 Welcome everyone to the 50th edition of Bogleheads On Investing. Today our special guest is Craig Lazera. Craig is the managing director in the core product management group at S&P Dow Jones Indices where his responsibilities focus on providing thought leadership and educational outreach. Today we're going to be discussing the SPIVA report, S&P Index Versus Active. Hi everyone, my name is Rick Ferri and I'm the host of Bogleheads On Investing. This episode, as with all episodes, is brought to you by the John C. Bogle Center for Financial Literacy, a 501C3 nonprofit organization dedicated to helping people make better financial decisions. Visit our newly designed website at bogelcenter.net to find valuable information and to make a tax deductible contribution. And don't forget about our Bogleheads conference coming up this October 12th through the 14th featuring many speakers that I've heard on this podcast and more. Today our special guest is Craig Lazera. Craig is the managing director in the core product management group at S&P Dow Jones Indices where his responsibilities focus on providing thought leadership and educational outreach. Today Craig and I are discussing something near and dear to the Bogleheads heart. The active versus passive debate. Should you only be using index funds or should you be using only active funds or should you be using a mixture of active funds and index funds? There's over 100 years of data on active management versus index returns and this data has been remarkably consistent through the entire 100 year period of time. Today we're going to be looking at the last 20 years when S&P has been publishing their detailed index versus active report and a second report which analyzes the persistence of outperforming actively managed mutual funds. So with no further ado, let me introduce Craig Lazera. Welcome to the Bogleheads on Investing podcast Craig. Craig, thank you. I'm delighted to be here. Craig, we've known each other for a long time and you've been at S&P Dow Jones Indices for a long time. And I want to get into a little bit of your background as to what brought you there but in order to talk about S&P Dow Jones Indices I think we need a little explanation of who is S&P Dow Jones Indices and why is it S&P Dow Jones rather than just S&P? Well, as you suggest Rick, they're originally two different entities. S&P of course stands for Standard and Poor's and the roots of the Standard and Poor's company go back to the 1860s when Henry Varnum-Poor published the first railroad ratings. S&P obviously is well known for its ratings business. There was a company called Standard Statistics that I believe in 1923 began to publish an index of the U.S. stock market. I think it had a relatively small number of names and it was the first capitalization weighted index ever computed on a daily basis and I believe that was 1923. Somewhere along the line the Standard Statistics company and Henry Varnum-Poor's ratings company merged. That's where it came to be Standard and Poor's and then this initial cap weighted index product that I mentioned that started in 1923 morphed by, we added names as they added names as computing capabilities allowed. But in 1957, I believe in March 1957, the Standard and Poor's company launched what was called then the Standard and Poor's 500, now the S&P 500. And so the history of the 500 goes back to March 1957 and that obviously has come forward from there. On the other side of the merger, Mr. Dow and Mr. Jones started publishing their iconic indices. I think in the 1880s the Dow Jones Industrial Average started I believe in 1896 but the transportation average is actually a bit older than that. Those were price weighted meaning in the days when the only computational equipment you had was a piece of chalk and a blackboard, you could add up the names or the prices of, I think it was a dozen stocks to begin with and divide by 12 and got your answers. So the Dow company continued to evolve and obviously as indices and index funds became more important in the investing landscape, both companies developed substantial businesses and licensing indices for the creation of investment products. In 2012, what was then called S&P Indices acquired Dow Jones indexes and formed S&P Dow Jones Indices, which is technically now a joint venture company about three quarters owned by S&P Global, which is the parent company that evolved out of the old S&P and one quarter by the Chicago Mercantile Exchange. So we've been S&P Dow Jones Indices since 2012. It's interesting because both S&P had a total stock market index and Dow Jones had a total stock market index and I think the difference between the two indices, even though they both have 4,200 names, one of them might have 4,201 names and the other one might have 4,200 names. I bet it is so similar, but when you're looking at something like the Fidelity Total Stock Market Index or iShare Total Stock Market Index, one of them actually tracks the S&P Total Stock Market Index fund and the other one tracks the Dow Jones Total Stock Market Index fund, or the Total Stock Market Index, I should say, and they're so very similar. Yeah, I remember when we did the merger, I was product manager for our U.S. equity product, which basically means the S&P 500. And I remember looking at those very things, you graph the S&P Total Market Index and the Dow Total Market Index, and if you could put a speck between the graphs, you were lucky. They both include all the stocks you can grab hold of and they're cap weighted, so it's going to come out to the same thing if you do the numbers correctly. Yeah, you know, on this point, I get outside of Dow Jones S&P and let's go to CRISP, which is the indices that Vanguard uses for their Total Stock Market. It's a University of Chicago Center for Research and Security Prices, CRISP. Same thing. I mean, very, very similar tracking, almost all the same stocks, maybe entering the index, coming out of the index, slightly different, but negligible. And then even