More than 50 percent of mutual fund and ETF assets track passive indices, which makes index methodology important. Episode 68 is all about index history and evolution and how the usage of indexes as investment products changed the industry.
Rolf Agather has been in the indexing industry since the 1980s. He started Russell, moved around the industry a bit, then went back to Russell Investments and in 2002 became the Managing Director, Research and Innovation, Russell Indexes. He remained there after a merger with FTSE and became the FTSE Russell Managing Director, Research, North America. Rolf moved to Morningstar in 2020 as Morningstar’s Head of Research and Product, Indexes.
This podcast is hosted by Rick Ferri, CFA, a long-time Boglehead and investment adviser. The Bogleheads are a group of like-minded individual investors who follow the general investment and business beliefs of John C. Bogle, founder and former CEO of the Vanguard Group. It is a conflict-free community where individual investors reach out and provide education, assistance, and relevant information to other investors of all experience levels at no cost. The organization supports a free forum at Bogleheads.org, and the wiki site is Bogleheads® wiki.
Since 2000, the Bogleheads’ have held national conferences in major cities nationwide. There are also many Local Chapters in the US and even a few Foreign Chapters that meet regularly. New Chapters are being added regularly. All Bogleheads activities are coordinated by volunteers who contribute their time and talent.
This podcast is supported by the John C. Bogle Center for Financial Literacy, a non-profit organization approved by the IRS as a 501(c)(3) public charity on February 6, 2012. Your tax-deductible donation to the Bogle Center is appreciated.
Listen On
Transcript
00:00:10 Rick Ferri
Welcome everyone to the 68th Edition of Bogleheads on Investing. Today our topic is indexes. My special guest is Rolf Agather, Morningstar’s Head of Index Research and Products. Rolf has been in the business for 40 years and is a walking encyclopedia on index methodology and industry evolution.
Hi everyone. My name is Rick Ferri, and I am 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 nonprofit organization that is building a world of well-informed, capable, and empowered investors. Visit the Bogle Center at boglecenter.net, where you will find a treasure trove of information including transcripts of these podcasts.
Today, we’re going to be talking about indexes, index evolution, and index funds. Why is this an important topic? When I first started in the business back in the late 1980s, less than 3% of the money in mutual funds was managed to a benchmark or an index. Today, that number stands at 53%.
That means indexes are very important. But what do we really know about them? Today’s episode is all about indexes, index creation, methodologies, rebalancing, reconstitution, what makes a good index, the evolution of indexes – starting out as a price indicator for the general direction of the market and moving on to a value indicator for valuations of the market and economic indicators that were used in academics and then became benchmarks for asset managers to measure their performance against, and then the indexes themselves started to become blocks being replicated in index funds and ultimately in exchange traded funds, and how that has changed the indexing industry.
It would be hard to find anyone better to talk about this than Rolf Agather. Rolf has been Morningstar’s Head of Index Research and Products since 2020, but before that he spent most of his career at Russell, previously when it was Frank Russell & Company during the 1980s and then became Russell Investments from 2000 to 2015, where he was the Managing Director of Research and Innovation for Russell indices.
Then, in 2015, there was a merger between Russell and FTSE, who was another index provider, and Rolf became the Managing Director of Research for North America and then ultimately moved to Morningstar at his present position. So, with no further ado, let me welcome Rolf Agather. Welcome, Rolf.
00:02:57 Rolf Agather
Yeah, great to be here. Thanks.
00:02:59 Rick Ferri
Rolf, you and I go way back, we used to do index conferences 20 years ago where you were on panels as the Head of Index Research and Development at Russell, and I was the moderator and we had people like David Blitzer from S&P and Sanjay Arya from Morningstar. All of the index creators. And we discussed the creation and management of original traditional indexes, which you were very much involved in, even back in the 1980s. So, could you tell us a little bit about your background and how did you end up at Morningstar back in 2020?
00:03:42 Rolf Agather
Sure. Just to tie into your comments, I think clearly this has been a great industry to be a part of for the majority of my career. It’s the rising tide that has floated many boats. So, I feel very fortunate to have been able to be part of what was once a sleepy backwater. I can talk about what indexes really were back in the early days, but it really has come to dominate the investment management landscape.
I actually started at what’s now called Russell Investments, which is much more known as an investment manager. But when I started in 1984, Russell was the preeminent consultant to basically institutional pension plans around the world. ERISA had just passed, I think in the late 70s, and so there was quite a bit of science developing around institutional asset management.
