it was telling me that everything I built my career on is wrong. And I said, so why do you not hate me now? And he said, cause you were right. So, so anyway, um, people are resistant to ideas, new ideas that are unfamiliar because you have to, uh, get off your butt and do some work to learn about it.

[00:14:01] And, um, because they often have built a career on those ideas. And if the ideas are wrong, you're basically saying, um, you built your career on a foundation of sand. And, uh, so part of it is laziness. And part of it is, uh, deep commitment, commitment of your life to a set of ideas. And so people are reluctant to, uh, toss out those ideas.

[00:14:28] Uh, Albert Einstein late in his career was, um, asked about his contribution to the world of physics. Uh, and he said, I have no particular advantage over other physicists other than a deeper curiosity, um, than most. And, uh, I love that quote because curiosity is the way things advance. Uh, that said curiosity killed the cat, so to speak.

[00:14:58] And so it can be a dangerous path. If you're curious, innovative, and your innovations are wrong, uh, or don't work, um, then you can be doing damage to yourself. It's, it's a lot easier to follow the conventional path of just coasting with what you already believe. I love this perspective on curiosity, uh, because curiosity is something that every firm claims to have.

[00:15:23] Everybody claims they have, but yet is this openness to new ideas that, that you're connecting with it. So I love that. So we wanted to shift and discuss fundamental indexing and smart beta, which is, was, was a new idea at the time. And I remember some people getting amazingly angry about it. Yeah. Yeah. Uh, but before we do that, can you define the idea of fundamental indexing? Sure.

[00:15:47] Um, well, I'd long wondered why everyone anchors on cap weighting. Now, of course the market is cap weighted. And of course, if you want to just own the market or match the market, you're going to cap weight. So that's a given. Um, but why would you want to do that? If, if, if a stock doubles in price, um, and it's fundamentals haven't moved, your exposure to that stock will double as a share of your portfolio.

[00:16:17] Meaning that you're going to have maximum exposure when the price is maximally wrong. Now, indexers have heard that criticism since the dawn of, uh, uh, indexes back in 57, when the S and P 500 was created. Um, and they're, they have a very simple retort to that. Unless you can tell me which stocks are overpriced, you, you haven't said anything useful.

[00:16:45] What you're saying is absolutely true, but it's not useful. And, uh, I'd always thought that weighting companies, according to their economic footprint, how big the business is, would make more sense. Um, uh, a friend of mine who was, uh, president of the common fund and on the board of, um, uh, New York.

[00:17:07] And I think Virginia state pensions, uh, came to me and said, people are being pulled into massive exposure to. Dot-com companies. This was 2002. Uh, and they're getting crushed when the bubble bursts. Um, there's gotta be a better way to index. And, uh, Marty Leibovitz was also part of that conversation. And, um, um, that's when I decided to push fundamental index to front burner up from a research perspective.

[00:17:36] I, I'd long thought that weighting companies, according to their sales, for example, would be just as legitimate as weighting on market cap, but wouldn't have the problem of putting more money into stocks that are overpriced. Some stocks are overpriced. Some stocks are underpriced. If you tie the weight to the price, you're guaranteeing that most of your money is in overpriced stocks. If you weight the stocks by their sales, some overpriced stocks will be overweight. Some will be underweight.

[00:18:06] The errors cancel. And so, um, we tested that idea and we found that sales weighting or book value weighting would have added two and a half percent a year over the prior 30 years. Um, and so we then tested a host of other ideas, number of employees, uh, uh, dividends, uh, free cash flow. Everything worked. Everything worked about the same.

[00:18:34] We didn't go back and data mine and say, which works better? And can we tweak it to make it better? Because even 20 years ago, I was amply aware that data mining was dangerous. So we came up with a very simple process where you find out what percentage of the economy a company is by its sales relative to all publicly traded companies.

[00:19:00] Um, how big is it relative to dividends or now we prefer to use dividends plus buybacks relative to all dividends and buybacks, um, that are tacitly distributed to the shareholders. Um, and like I said, they all were. And by using multiple measures, you smooth off some of the rough edges. Uh, Walmart has huge sales, but it's profit margins are skinny, that sort of thing.

[00:19:31] And, uh, so you wind up with more consistency by using multiple measures. So fundamental index was just a matter of saying, Hey, this is, we have indexes that studiously mirror the look and composition of the stock market. And, uh, none that studiously mirror the look and composition of the macro economy. Now the controversy, some of the controversy, um, I, I spoke at the super bowl of indexing.

[00:20:00] It was called back in 2000. Remember those days. Yeah. That's a long time ago. Yeah. What was that? Yeah, I'm getting old. And, uh, the, I presented the idea and showed the historical results.

[00:20:18] And, um, one fellow who was the, uh, head of index fund management for a very large bank stood up after the presentation and he was literally trembling with rage and he could barely control his voice. Uh, he said, how dare you call this? He said, how dare you call this? And I said, uh, I'm sorry. And I said, uh, I'm sorry, my Merriam Webster doesn't say cap weight in its definition of index.