if you look at the Morningstar Broad Market Index and all of these companies that are creating indices for, say, the Total Stock Market or the Broad U.S. Market, the Broad Market would be a little bit of a smaller subcomponent than the Total Market, but they're all very similar. I mean, the correlations are 99%, and they might be a 0.1% difference in return, but they're all very close. That's the history of S&P Dow Jones indices. And what about your history? I mean, how did you end up there? Yeah, well, I went to college at Princeton, I majored in what we called them Public Affairs and Economics. Went from there to Harvard Business School and graduated more years ago than I care to admit to, but quite some time ago. And did a year with a consulting firm in Boston and then joined an investment management company and had been in the investment management business ever since. And you received your chartered financial analyst charter along the way? Oh, yeah. I received the charter in 1983, yes. I have a four-digit charter number, which, as you know, Rick, is pretty low. Oh, well, let me think. Do I? No, I don't. You're ahead of me. And then you went into work in the investment industry. I started my career in the investment business on what we'd call the buy side as an investment management companies. I'd spent a number of years at a variety of firms tending to specialize in what was then considered, you know, quantitative analysis. I think by today's standards it wasn't all that advanced, but by 1984 standards was pretty good. And so I was an equity quant for a while. And in the mid-90s I had the chance to join the old Solomon Brothers brokerage firm, kind of in a job that related to marketing and use of their quantitative research. Solomon had started in 1989 a set of indices that were designed to be float market cap weighted indices of the entire global stock market. I got to know that group. They were related. I eventually transferred into the department that managed and maintained those indices. And those indices were ultimately acquired by Standard and Poor's in 2002, I believe. Now I had left Solomon prior to the acquisition, but luckily I've met a number of people at S&P. And when they were looking for some senior help in 2008-2009, I was available and were able to make the match. So I ended up joining S&P in June of 2009. So now you're at S&P indices, which became S&P, and you and I have been talked for many years about the results of the SPIVA report that was started 20 years ago. So let's talk about what SPIVA stands for and why this report got started and why it's important. Easy question for SPIVA stands for S&P index versus active. It's an acronym. Sometimes in the UK pronounce SPIVA, so take your pick depending on where our listeners are. But the point of SPIVA then 20 years ago and now is to ask the question, how have actively managed funds performed relative to benchmarks that are appropriate for their investment styles? What I mean by that is if you're trying to evaluate how large-cap US managers have done, you compare them to the S&P 500. You want to ask how mid-cap managers have done. You might compare them to the S&P mid-cap 400 and compare growth managers to a growth index, value-to-value index, and so forth. But that was the notion that underlay SPIVA in 2002 when it was first published, and that is the notion that we continue with today. So when you talk about managers, you're talking specifically about mutual funds in exchange-traded funds, correct? In the case of US SPIVA, yes, it really mutual funds, actively managed mutual funds. The data for which we access from a database called the CRISP, you mentioned CRISP earlier, the Center for Research and Security Prices at the University of Chicago. CRISP maintains what is called a survivorship bias-free database of funders. So it's a mouthful to say, right, CRISP survivorship bias-free database. In the initial generations of SPIVA, it was focused on mutual funds. We've expanded it somewhat institutionally since then. You think of it as a mutual fund service initially and you won't go wrong. Let's talk about the CRISP survivorship bias-free database because it wasn't an important database. This didn't exist until 1997 when Mark Harhard, who was a University of Chicago Booth School of Business, a PhD student, was doing a study on mutual funds for his thesis. And there wasn't any good historic mutual fund database that captured all of the mutual funds going back in history because so many mutual funds have merged or gone out of business. So what he did with funding from Gene Fama, Nobel Laureate Gene Fama from the University of Chicago, went out as part of his PhD thesis, gathered information on mutual funds going back, as far back as the mutual funds existed in the United States and most of them have gone out of business over time. It's included in the database. So it is a survivorship bias-free. And the idea of survivorship bias, could you explain what that is in a mutual fund database that doesn't have a survivorship bias-free? What happens with survivorship bias in these databases? Let's suppose I'm working with a database that is not adjusted for survivorship bias, so it has survivorship bias built in. What that means is I'm only able to look at funds that exist today. So if I want to say, well, how did these funds perform over the past 10 years, for example, the past 20 years, I'm only looking at the funds that did well enough to survive for 10 or 20 years. What you really want to do and what this database does, as you suggest, is I want to be able to go back 20 years and say of all the funds that existed then, how did they do? Because as you pointed out, Rick, a fair number of them do not survive. So if you examine only the returns of currently active funds, you're building in a bias because you're only looking at the funds that were most successful, or to be precise, that were successful enough to survive. And we know very well that there are many funds that are not successful