00:04:23 Rick Ferri
Just for clarity, could you tell us what ERISA is?
00:04:27 Rolf Agather
Yeah. So, ERISA stands for the Employee Retirement and Income Security Act. It was a US regulation that basically established the concept of a fiduciary. The notion that a fiduciary must only operate in the best interest of a beneficiary of some sort. In this case a pension plan or other sort of investment vehicle. So, ERISA really established the need to have more of a science and a deliberate process around investment management.
Anyway, I spent a number of years at Russell. I also worked at Fact Set for some time, so I was able to dive into the software world, really helping asset managers pick stocks. So definitely worked on the active side as well. But then as you mentioned in the early 2000s, basically went back to Russell, got back into the index business. Russell went through an acquisition by FTSE, so it became FTSE Russell. I basically continued on in a capacity basically managing product development, product management research and then ultimately came to work at Morningstar. So, about four years ago I actually joined a number of my former Russell colleagues at Morningstar. And I’m right now the head of the product and research team for Morningstar indexes.
00:05:29 Rick Ferri
And I understand you have 40 years in the business this year.
00:05:32 Rolf Agather
Yeah. In October of this year, I will celebrate 40 years. If you knew how old I was, you could do the math. I actually started at Russell when I was in college. It was basically just a part time job. I was looking for work to help pay the bills. And so, I got a job in in the performance measurement group, basically keying in data and doing a variety of activities in what was largely the back office of Russell as a consultant.
00:05:52 Rick Ferri
Well, it’s great to have you on the podcast because we’re trying to get a view of where indexes began, how they evolved, what is involved in the construction of a good index, the different types of indexes. I believe it’s important to do this conversation about indexes and index methodology in the early years and the evolution of it because more than 50% of the money now in mutual funds and ETFs is passively invested in indexes. So, it’s a huge industry.
When we started out it was less than 3%. Now it’s more than 50%. So, I think that it’s important that we go over what makes a good index. What are we looking for in a good index? Because that is the underlying foundation for what the fund is trying to replicate. And there’s a big difference between the index and sometimes the index fund. Because one of them is theoretical and the other one actually has to be created.
So, there is some cooperation between the index providers and the fund providers to make that happen. And we’ll get into some of that in a little bit. But let’s go back way, way back. The first – you could call it an index – was the Dow Jones Industrial Average, which was created in 1896 by Charles Dow. And it had just 12 stocks. Can you tell us a little bit about that market indicator?
00:07:18 Rolf Agather
Yeah. I think at the time the goal was really to try to get some sense of how is the US market doing, and even more specifically, it’s interesting, those are actually more sector-oriented. What we would think of as sector-oriented indexes. You had the industrials, you had the transport, the utilities, and I think financials. So, ultimately it was interesting, they were really more sector based. But obviously the industrial average is the one that was the forebearer of what we now see as the DJIA.
One of the other themes here throughout the history of indexes is technology. So, at the time you didn’t have computers, Charles Dow computed the index on pencil and paper. Collect prices from the companies trading on the New York Stock Exchange, had to do all the calculations with pencil and paper. But there was again the first instance where we were now able to get an overall sense of how are US stocks behaving?
00:08:06 Rick Ferri
And the Dow Jones averages were price weighted, so it didn’t matter how big the companies were. What really mattered, and still does today, is what is the price of the stock. So, a stock that had a higher price, let’s say $50 a share, that went up 10%, it went up five points versus a stock that was at $10 a share and went up 10%, and only went up one point. So even though both stocks went up 10%, the one that went up five points counts more. It’s kind of archaic, but it worked at the time.
00:08:35 Rolf Agather
Yeah. If you think about it, the implication there, it was the way you would have constructed that in an actual portfolio is, it implies that you had the same number of shares in each company is how you achieve what we would call a price weighted index. And so obviously the problem there is shares are somewhat arbitrary. If a company’s worth $1,000, you can issue 10 shares at $100 or 100 shares at $10. So, using shares as a weighting mechanism you could argue is somewhat arbitrary.
It doesn’t really accurately reflect the economic reality of how big companies are. So, and to your point, that will have an impact on how accurate the performance is of the index that you’re getting.