[00:20:49] Um, and I said, um, and I said, from your perspective, this is an active value tilted strategy. We're shifting our weight. We're buying and selling. If a stock soars and its fundamentals don't, we're going to say thanks for the gain. We're trimming it. If the stock tanks and its fundamentals don't, we're going to say thanks for the discount and top it up. And if the market's right with this stock being worth more because its growth prospects have improved, then wonderful.

[00:21:17] But it's already in the price. So it won't help you with a cap weighted portfolio unless it exceeds lofty expectations. The company that you just bought that had tanked the market's predicting tacitly predicting that the fundamentals are going to erode and maybe they do, but it's not going to hurt you unless it underperforms leak expectations. So you get a rebalancing alpha.

[00:21:40] So the, the cool thing about fundamental index is it has a rebalancing alpha that is very robust and very reliable. A lot of people think it's just a warmed over repackaging of value investing because it has a stark value tilt. Growth stocks are re-weighted down. Value stocks are re-weighted up. So you have this stark value tilt always. Now, um, that's all well and good, but, uh, uh, what you wind up with is the rebalancing alpha from contra-triving.

[00:22:10] And you're trading against the market's constantly changing opinions. You also get a value factor exposure that ironically is almost exactly the same as that of the value indexes. And so one of the things that's fun about Rafi is that if you compare it to value, you find that the outperformance is relentless. Um, FTSE Rafi was the granddaddy of, uh, all Rafi indices launched in 2005. So it's live experience 2006 to 2024.

[00:22:40] Um, it has beat Russell value and this is live. It's beat Russell value in 16 out of 19 years. Wow. By two and a half percent per annum with 2% tracking error and a worst ever drawdown of a little over 2%. So it does have a value tilt.

[00:23:04] And when value works against you as it has during much of the last 15 years, we have a headwind, but we do relentlessly beat the value indexes, whether they're winning or losing. Yeah. And you talked about in here as well, the term smart beta, how it came about. Um, and do you think that the term smart beta has outlived his welcome? Oh, smart beta has been applied to so many ideas, some smart and many of them, uh, not smart and some of them downright stupid.

[00:23:32] So the idea has been attached to so many ideas that the term is meaningless. Um, towers Watson, the consultancy out of, uh, London coined the expression back in 2007 and they did it based on fundamental index. They wanted their clients to use strategies that break the link with price.

[00:23:55] Co-equal with cap weighted indexing is basically half of their core, but they didn't want to say, go buy fundamental index because that would be appearing to endorse a single strategy. So they looked around and they said, there's a lot of strategies that have this same alpha. The alpha is not from the fundamentals. The alpha is from rebalancing and contra trading against noise in the market price.

[00:24:23] And well, gosh, the same thing applies for minimum variance strategies, unless they're anchoring on cap weights. Uh, the same thing applies to equal weighting. The same thing applies to a company called Tobam had, uh, uh, diversity weighting. Um, I believe they called it and maximum diversification. Yeah. Maximum diversification. Yeah. Um, and they said, these are all good ideas and they all have the same alpha engine.

[00:24:53] And so let's, let's call this smart beta. Smart beta is any strategy that's mechanistic, uh, formulaic, uh, replicable, historically testable, but breaks the link with price. Pretty soon the term was attached to everything. Uh, uh, the folks at Russell, uh, said, um, Russell value and Russell growth are smart beta.

[00:25:21] Uh, on the original definition, they aren't, but who controls the definition, the broad market and the broad market loved the term smart beta and started attaching it to everything under the sun. Yeah. Um, so strategies like momentum where you're buying more because the price soared. Um, that's the antithesis of smart beta on the original definition that was attached to momentum strategy. So the term is, is utterly meaningless now.

[00:25:49] But it's fun while it lasted. Yes. It's fun when you actually create a whole category of a term and then the term becomes so big that is meaningless. So what, what are you doing next? What areas of research excite you the most going forward? You know, we're having a lot of fun doing work on, um, uh, multi asset and on indexes.

[00:26:13] And the index work, uh, you and I are both avidly interested in the index arena and, uh, uh, you created, um, uh, strategy. That simply chooses the most, uh, the freest of the emerging markets economies on the presumption that economic freedom and rule of law.

[00:26:38] Means you're probably going to have more of the return float, float of the shareholder and less of the return float of the kleptocrats. And, um, uh, very cool idea. Uh, and, uh, by the way, um, shout out, it's done phenomenally well. Um, I think inception to date, you're five or 6% per annum ahead of the emerging markets cap weighted indexes. Very cool.

[00:27:07] Uh, you really need to launch a fundamentally weighted variant of that strategy. But be that as it may, I'm a, I'm a huge fan and I, uh, uh, interests of full disclosure. I'm not only a seed investor, I'm also an investor in your company and, um, uh, proud of it. Um, so the index arena is just fascinating.