enough to survive, and since you didn't know which ones those were 20 years ago, the only fair way to do an evaluation is to use a survivorship bias-adjusted database to go back 20 years and see what the fund landscape was like then. And in your SPIVA report, you have all this information for all the different categories of how much, over one, three, five, 15, 20 year period of time, how many of these funds survived, or changed styles. That's another thing that you look at as well. Some funds that say that their large cap growth end up being something else, or maybe they start out as mid-cap core, end up becoming something else. So you also track a style survivorship. Yeah, and that's important. You might have a fund. Let's say a successful fund starts out as a mid-cap growth fund. It's successful. It begins to acquire assets. The manager thinks, well, if I'm going to continue to be a mid-cap growth fund, I've got to stop taking assets, or else maybe I should migrate up the cap scale and use larger stocks, which then I'm able to deploy the liquidity that I now have access to. So that's a natural sort of thing to happen. It can happen organically as funds grow. Or as managers change, manager tenure is limited in most of these places, so things change. So you looked at all of this. You were very unbiased at how you tried to do it. You had to do it as fair as you possibly could to try to say, okay, let's take these funds and compare them to this indices, and let's see how many survived, how many were style consistent, and how many outperformed the index, and how many underperformed. And here the data gets interesting because it is so consistent, not only over the last 20 years, but I've done a lot of work on this. It goes back 100 years, how consistent this data is. So why don't you tell us? Yeah, the quickest way to answer your question, I think the adequate answer, a good answer, is what the data show is that most active managers underperform most of the time. So if you look for the last 20 years of SPIVA for the large cap US manager category, which is the largest category, that's why I picked that one. If you compare those managers to the S&P 500 year by year, I think the average is something like 64%, and the average year, 64% of the large cap US managers underperform the S&P 500. Now, some years is more, some years is less. The last year in which a majority of large cap managers outperformed the S&P 500 was 2009. In the 20 years of SPIVA, I think it was only three years when a majority outperformed. So the result is suggests, and Rick you alluded to other research that goes back 100 years, and a lot of research was published on this topic before index funds started in the 60s and 70s. It's a very consistent finding that most active managers underperform a benchmark that is appropriate to their investment style. Now, one of the things that occurs is that because less than 50% outperformed, that over time this compounds, so by the time you get out to say five years, it's more than 65%. Exactly right. In fact, we just released in early September our mid-year SPIVA. So it covers the first six months of 2022. And it was a relatively good six-month period, only 51% of the large cap managers underperform. And that's really quite good considering the history I mentioned to you. Going back a full year, so the year ended June 30th, 2022, 55% of large cap managers underperform. Going back five years, 84% underperform. Going back 20 years, 95% underperform. And this is again very common. Whatever you look at SPIVA data, you see that the percentage of underperformers goes up as the time horizon extends back. And you see the same thing if you look at mid-cap managers or small-cap managers or growth specialists or value specialists. Very common effect. So this is important for investors who are long-term investors. I mean, if you're going to pick an active fund, the longer you hold it to lower the probability it'll outperform the indices. So if you're going to be a long-term investor, if you just bought an index fund, the longer you hold it, the higher the probability that index fund will outperform the active managers in that category. Absolutely. What I would say that is if you look, say, at our large cap US category where 95% of the funds underperformed the S&P 500 over the past 20 years, to turn that on its head and say if you were an investor choosing a large cap fund 20 years ago, the chance that you chose one that did better than the index was one in 20. So the odds are really quite overwhelming in the historical data that the longer you're holding period, the more likely it is that an active manager underperformed. Let's go to the next dimension of this, and that is that small minority that did outperform that 5% or 10% that did outperform. They didn't outperform by very much relative to the 90% that underperformed. And the reason I say this is because if the payoff for picking a manager that outperformed was like 10% a year excess return, then it might be worth trying to find those 10% managers because the payoff is so great. But that's not what happens. The payoff, if you actually find one of those 10% is significantly smaller than the 90% that underperformed, how much they underperformed by. So not only is it very difficult to find managers who are going to outperform and we'll get into that in a bit, but the payoff for doing it, you're not getting paid. No, exactly. I mean, the odds are against you. I mean, I think that's a fair thing to say to an investor who's contemplating, you know, should I invest in an index fund or in an active manager who follows the same style. If you want to invest in an active manager, that's fine, but realize the odds are against you. So there's a low probability of picking a manager and then the payoff is low. You also, though, look at this and say, well, you know, some of those funds that are not performing well don't have much money in them. So why are we even counting them in this study? So you actually do the