00:09:08 Rick Ferri
And yet, as outdated as that methodology is, it’s still quoted in the media. The first thing that the media quotes, even Bloomberg, is the Dow, the Dow, the Dow.
00:09:19 Rolf Agather
Yes, it’s a universally known measure. I think that is probably one of its biggest strengths. It’s reported on most newscasts every day in all the newspapers, even in the index industry as vendors, we try to chastise people for using competitor indexes in any of our conversations. And yet we all know where the Dow is, right? If I asked you what level was the Morningstar Large Cap Index at yesterday, I don’t even know. I shouldn’t admit that. But generally, people know if I tell you the Dow is at 35,000, you kind of know what that means relative to history. So, it’s this sort of universal constant that we all can reference around the world, actually, it’s not just the US.
00:09:51 Rick Ferri
And the dominance of the Dow lasted a long time, because, again, technology didn’t allow anything to take its place until the first price weighted index was created by the Cowles Commission for Research and Economic. And Alfred Cowles created this index in 1938, which was merged with Standard and Poor’s data in 1957. And this was the basis of the S&P 500 Index. If I recall correctly, it was only the S&P 200, but then it became the S&P 500. And this was quite a different index.
00:10:21 Rolf Agather
Yeah. And as always, if people want to challenge you on your dates, recognize that in some cases there were precursors. I mean, there may have been other sorts of things going on, but what we all acknowledge as the real beginning of the S&P 500 was, for purposes of comparing over time, was in 1957.
What really made the S&P different than the Dow? Maybe a couple of things is it was clearly more broadly representative of the US market, so it definitely captured more than just the industrials. It would have been more representative of all the various sectors in the US economy at that time. And then also it was capitalization weighted. So even before you had all this academic theory which talked about why you might want to hold stocks in proportion to their market capitalization, which is shares times price, the S&P started to do that at that early date.
00:11:05 Rick Ferri
And about 25 years later, Frank Russell came in the picture with the broader Russell 3000 that covered 3,000 stocks and then began slicing and dicing that into the Russell 1000 Large Cap, 2000 Small Cap, and later on value and growth indices. You were there just about that time when this evolution took place.
00:11:27 Rolf Agather
Yeah. So, the first set of Russell indexes were launched in 1984, as you mentioned that was the broad market Russell 3000. The growth and value indexes came in the late 80s. And really what was going on there is we talked about the Dow and then even the S&P 500 really were designed more to sort of reflect the US market. So, if you think about the US stock market, US equity as an asset class, they were what we would call an asset class proxy.
The S&P 500 was still just an index. It was sampling from all the companies. It didn’t represent all the companies, but really the goal of the Russell 3000 was to take that to the next level and basically hold just about all investable companies in the US market. It was less about being a sample index and really being more of a comprehensive capture of the universe.
And what was happening at the time is Russell was heavily engaged in manager research. So, if you think about – this practice continues today – people that really believe in active management are, first you have to find good active managers who you believe can pick stocks. And so typically the way a manager research team does that is they have to compare those managers to a benchmark, right? It’s all relative to a benchmark.
The other thing that was happening back in that day is traditionally a lot of these large institutional pension plans were generally run by the banks as just these big sort of typically balanced funds, and you had this big equity component, but what you saw is this is also sort of the evolution of the asset management industry where you now started to have asset managers, distinct organizations outside of the bank trust departments, and they started to specialize.
So, I think the real key here was what Russell was observing in their manager research function is asset managers weren’t just buying the entire market. Some managers were starting to buy large stocks. Other managers were starting to buy small stocks, growth stocks, value stocks. So, if you had to think about a chicken and egg scenario, the chicken was what they called manager universes. The peer groups of managers were being established by the manager research teams.
And then what you found is you needed a good benchmark to represent that peer group. So as Russell was doing its manager research comparing, one of the things they recognized is you can’t use the S&P 500 for all of these managers. The S&P 500 might have worked in the large cap space, but now you have managers investing in small companies. Coincidentally, if you look at performance of small companies, it got really bad in the mid-80s. So, when the Russell 2000 was launched, small cap managers wanted a more fair benchmark. They looked really bad when you compared them to the S&P 500.
So having a better benchmark actually enabled them to not lose a lot of business. It really was the launch of the Russell indexes was really based on coming up with better benchmarks to compare active managers to.