[00:27:34] Um, index funds, think about them for a minute. Uh, S and P index funds collectively own about 25% of the market cap of every stock in the S and P. Okay. And they're cap weighted. So it's the same 25% for everyone. Now a stock gets added to the index.

[00:27:56] Um, they have to buy 25% of the total market cap fast. They all want to do it on the day that it's added to the index. Typically the S and P announces pre announces a change. It says the change will happen at the close of trading on, uh, thus and such date.

[00:28:20] And if they want no tracking error, uh, the industry has done a good job of educating their clients to believe that tracking error is God. And if there's no tracking error, you're competent. And if there's tracking error, well, it could go either way. So you better not try to beat the index with your trading, even though, you know, that between the announcement date and the effective date, the stock's going to have to go up. Why does it have to go up?

[00:28:48] Because they're buying 25% of the market cap all at once from active managers who love the company because it's gone up a lot. It's performed brilliantly. They don't want to sell it. And so they have to be induced to sell it by with a higher price. A beautiful example was, uh, Tesla when Tesla, uh, uh, S and P has a rule that they, uh,

[00:29:17] almost never violate, which is that a company has to have at least four consecutive profitable quarters to be added to the index. So Tesla was getting bigger and bigger and bigger without being added to the index because it was running operating losses. Okay. Well, finally, March of 2020, um, they'd had just had their third consecutive profitable quarter. So people were saying, aha, this is going to be added maybe next quarter.

[00:29:47] The next quarter was profitable and they didn't add it. But meanwhile, after the three quarters, people started talking about it's going to be added and it's huge. By that time, it was already one of the 20 largest market cap companies in the world. Um, it rose by well over a hundred percent before they announced that it was going to be added

[00:30:12] in late November of 2020 with an effective date of December 18. And they were going to add it all at once because they've always done that. And they did have discussions about. Staggling it to allow the massive purchase to not be so disruptive, but the stock went up 47% between the announcement date and the effective date.

[00:30:36] People could have bought it in March for a third, the price at which it entered in anticipation. People could have bought it on the announcement day, the day after the announcement date for about 30% less than they paid, uh, three weeks later. Um, but for the most part, they don't about 20% of the market cap of Tesla changed hands, uh, in

[00:31:01] what's called a market on close trade, which means I want to trade this stock. I don't care what I'm going to pay for it. That's what a market order means. Wow. I don't care what I'm going to pay for it, but I want it to be at the close because that's the price at which it goes into the index. So a market on close trade, 20% of the market cap of Tesla changed hands in a single humongous block trade. Um, all right.

[00:31:27] So index funds do a great job of matching the market and do a terrible job of trading. I wrote a paper in 2018, buy high and sell low with index funds that pointed out that stocks get added after they've soared enough to loft into the top 500 by market cap. And, uh, those stocks are pricey.

[00:31:54] They're priced at a premium multiple, typically about twice the market multiple. Um, they are pushed up in the process of getting added. And, uh, Tesla was actually underperforming, uh, the S and P in the subsequent, uh, four years until, um, uh, just after, uh, Musk endorsed Trump.

[00:32:20] And suddenly it surged as people thought, okay, Trump might win. This could be really good for the company. Anyway, um, so it was behind the S and P for, for some years. And even now it goes only modestly ahead. Um, so that's what happens with additions. Now these replace stocks that are kicked out. Some of them are kicked out because they're a takeover or, um, um, or a bankruptcy or something like that.

[00:32:50] But if it's a corporate, if it's a, if it's a, uh, discretionary deletion, then the indexes have to sell 25% of the market cap of a, what's now a newly small cap stock. That's illiquid. And that they want to sell it at the close market on close trade to sell it, which means I don't care what the price is.

[00:33:20] I want to get out at the closing price on this day. And the result is between announcement date and, uh, effective date, the additions beat the deletions by roughly 15%. Index funds say we don't move markets. Um, bullshit. Um, then the aftermath is interesting.

[00:33:45] The additions, uh, are now part of the index and anyone who buys the index winds up buying a little bit of that stock. So as money cores into indexes, it helps to prop them up, but the additions tend to underperform the market by one to 2% on average over the next year. Eventually they become market performers again. Um, the deletions tend to be hammered down far enough.

[00:34:14] That they snap back by an average of 20% over the next year. So one of the things that's really fun is work that we've done on deletions. Uh, if you go back historically. And by stocks deleted by the S and P 500 or by the Russell 1000, um, and hold them for five years on average, they, they outperform.

[00:34:40] Uh, I said 20% that's averaged across samples and it's heavily clustered in high turnover years like 2000 and 2009, but averaged across time. They win by 7% per annum in the first couple of years, trending down to 5% and then there's 3%. After five years, it's, it's a fairly anemic alpha. But over that five years, the deletions from the S and P and Russell win beat the market by an average

[00:35:09] of 28% over five years. So if you go back historically, what you find is that if you bought the deletions from the S and P 500 and the Russell index, uh, and just hold them for five years and then rotate out into the new batch, um, you would have made, um, I believe the number was 73 times your money between 1991 and 2023.