study two different ways. You do it based on equal weighting all of the funds as though all these active funds have the same amount of money in them. And then you do cap weighting, which is, okay, let's look at the funds and see how much money there is actually in there so we can really find out what the investor experience is. So could you explain the difference between the two? If I'm trying to answer the question, how did the average large-cap manager do? I've got, you know, let's say a thousand, I'm making the number up there, a thousand funds to evaluate. One way to do it, as you say, is to take the returns of all 1000, add them up, divide by 1000. That's equal weighted. The problem with doing that is that you may have, you know, some funds that control $50 billion and another fund that controls $1 billion and you're treating them as if they're equally important. So the other way to do it is what we call asset weighting, where you weight each fund's return by the amount of money or the amount of assets that the fund has and then divide by the total value of all the assets across all the funds. So that gives you an asset-weighted rate of return. And the difference between those numbers will tell you whether the larger funds are doing better or doing worse than the smaller funds. If the asset-weighted number is 12% and the equated number is 10, that tells you the larger funds, or at least some of them, the larger funds did better and the reverse if the numbers are reversed. And over the long enough time horizon, what you quite often see is that both the simple average and the asset-weighted average return of these funds is less than that of the index to which it's being compared. But how much of a difference is it? I mean, is it better to get in bigger active funds than smaller active funds? It depends on the period. I don't think I could make a blanket judgment about it. This last six months, for example, I mentioned we just published a speech for the first six months of 2022. The last six months, the asset-weighted average did less well than the equated average, meaning the smaller funds did better. But there have been years when the reverse has been true. So it's probably pretty equal then over time. I would think so. Yeah, I mean, I would think so. And you've also done, you've added this over the last few years, you tried to do some risk-adjusted numbers. So you're looking at sharp ratios, which is a function of adjusting for, call it the volatility of the funds. Yes. And what has that shown? Has that, in other words, are the active funds, even though they render performing, they're less volatile? Is that a factor? No, that is, in fact, just the opposite. The average actively managed fund in the speed of a database is more volatile than the index to which it's being compared. And Medwars and I did a paper about this some years ago now. It's called the Volatility of Active Management. It's very clear that most active managers do not produce less risk than the benchmarks that you're comparing them to. The one thing that is particularly interesting about that paper, as I'm recalling it, is that unlike returns, where a manager might have good returns one year and bad returns the next, and that fluctuates, the volatility profile of funds is fairly stable. So if you have a fund that is relatively more, say, a large cap fund, relatively more volatile than the S&P 500, this year it's probably going to be more volatile next year and the year after that, and probably the same for the years in the past. So there's some stability and volatility of active funds, but there's no, I mean, some funds, yes, are less volatile, but the majority are not. So you've also do this for international equities and for U.S. bonds. So let's start with the international equities first, which is a completely different database. I mean, you're in a completely different market. It has nothing to do with the U.S., nothing to do with the original U.S. SPIVA. Is the data the same? The conclusion is the same, yes. I mean, I can answer the question really in two ways. One is that as part of U.S. SPIVA, we look at international funds or global funds that trade in the U.S., so same database, and that gets you the same answer. But our business has expanded internationally. Other clients in other regions have said, well, what about Canada? What about Europe? What about Australia? And so we've begun to publish, we have now published SPIVA, I think, in 10 or 12 different regions. Europe, Australia, Latin America, Canada, I'm sure I'm leaving things out, India. And the conclusion is remarkably consistent. I mean, there are exceptional years, yes, in all these places. But if you look at long periods of time, even an interval as short as five years, the majority of active managers underperform a benchmark that is appropriate to their investment style. It's very consistent. So let's go back, circle back to investors here in the U.S. or investing in international or foreign stocks through foreign mutual funds that are managed here in the U.S. The conclusion is the same, yes. There's no evidence that the managers of international funds do any better than the managers of domestically focused funds. Okay, I have to ask you this because a lot of people say, well, yes, but not in emerging markets. I mean, emerging markets, you can go out and you can find active managers that are going to outperform. I mean, do you find that to be true? Well, some, sure. But there's no consistency there either. I mean, I think the thing that makes SPIVA results what they are is that in most markets, including emerging markets, including all the international markets I mentioned, in most markets the investment business is very largely institutionalized. Most of the money is controlled and managed by institutional, by which I mean mutual fund, pension fund, endowments, you know, professional asset managers, which means that the managers of emerging market funds or the management of Canadian funds, the management of