[00:35:35] Um, 32 years, um, 73 times your money, not bad. How do you, how would you have done with the S and P? Uh, you would have done awfully well. You would have added between 25 and 30 times your money. But last I checked 73 times, your money is a little archer than 20, than 28 times your money. So, so this is an exciting idea.

[00:35:57] And one of the things that's fun and Perth, you were instrumental in this is that the same folks who decided to launch, uh, an, uh, ETF based on your index decided to launch an ETF based on ours. And so, um, uh, that's a fun bit of work.

[00:36:18] Another area, and this is, um, kind of, uh, um, tipping our hand as to stuff that's coming in the not too distant future. It also turns out that if you just change your paradigm for how you add or drop stocks, if you want a cap weighted index, because you want to match the market.

[00:36:41] What if instead of choosing stocks that have soared and lofted into that large market cap arena, you, you add stocks that have gotten big enough as a business to matter and replace stocks that have, uh, where the business has slowly eroded and it's no longer big enough to matter. Okay. You're, you're still going to be buying stocks that are priced at a premium because they're growing and dropping stocks that are priced at a discount.

[00:37:07] But instead of a four to one valuation ratio, it's more like a one and a half to one, four to one versus one and a half to one. It's not by massively expensive and sell massively cheap. It's going to be by companies that are growing and sell companies that are, uh, seem to have lost their way. Well, if you think about that, what is that? That's just a cap weighted index based on fundamental index stock selection.

[00:37:35] So wrap use Rafi to choose the stocks and use cap weighting to weight the stocks. What comes out of that? Um, testing it over the last 30 years, we find, uh, it adds about half a percent per year per annum with about 1% tracking error. Now, the question is, is it 1% tracking error where we're not tracking the cap weighted index? No, wait a minute, we're cap weighted or is it one and a half?

[00:38:02] Is it 1% tracking error where the conventional indexes are missing the mark by concentrating their purchases on frothy bubble stocks and their sales on deep value stocks? Yes. There's a beautiful example. Dillard's is a stock that, um, uh, the small department store chain, uh, they've been in the Russell 1000 four times in the last 30 years. They've been kicked out four times in the last 30 years.

[00:38:33] Each time they're kicked out, they seem to have a knack of regaining their footing and getting back into the index. So buy high, sell low, buy high, sell low, rinse, repeat. And, um, they were last dropped in 2017. Have they done since then? They're up 500%. Uh, they've more than doubled relative to the market.

[00:38:56] Uh, if my math is right, they will be added to the Russell index for a fifth time next year. Yeah. So, uh, that's a deletion, but it's also, it also highlights the advantage of, um, a better form of cap weighted indexing. Uh, Dillard's never got small enough as a business to fall out of the top thousand. It's just that. It's profits soared when things were good. They tanked when things were bad.

[00:39:26] The price soared, the price tanked, but it's always been big enough to be in the top thousand as a business. So, um, Rafi would have had it in nonstop through thick and thin instead of buying when it's expensive and selling when it's cheap. Um, and so if you take Rafi and then cap weighted, we call it RACWI research affiliates, cap weighted index.

[00:39:52] Um, if you cap weight a fundamental index, you have a portfolio that is cap weighted that spans the economy that spans the market. It just leaves out some frothy, bubbly stocks until they get big enough to matter. And it doesn't give up on the Dillard's of the world just because they're having a bad couple of years. And so the result is, uh, um, I think a better cap weighted index.

[00:40:21] Now the challenge is how do you monetize an idea that makes indexing work better when most indexes are almost free? So monetizing the idea is the big challenge, but, um. Personally, if the cap weight has a component in there. Right. And it is cap weighted.

[00:40:45] The, we launched the index live as an index, uh, back in September of 21. So we now have 15 quarters of live experience and measured against the S and P 500. It's added 92 basis points per annum with 50 basis points of tracking error. Who would care about 50 basis points of tracking error if you can pick up 90 basis points of return? And that's live. That's not a back test.

[00:41:13] I mean, so to answer the monetizing question, you need a scale. So if iShares or Vanguard is listening, um, give us a call. Well, yeah, I'm highly confident that they won't. Yeah. If they want to, I'm happy to chat. The other way to do it is a performance fee, which is tough in the ETF arena. Yeah. Uh, but, uh, if we make an extra basis point every time our customers make 20 basis points,

[00:41:42] I would think the customers and we can be very happy about that. I want to ask you, we've had Mike Green on the podcast and I believe you've spoken to Mike as well. Um, and he's done some really interesting, he's, he's really smart. Done some really interesting research about what the impact of all of this passive, you know, the rise of these market cap weighted indexes and the default options on 401ks and all these flows into passive funds have. And I'm just wondering what your thoughts are on that. I mean, if you, if you agree with his research, I mean, this is having, this is probably driving up the biggest stocks relative to the rest of the market.

[00:42:12] And also, you know, it could have implications for the market as a whole eventually, if this gets too far out of control. So what are your thoughts in general on that? Um, a lot of people have complained about indexing from the vantage point of, well, it's the investment world's equivalent of socialism. You're pre-walking off the market, making the decision of what things are worth and you're not paying for it. Okay.

[00:42:40] Well, um, that's on one level, that's absolutely right. On another level, it doesn't address the question of, well, why should people not freeload if they can? Um, the near zero pricing and the assured market performance is a very cool, compelling package, but is it making the market more efficient or less efficient?

[00:43:07] I think it's making the market much less efficient, but on a long cycle fashion where it's hard to exploit that, that mispricing. Right now, um, the top 10 names in, uh, the S and P comprise just under 40% of the index. When has it ever been that concentrated? Never, never in history.

[00:43:36] Even back in the 18th, 19th century, it never got this concentrated. And so, um, you have huge concentration. Is it matched by concentration in the economy? If you look at Rafi, the top 10 names comprise a little over 20%, meaning that the top 10 businesses, publicly traded businesses are a little over 20% of the business economic footprint of all publicly traded companies.

[00:44:03] They're big, big, big businesses, but that's a mismatch 40% versus 20%. That means that these companies are priced to reflect an expectation that they will be 40% of the future U S economy in the decades ahead. That's a pretty extravagant claim. So fundamental index simply says, I'm going to reweight it back down to the 20%. I'm not going to throw them out the way a value index does. I'm going to reweight them down.

[00:44:34] But back to the question of, does this make the market more or less efficient? I think it makes it less efficient. Now, uh, green has done some really interesting work on, uh, how it breaks. Um, and the way it breaks is kind of interesting. Uh, people have asked me for a long time, whether I think what I think the limit is for indexation before it does economic damage.

[00:45:03] Um, and my answer is it can be very, very large, um, 50% sure. 70% maybe, but think about it this way. Suppose indexes comprise 70% of the market. Um, there's members of the index. There's non-members. You're going to drive a wedge in valuation between the members and the non-members that has nothing to do with the underlying fundamentals. That's an opportunity.

[00:45:32] Higher price means lower future return. You're front end loading that future return. So by the companies that aren't in the index and you should have much better long-term performance is once the flow of money into these stocks slows, the flow of money into these stocks props their relative valuation up. Now think about hypothetical world in which it's 70% of the market. Okay.

[00:46:02] Stock gets added. Indexes, 70% of the market have to buy 70% of the market cap of a stock that they don't own. They buy it from active managers who are now less than half their size. What kind of price appreciation will induce that little sliver of active managers to sell to these whales?

[00:46:25] Um, the, the, the price impact of a stock being added or dropped will increase drastically. It can even reach a point where there's no market clearing price. And that's where things get very interesting and very fragile. Um, suppose, um, a stock gets dropped. It's indexes have to sell 70% of its total float.

[00:46:54] To active managers who don't want it because it's been a lousy company. Uh, it's been performing badly enough to get kicked out of the index. So they have to get a very low price and you can actually paint scenarios in which there aren't buyers sufficient. And the price, the, the clearing price can in theory actually be zero. So would you want to buy it at zero? Yeah, of course.

[00:47:20] Uh, but, but that's, we're pretty far away from that situation. But as indexing grows, uh, it, there should be some serious conversations about, um, uh, where's the breaking point where the market clearing price doesn't exist. The back half of the conversation here, I wanted to go through some of these other myths you had in the article because you, you had some really great ones.

[00:47:47] Um, so Perth and I are just gonna go back and forth here and we're, we're gonna list some of these for you. And we want to have you talk about your opinion on these. And the first myth you highlighted in the article is historical excess return can help determine the equity risk premium. Um, people look at past returns and say, okay, over the period of time covered by the Ibbotson data, going back to 1926, you're now fast approaching a century of data.

[00:48:14] And over that period of time, stocks have beat, uh, bonds by, um, on the order of foreign, four and a half, four and three quarter percent per annum compounded for a century. Um, have beat T bills by over 5% per annum for a century. So, well, obviously we should assume that stocks will beat bonds by 5% because that's been a

[00:48:39] long-term history over a span of time, uh, rather longer than any of us should reasonably expect to live. And, um, so the reliance on past returns is very deceptive because this gets back to revaluation alpha revaluation alpha exists within stocks. I talked about it for low volatility stocks versus high volatility stocks. It can also be for stocks relative to bonds.

[00:49:07] If you take second half of the 20th century, uh, most people would consider 50 years a reasonably long span of time. And long-term as an investor, stocks beat bonds by 8% per annum for 50 years. Uh, this helped to fuel the view in the late nineties and early two thousands. Why would I put any of my money anywhere, but stocks, stocks for the long run?