U.S. funds, are competing against other professionals who have the same skillset information, access, computational ability, knowledge of the market. It's a fair game. It's not like, as it might have been in the 50s, for example, it's not like one set of investors, the professionals, have access to information and trading data that is superior to that of others. And so the ones with superior knowledge can take advantage of the ones with less knowledge. Here it's professionalized pretty much across the board. Well, here in the U.S. anyway. Certainly in the U.S. and increasingly globally too, yeah. Let's get into fixed income. So you also do this with fixed income. You look at treasury indices versus managed treasury funds and corporate bond funds versus corporate bond indices and does the data hold there? Yes, although I would say it's more volatile there, because in the following sense, if you're in an environment where interest rates are increasing, as we have been down, if most fixed income managers have a duration in their portfolio that is less than that of the index to which they're being compared, then the majority can outperform in periods when rates are increasing and they will then underperform in periods when rates are decreasing. So the importance of the maturity slash duration decision, what's the average maturity of the bonds in the portfolio? What do I mean by duration? The importance of duration in fixed income analysis is just overwhelming. If you get that decision right, you can get a lot wrong and still be a really good bond manager. And because it's so important, I think you see more fluctuation. You'll see in some categories a large majority outperform one measurement period and then underperform the next measurement period simply because the direction of interest rates is changing. The other thing to keep in mind about fixed income markets, certainly the Treasury markets in particular, is unlike the equity markets where basically all of the players are, you know, what an economist might call rational profit maxim, I'm trying to make a lot of money, you're trying to make a lot of money, in the fixed income market there is one very large player who is not a profit maxim, or that being the Federal Reserve. You have another presence in fixed income that sometimes, depending on his interest rate decisions, sometimes helps the managers, sometimes hurts the managers. But there is this other factor which you don't see in the equity world. I guess it would be also a little different because the dispersion of returns among bond managers is going to be much narrower than the dispersion of returns of equity managers. Yeah, oh, very much so, very much so. In fact, David Swenson, the late head of the Yale endowment, I think at one point was quoted as saying the difference in performance between a top-desert bond manager and a bottom-desert bond manager was so small that it wasn't worth your time to try to figure out who was who. Very good. Okay, so, Spiva, congratulations on 20 years of data. I will say that, you know, I've been following this market for 30 years, 35 years, and it's remarkable that what S&P Dow Jones has done, your data has correlated so highly with others like Morningstar does the same study and the same results. No, they might use different indices, but it's the same result. And Vanguard does annual study too, and it's the same result. And they're using different indices, but again, it doesn't really matter that much. The results all come out to about the same, and the results are this. Can active funds beat the benchmark? The answer is yes, but not many, not by much, not for long, and the winners are not predictable. And with that, let's get into the second study that you do, which is called a persistence study. So tell us about persistence. The persistence scorecard, so-called, uses the same database as Spiva, so it's the same crisp survivorship bias-free database. And the question that we asked in the persistence scorecard is really very simple. It says, for example, let me identify all of the funds who were above average two years ago, and say of those that were above average two years ago, how many were above average last year? We go back five years and say of those who were above average five years ago, how many were above average four years ago, three years, two years, one, and so forth. So what you're trying to measure here then is it does the outperformance carry forward? Yeah, exactly. Recognizing that only a minority of active managers outperform in a given year, you have to focus on the more successful active managers in the historical data. Will I be more successful going forward? A simple way to say, question that the persistence scorecard tries to answer is, do winners continue? Do losers continue? Is there persistence in skill? Or does it ask, is there actually skill or is it randomness? The way to think of it is this. When you identify a manager who has outperformed, how can you tell whether his outperformance is a result of genuine skill or simply of good luck? And the answer to that question is, genuine skill should persist. Good luck is ephemeral, comes and goes. You're lucky this year, not last year or not next year. And so what the persistence scorecard does is to, at various time horizons and various breakpoints, look at funds which have outperformed historically and ask, did their outperformance continue? For example, there are many, many cuts in the persistence scorecard. One thing we do is to say, let's go back to 10 years of data, take all the managers who were above average in the first five years and say, how did they do in the second five years? And obviously if you were in the top half of the universe five years ago, in the first five years, and skill persists, the likelihood is you should have a lot higher probability of being in the top half of the distribution in the second five years than the managers who were in the bottom of the distribution in the first five years. And what the persistence