[00:49:34] Um, the interesting thing is that, um, the relative valuation of stocks, stocks were priced at an 8% dividend yield at the start of 1950, a hundred dollars invested. You, uh, you get, um, uh, eight, eight dollars per year in income. And this was at a time when there was basically no inflation.

[00:50:00] Um, bonds were priced to give you about a 2% yield. Now, uh, roll the clock for it. End of 1999, you, you were down to a 1% yield. That means the rise in the value of every dollar of dividends went up eightfold. Do the math. That's 4% per annum.

[00:50:25] So of the 8% return difference between stocks and bonds, half of it came from revaluation. So maybe we should take that out and just assume a 4% spread, but wait a minute. Now you're at the end of a century and the yield on stocks is 1%. The yield on bonds is 6%. The yields on inflation linked bonds is 4%. This is when I wrote that paper death of the risk premium.

[00:50:55] So maybe you shouldn't assume 4%. Maybe you should look at current yields and ask the question, what's the relative opportunity? Well, stocks historically have earnings and dividends that grow about 2% faster than inflation. Tips have income that grows per a passu with inflation contractually linked to CPI inflation. So you take that 1% yield, you add 2% for grow.

[00:51:23] That's the long-term historical growth rate. That's a 3% real return. Tips give you a 4% real return. Your risk premium at the end of that 50 years would probably minus 1%. Now what's happened since then? Stocks are back to a valuation level not dissimilar to the valuations of 99 and 2000. Shiller P ratio peaked recently at 38 or 39 times.

[00:51:51] It was 44% at the top of the dot-com bubble. So when you look at the world today, you have similar problems with the likely return for stocks. You've got a yield of 1%, 1.25%. You've got historical long-term growth, 2% above inflation. That means a 3.25% real return from today.

[00:52:20] You can get 2.25% from 10-year tips. So that means today the risk premium is only about 1%. And that's if current valuation levels hold. And 10 years from now, we're at the same level. Do you want to own stocks with a 1% premium over what you can get with government-guaranteed inflation-linked bonds? I don't find that very interesting. It's heavily skewed by the Magnificent Seven.

[00:52:48] The value side of the U.S. stock market is cheap relative to the market. Historically, just about as cheap as it can get. Small cap is cheap. So I see this as an opportunity-rich environment where stocks deleted from an index or small cap stocks or value stocks or non-U.S. stocks are all priced pretty reasonably.

[00:53:15] So I think there's a lot of places to invest, but U.S. growth stocks aren't on that list for me now. Not at these prices. Yeah. So another myth that you covered was that long horizon forecasts are hard. A lot of people think forecasting the future is difficult and gets harder and harder the further out you look.

[00:53:41] When it comes to things like geopolitics, I could reasonably forecast a relatively benign next four weeks. Pretty high odds of that. I have no clue what 10 years from now is going to look like. I have even less clue what 20 years from now is going to look like. So long horizon forecasts are generally perceived as difficult. For investments, it's the opposite.

[00:54:10] Anything you invest in produces return in the form of income plus real growth in income plus or minus changes in valuation levels or spreads. And the changes in valuation levels are the tail that wags the dog on a short-term basis.

[00:54:32] So I can look at U.S. stocks and say, I think I'm going to get one and a quarter from the yield. And I think the income on a stock portfolio is going to grow at CPI plus a little bit. All right. If I think CPI is going to be two and a half, let's call it 5% from growth and one and a quarter from the yield. So 6% return.

[00:55:01] Plus or minus 20%. That plus or minus is huge. 10 years, you're looking at the same math. Something on the order of 6% returns. 3% real. I already described earlier. On the order of 6% returns. Plus or minus 2% or 3%. And so long-term returns are much easier to forecast.

[00:55:31] 1950, using that same logic, you would have said, I'm going to get 8% from income. I'm going to get 1% from inflation. I'm going to get 2% from real growth and earnings and dividends. That's an 11% return. And that's if the valuation levels stay the same. If the yield comes down by half, then that 11% return becomes 20 over the next 10 years. Well, that's what the 1950s was.

[00:56:00] That's exactly what the 1950s looked like. Looked at from today's vantage point, if valuation levels stay right where they are, 6% is a reasonable return expectation over the next 10 years. But if valuations return towards historic norms, you'd better shave those expectations. So right now, we're looking at a likelihood that valuation multiples mean revert downward. And you still get your yield. You still get your growth.

[00:56:29] But valuation levels mean revert downward. And we have a website, asset allocation interactive. If you Google asset allocation interactive and skip past the sponsored ads, the first link you'll get will take you to that website. And it shows forecast returns, 10-year returns for 130 different asset classes. And for U.S. stocks, it's showing about a 3% return. Well, you can get almost five from 10-year government bonds.