scorecard tells us is that there's relatively little persistence. In other words, the example I just posed, the last time we ran persistence, if you take again all large cap managers, go back 10 years, take the first five years and say of the managers who were above average in the first five years, how many of them were above average in the second five years, the answer was 42%. So less than half. So this is like a random event. I mean if you think about it, it seems like half should be, at least half. Yeah, no, exactly right. That's the default is half. In other words, if the results are completely random, half of the managers are going to be in the top half. And it turns out that somewhat less than half of the top half managers from 10 years ago are still in the top half in the second five year period. Now what about the bottom half? Did any of the managers in the bottom half end up in the top half? Oh sure, sure. Well I'm actually looking at the data right here and it does look like the bottom half, a little less than 20% of the bottom half ended up in the top half, but about 15% of the bottom half ended up going out of business. Whereas the top half only about 6% went out of business. So I guess what you could say about the top half is that the lower probability they will merge or go out of business so they have that momentum probably because they have a lot of assets in the fund if you're in the top half. Yeah, and there's some persistence I think, not of performance but of stickiness of funds. If you're in a fund you may not want to get out for a variety of reasons. So historical success, if you look let's say at the large cap funds do this five year exercise, if you're in the top half in the first five year period you're less likely to go out of business or to liquidate. You're not particularly likely necessarily to repeat in the top half but you're likely to live to persist in terms of still being around. Now you also divided these into quartiles so you can look at the top quarter and then the second quarter and then the bottom three quarters and then the bottom quarter. And it doesn't appear overall that it's much different than random what happens to a fund. So that's a very fair summary. Again coming back to the way the exercise works, if skill is randomly distributed or results are randomly distributed in any given period 25% of the managers are going to be in the top quartile. So if you look at large cap U.S. managers again and say in the first five years what percentage of large cap U.S. managers who were in the first quartile in the first five years repeated in the first quartile in the second five years the answer again most recent persistence score curve is about 27% not really much better than random. Yeah and if you go to look at mid cap and small cap it's even worse. A manager who has skill like a bowling team or a top tennis player should continue. I mean they should continue to win. Exactly, exactly. The fact that the persistence scorecard says what it says. In other words that there is no predictive value in historical performance. I mean I think it says two things Rick. One is it reminds us that what active managers are trying to do is very difficult. It's so difficult that most of them don't do it particularly well. And secondly it reminds us as investors who are potentially identifying funds to buy that historical performance is not a good gauge of what will happen in the future. Some have said that fees are a good indication. In other words if you have low fees you have a higher probability of outperforming it. Do you work any of that into your studies? Yes there's a I guess I'll give you an answer to levels. We haven't done the study directly I know Morningstar has and the summary is exactly what you say. If you were instead of picking a fund based on past performance if you picked a fund based on I want something in the lowest quartile of fees that's a more sensible strategy than picking a fund that has outperformed by a lot recently. What we have done in SPIVA is and we don't do it every six months but we'll every year do what we call an institutional SPIVA and what we do in institutional SPIVA among other things is to take all of the funds that were in SPIVA, classic SPIVA that we were talking about and add back their fee. You do a gross. We do a gross of fees and not surprisingly somewhat fewer managers underperform when you don't count their fees but it's still a majority. I remember the very first institutional SPIVA came out. I remember doing a meeting with a client and said I'll make it as simple as I can. SPIVA tells us, classic SPIVA tells us that most mutual fund managers net of fees underperform most of the time. Institutional SPIVA tells us that most institutional managers and most mutual fund managers gross of fees underperform most of the time. The conclusion doesn't change. I mean the numbers change a little bit but the conclusion doesn't change really at all. That's interesting. I wonder how much of that is related to just the amount of cash that they have to have in a portfolio and bull markets occurring at times when there's cash in a portfolio which is hurting performance and kind of an asset allocation decision as opposed to a stock selection decision. Part of it could be that but we've done some work on this topic because we hear this objection all the time that index funds will outperform because they're fully invested in a rising market but not in a falling market. And if you look back at the falling markets in our historical database I look for example at 2001, 2002 the majority of managers underperformed, 2008 the majority underperformed so the data don't really support. Can't make that argument about cash then. I mean I think there's a slight advantage, sure. If the other thing to keep in mind of course is that we're at a point in the investment business now where sophisticated mutual fund managers which I would think would include most, if not all of them, there are ways to equitize your cash. You have to have a lot of cash on hand because you might get redumptions. You can buy index futures to so-called equitize the cash, give you the return of the equity market so I think that wasn't possible 50 years ago certainly but certainly it is today and I think that also rebuts that argument. Past performance is not an indication of future results. This is really what the persistence studies show. That is a very good summary, yes. A lot of it is random. Skill is very difficult to discern, very hard to go out and pick a manager that actually has skill and I get one of the problems with trying to pick a manager that has skill is everyone is looking for these managers and if you actually identify somebody that has skill the money is going to just pile in and that in itself could harm the performance of the fund. We see that quite frequently. One thing I think we know for sure about fund flows is that the vast majority of fund flows come into funds that have recently done very well. So especially if you think back to what you just said, Rick, that past performance is not a good indicator of future performance to allocate money based on past performance is just saying you're almost asking to underperform and of course that's what happens. It's understandable that people want to buy something that has done well except that the fact that it did well last year does not tell you much about how it's going to do this year. So let's talk about a report that I did that you've read. Yes. As an advisor for 35 years, if I just took client money and allocated it between stocks and fixed income and then within the stock side, the U.S. stocks, international stocks and on the fixed income side, treasury bonds, corporate bonds or total bond market fund and all I did was buy the cheapest index fund I could get on the U.S. stock side, a total stock market index fund or even some P500 fund on the international side, a total international fund on the bond side, a total bond market or if I wanted to have a municipal bond fund, a municipal bond index fund or something similar to it because Vanguard actually has actively managed municipal bond funds which are basically index funds because there's so many bonds in there but if that's all I did, what the study that I did which is called the case for index fund portfolios which actually came out 10 years ago with Alex Banky as the co-author, if all you did was buy a portfolio of index funds and nothing else, forget about all, forget about trying to pick active managers, forget about trying to pick managers and outperform the international market or the small cap market or whatever, just forget it, just buy all index funds the cheapest you can and maintain your asset allocation. The probability of the portfolio outperforming a portfolio that has either all active funds or some active funds in it is well over 90% and it's even higher in a portfolio sense than it is in each one of these silos like large cap, mid cap, small cap, when you put it all together in a portfolio the probability of the portfolio outperforming a portfolio with active funds in it is actually higher than the individual silos because there might be a couple of active funds that you own that outperform but the underperforming funds drag everything down. So I've been pounding the table on this and of course Jack Bogle did for years and the Bogle heads pound the table on this. Just put together a few good index funds in a portfolio and hold it for the long term and you'll be far better off. Do you agree with that? Absolutely, absolutely. I think the mistake people make and what you addressed in your study, Rick, the mistake people make is to say that diversification will help me. So I'll pick an active U.S. fund for example or maybe two active U.S. funds and then I'll diversify by picking an active international fund, an active bond fund and so forth and the difficulty is that that works if the expected return or the expected benefit of buying those active funds is positive but if the expected benefit of buying the active funds is negative which SPIVA and other research demonstrate very clearly that it is you're basically compounding the mistake. I mean another way to say that is let's suppose I go into a casino I go up to the roulette wheel and I put $10 on red then I spend and lose my money and the next roll I go I'm going to put $10 on black now because I'm going to diversify. I mean you're not diversifying anything except randomness. What you identified in that paper which is really important is to my mind is conceptually similar argument to the reason why SPIVA results are so much worse over a 20-year horizon than over a one-year horizon and that is that the probability of success is less than 50%. I am a lousy basketball player so let's suppose that I was to get into a free throw shooting contest with Michael Jordan it's possible he might miss his first shot it's possible I might make my first shot so as the unskilled player I don't want many if we have to shoot 100 free throws he's going to beat me easily even 10 he's going to beat me easily he might get lucky the first time or the second time so if you're a low skill player you don't want many trials you want relatively few. You want to rely on luck. You want to rely on luck. If you're a high skill player you want lots of trials and so in SPIVA's case lots of trials means let's look not at one year but at 20 years in the case of your paper lots of trials multiple asset classes not just one asset class but they both point to the same conclusion which is that the probability of success in picking an active manager is less than even that's why the results get worse over time that's why the results get worse when you use more asset classes I think we made the conclusion in the paper that if you were going to go with active management you put all your money on red and spin the wheel like you're saying. The more active managers you put in your portfolio the lower the probability is that that portfolio