[00:56:59] So this looks to us to be an inauspicious time to load up on equity market risk. And congratulations, by the way. That tool, I believe, just crossed its 10th anniversary. Is that right? It sure did. And we've gone back and looked at how it's done. And the rank order of asset classes and the 10-year subsequent returns are highly correlated. The outliers are all due to revaluation alpha.

[00:57:27] The outliers are U.S. stocks performed better than we would have predicted and emerging market stocks and bonds worse. And those outliers are because U.S. stocks revalued upward from a Shiller PE ratio 10 years ago of about 23 or 24 to 50, 60% higher. And emerging markets the other way.

[00:57:53] So the places we got it right were most.

[00:57:59] The places we got it wrong were strictly due to revaluation where the market likes U.S. stocks far more than it did 10 years ago and likes emerging market stocks far less in large measure thanks to China and Russia, which, Perth, you've been dutifully avoiding in your index ever since the launch of the index. Yes, thank you. Thank you.

[00:58:30] I just want to get through a few more myths before we wrap up here because we are running a little short on time. But this one's interesting to me because you always see this stat brought out there that there's never been a 20-year period where you lost money in U.S. stocks. And that's an accurate stat, but it can be somewhat misleading as well. So your myth here was if your time horizon is 20 years, you can't lose with stocks. So what were you saying with that?

[00:58:53] If your time horizon is 20 years, it's hard to lose money with anything because almost everything produces a real return. But the relative opportunities can be shockingly unreliable.

[00:59:15] We went back over the last 200 years, actually 220 years, and looked at rolling 10-year returns for stocks relative to bonds. And for the first 100 years of that, the 19th century, stocks sometimes beat bonds by 10-15%, sometimes underperformed by 10-15%. They underperformed more often than they outperformed on a 10-year basis.

[00:59:46] The 20th century saw a change in that. And of course, the 21st century is just overwhelmingly up if you ignore that first lost decade. But the simple fact is stocks can underperform other asset classes very easily.

[01:00:04] The U.S. has been in a very auspicious circumstance of having no wars fought on our soil since the Civil War. And that's not true of most countries. So the U.S. has had particularly good luck on that score.

[01:00:27] But if your time horizon is 20 years, stocks historically produce positive returns in almost all cases, but superior returns to bonds most of the time. I'm a believer in a stock-centric portfolio most of the time, but not when the Shiller P.E. ratio is 38. And I'm still stock-centric, just not U.S. stocks.

[01:00:56] But this is a myth. I remember at the market low in 2009, I sent a note to Jeremy Siegel in which I said, have you noticed that long bonds have outperformed stocks on a 40-year basis now?

[01:01:19] Roll the clock back 40 years, and you were looking at February of 1969, which was 69 was a bear market year, but it hadn't really started by February. So stocks went from bear market high in 69 to a bear bull market high to a bear market low in 2009.

[01:01:42] And meanwhile, bond yields had gone from relatively high levels to relatively low levels, with the result that they produced capital gains on top of rich starting yields. And his initial response was, that's not possible. And then he went and checked and got back to me and said, yeah, it's true.

[01:02:04] So you can't lose with stocks on a 40-year basis unless your country's obliterated in the war. But you can lose relative to other asset classes, even over a span as long as 40 years. 20 years is, this is going to sound very weird, 20 years is not a long horizon. Yeah, that does sound weird.

[01:02:31] So another myth that you had in here is the equity risk premium is static. You know, I debated Roger Ibbotson about 15 years ago about equity risk premium. And Cliff Asmus and I had written a paper in which we showed that when companies retain most of their earnings, the subsequent dividend growth is better.

[01:03:01] Yeah. And the subsequent returns are better. And the subsequent earnings growth is better. And that's so counterintuitive. It actually goes against Modigliani-Miller, which is one of the foundations of modern finance. But in point of fact, what you do with your profits matter.

[01:03:29] If you reinvest them in your own business, you have a limited roster of choices of where to put the money. If you distribute it to your shareholders, they have an unlimited roster of choices for where to deploy the money. So anyway, Ibbotson's comment was, well, when the dividend yield is low, what the market is telling you is that growth is going to be superior and that the growth will make up for the low yield. Why?

[01:03:59] Why should that be? Why should that be?

[01:04:31] That's not true right now. Then the only way to make it true is to have growth go much faster than it was in the past. And to make the argument that low valuation multiples mean you're expecting lousy growth, high valuation multiples, you're expecting great growth. And it doesn't, doesn't turn out that way. So, um, this notion of anchoring on a static risk premium is, is dangerous.

[01:04:57] Uh, there was one big public fund, one of the 10 biggest in 1999, uh, the actuaries came to them and said, um, uh, here's a draft of your year end report. And you're slightly underfunded. And they said, we can't do that. We can't have a report that says we're underfunded. It's been a huge bull market. And we just can't say we're still underfunded after a bull market.