will outperform a portfolio of index it's already low to begin with but as you add more active funds the probability actually decreases and I'll just tell people that you can find this paper at my website at rickferry.com it's a case for index fund portfolios again the paper is 10 years old now it was published in 2012 but the results are the same in fact if we were to redo the study right now going back 10 years I think you'll find that the case for index fund portfolios was even higher than it was that we found in our paper let me ask you about taxes now I know that you don't include taxes in your SPIVA report but you probably have done some work on taxes because a lot of individual investors have taxable accounts and they have to pay taxes on capital gain distributions from mutual funds so has S&P Dow Jones done any work on after tax returns? we have not done it ourselves I've certainly read some of the literature and your question is the right one the thing to remember about index vehicles is that they typically are much more tax friendly than actively managed vehicles especially if you access them via an ETF which has considerable tax advantages relative to a traditional mutual fund and since it's almost any index fund that you want to have access to can be got via an exchange traded fund I'm not a tax authority or tax lawyer but that's what I do personally and I can certainly recommend it as something for clients to think about it's interesting you think that we're off going on a little different topic here but in my own personal portfolio I only have exchange traded funds now Vanguard is a little different because the ETF and the mutual fund are all the same it's all the same and all treated the same for taxes but if you're going to buy an iShare or a spider or a Schwab fund you're better off in a taxable account with the ETF because it doesn't spin off capital gains at the end of the year you only have to pay capital gains when you actually sell your shares you do have to pay taxes on the dividends but not the capital gain distributions that you would see a lot in actively managed funds so again, actively managed funds in a taxable account creates another cost because of the capital gain distributions at the end of the year and so all index funds all the time whether it's your retirement account whether it's taxable account if you're going to use it in your taxable account using ETF certainly makes sense any parting words for our Bogleheads listeners? I think the thing to keep in mind and we've written about this a number of times at S&P is two things one is the conclusion we've been talking about for the past hour or so which is the majority of active managers underperform most of the time the second thing to keep in mind is this is not a coincidence it's a random that this happens there are good reasons why this happens we've talked a little bit about cost I mean index funds are cheaper than actively managed funds I think the investment company institute estimates every year the weighted average cost of U.S. actively managed funds versus index funds it's about a 60 basis point difference as of the most recent estimate that means on average an active manager starts 60 basis points in the hole it's a lot to make up a second reason again we mentioned this earlier is the notion that in most of the world certainly in the United States the investment management business is very largely professionalized if you hire an active manager let's say from Fidelity and he's making a trade against an active manager let's say from JP Morgan the guy from Fidelity and the guy from Morgan have access to the same information they have the same research they have the same Bloomberg's on their desk they probably went to the same MBA program they have the same CFA certificate there is a level playing field there's no reason to assume that one of these guys has an advantage over the other and that phenomenon of the professionalization was identified in kind of a famous article it's famous in the index world by Charles Ellis in 1975 called The Loser's Game and this article he surveyed the then post-war history of U.S. financial market so that was 30 years in 1975 and this is exactly what he said in the 50's when the investment business was largely dominated by retail investors and relatively few professionals it was possible for the majority professionals to outperform because they had advantages that the retail investor didn't have as the business became increasingly in the 60's and 70's that advantage went away because all of the managers got better professionalization took over and we got to the point even in 1975 let alone today when it was impossible for any particular active manager consistently to have an advantage over the others that's why by the way index funds started in the 1970's because professionalization was well along by then so there were some arguments were just as good 20 years later but 20 years sooner but that's the thing that happened so the fact that the majority of managers underperform most of the time is not random it's not just a quirk of fate it happens for very good reasons those reasons still exist which means that the phenomenon is likely to continue to exist going forward so even though past performance is now I feel safe saying that if you're in index funds the future performance of index funds is going to be higher than actively managed funds in this case past performance does predict future returns I think that's a very fair statement yes thank you Craig so much for being on Bogleheads on Investing appreciate it thank you Rick this concludes this episode of Bogleheads on investing join us each month as we interview a new guest on a new topic in the meantime visit bogelheads.org the bogelheads wiki bogelheads twitter listen live each week to bogelheads live on twitter spaces the bogelheads youtube channel bogelheads facebook bogelheads reddit join one of your local bogelheads chapters and get others to join thanks for listening