[01:05:28] Um, and, um, they said, well, the actuaries said, well, stocks beat bonds by 5% a year. So if you move 20% more into stocks, we can boost your return assumption by 1% and you will be fully funded.

[01:05:43] And so it's the actuarial tail wagging the investment dog, um, because of a, an assumption of a static risk premium, this pension fund moved 20% from bonds into stocks at the end of 99. I love when you and Asnaz agree on something. So. Oh, I have huge, huge respect for the guy. Um, he's, he's very smart. He's very capable.

[01:06:10] Um, occasionally he might have anger management issues. That's, uh, that's something. I think he would, he would admit about this. I don't think he would, I don't think he would disagree. And he was very complimentary of you, by the way, too. We just had him on. He's terrific. He's terrific. He does great work. Um, and I, I am very pleased that he's, um, uh, uh, uh, gone.

[01:06:39] Notably quiet on, uh, the flaws of Rafi. What, what, what? So I just want to do one more before we wrap up here. And I've learned that there's certain terms on Twitter that I have to avoid. Um, like anything with crypto, if you get involved with that, it's, it's a problem. Uh, if you challenge Warren Buffett in any way that that can be, that can be an issue as well. But this, this next one is one that you, I definitely steer clear of whenever I can, which is this whole buybacks versus dividends. Um, situation.

[01:07:05] So, uh, your, your myth here was that buybacks are a stealth dividend that rewards shareholders in a more tax efficient manner. So why do you think that's a myth? If you're a shareholder in a stock and the company decides to buy back some of the stock, um, if you take them up on that, you sell your stock to them. Now you don't own that stock. So you're not a shareholder anymore.

[01:07:31] Um, the, the notion is if a company pays 2% dividend and buys back 2% of its stock, that the effective yield is 4%. That's bunk. Um, the yield is two. The other 2% is a reallocation of the capital structure of the company.

[01:07:57] And the only way it will benefit the buy and hold investor is if it leads to faster growth in the future, in those future dividends. Otherwise it hasn't done anything. All it's done is change the ownership structure. So this is, um, uh, this is a very important myth. I do think buybacks are very useful.

[01:08:24] And I think companies that don't have better use for their capital than to buy back stock should do that or should pay it out as dividends. It is more tax efficient. That's absolutely true. It's tax efficient for those who are willing to sell their stock in order to avail themselves of the buyback. But if you're a buy and hold investor, you're not one of those. And the only way it'll benefit you is if it leads to faster future growth. So how does a buyback lead to faster future growth?

[01:08:52] If the, um, cost of capital, uh, if, if the growth of the business, um, uh, exceeds the cost of capital, a lot of buybacks are funded out of borrowing. You borrow money, you buy back stock. And, uh, it, it, it's just a shift from equity based capital to debt based capital.

[01:09:17] Uh, the only way that that magnifies growth is if the cost of that capital is less than the, than the benefits of investing that capital. So if, if, if a company buys back stock, uh, it must see faster future growth in order for it to reward the shareholders. And that's not assured.

[01:09:41] So when companies buy back stock, they're not rewarding the shareholders, uh, they may be rewarding management. Um, uh, I can come back to that in a second, but, uh, they're not rewarding shareholders unless the result is that the cost of capital is smaller than the, uh, reinvestment opportunities that are internal. Uh, one of the other things about buybacks is that they're often done at high prices.

[01:10:12] Why would buybacks accelerate when the price is high? Because management pays itself with stock options. And if you want to redeem a million shares of stock options, what better way to do that than to announce a million share buyback than redeem your stock options. Um, and the change in the flow for the company is zilch, but you created a buying opportunity where you're helping the share price when you're redeeming your stock.

[01:10:41] So, um, that's not, that's not a buyback. That's management compensation and drag. I have no problem with management having huge rewards if they are generating, uh, big profits for their shareholders. I'm, I'm a true capitalist on that score. If Elon Musk wants a $47 billion dividend and he's made his shareholders a trillion dollars, I'm fine with that.

[01:11:08] But, but if, um, uh, but if you use a buyback to push up, to hold up the price while redeeming and selling stock, um, that's just management compensation in drag. And it should be called management compensation out of buyback. Well, thank you, Rob, for joining us. This has been awesome. It's been really informative. I really appreciate it. Yeah. Great. Thank you, Rob. All right. This has been great fun.

[01:11:38] Thanks for reaching out. And, uh, I, I hope your, um, viewers enjoy this. And, and thank you, Perth. Uh, I assume I'm going to be getting an email from, uh, Matt and Justin shortly that I've been replaced. I'm the podcast host. So, uh, thank, thank you for helping me out with this. Thank you for having me. It's been my pleasure. Thank you. This is Justin again. Thanks so much for tuning into this episode of XS Returns. You can follow Jack on Twitter at practical quant and follow me on Twitter at JJ Carboneau.

[01:12:05] If you found this discussion interesting and valuable, please subscribe in either iTunes or on YouTube or leave a review or a comment. We appreciate it.