00:13:55 Rick Ferri
In summary, the Dow Jones was created more of as a price indicator. You could tell the general direction of prices. The S&P 500 came along, it was more robust, it was used more as a valuation indicator for the market and also began to be used as an economic indicator. In fact, it was added to the leading indicators index and so it was used more in academics. So, it became more of a, in a way, an economic index.
And it was Russell who came in with the idea of benchmarking. Let’s look at the entire market, as much as it of it as we can, 3000 stocks at that particular time and use this in ways and slice it and dice it – if you don’t mind me using that term – to what the asset managers were doing. So, we can create good benchmarks and with that in mind, AIMR at the time which is now the CFA Institute was watching all of this. And I was a member of AIMR since the early 1990s.
And there really weren’t any rules or structure to what is a good benchmark? And Larry Siegel, who I had on this podcast, who is now the research director of the Research Foundation of the CFA Institute, he wrote a book called “Benchmarks and Investment Management.” And this was published back in 2003. And it listed out what makes a good benchmark.
And I happen to have that list in front of me, and I’ll go over it real quickly here. Basically the number one thing is it has to have a simplicity and an objective selection criteria or a clear set of rules, how this benchmark is being created. The second thing is it needs to be relevant and relevant to the investors and that’s what Russell was trying to do. We were looking at it saying, well you’re small cap, let’s have a relevant index that’s relevant to you.
It has to be comprehensive, so it has to cover all of the stocks that are in that category. It had to be easily replicated so that other people can pick up the rules based methodology and replicate it. It had to be stable, so not a lot of turnover.
And then there had to be no barriers of entry like there were no illiquid securities. They had to be eliminated. And finally, there was last one called expenses, which were, if carrying these stocks for some reason had expenses, then that had to be factored into the performance. But anyway, there was a set of rules that was created, and it seemed to stand the test of time because the index conferences that I went to in the early 2000s, everyone was talking about this set of rules. Do you remember that time period?
00:16:32 Rolf Agather
Yeah. And I remember because those became codified in the CFA curriculum.
00:16:36 Rick Ferri
Yes. And the candidates were tested on it. But I want to focus in on that indexes must be rules based. Can you give us more insight as to what rules-based means?
00:16:48 Rolf Agather
If I was to summarize all those rules, the whole point is if you’re going to be trying to assess an active manager’s skill from just luck or being in the right place at the right time, the benchmark had to be fair. And so, what these are really are principles of what does it mean to be a fair benchmark? It implies that there’s a passive alternative, so that if you decided that you weren’t willing to take a risk on active management or didn’t want to pay the fees, that you had a choice to do something that required less information at a lower cost. The passive alternative.
So, all of those characteristics were sort of summarized that in theory an investor should be able to say, you know something, I don’t think active management works. I don’t want to pay the fees. I can invest in something else, even if it’s in theory, the benchmark at this point was still a theoretical portfolio, but it should exist. So, the whole concept of being rules based was I should know in advance how the stocks in this index are going to be constructed. As you said, I should be, in theory it should be able to be replicated if I wanted to invest in those stocks.
It needs to be transparent as part of understanding all of these things. Those rules need to be published. And the whole point is there should be little to no subjectivity or discretion on behalf of the person managing this list of stocks. That’s the whole concept that persists to this day is there may be a lot of insight that’s used to construct the rules, but once those rules are put in place and published for everybody, that’s what’s followed. You have certainty in terms of we know how the index is going to be constructed and maintained over time. That’s the whole point about being rules based.
00:18:12 Rick Ferri
If I recall, when the Russell 1000 and the Russell 2000, which is the Russell 3000, basically split between large cap and small cap – and it was just a benchmark at the time, it wasn’t a product, there were no index funds following it – it only was reconstituted once a year. If I recall, it was based on prices at the end of May, and it was redone at the end of June. Is that correct?
00:18:38 Rolf Agather
Yes, that’s correct. Yeah.
00:18:40 Rick Ferri
So, when this became a product, when index funds started using the Russell 1000 as their benchmark, so it was a Russell 1000 index fund and a Russell 2000 index fund. So, money was actually being traded around these two indices, then front running started to occur.
In other words, hedge funds started to say, well, we know a month beforehand what’s going to go into the Russell 2000. What’s going to come out. What’s going to go into the Russell 1000, and we know how much money is being managed in these indexes. So, let’s jump the gun. Let’s get in front of them. And this caused some problems. Can you talk about that era of Russell?
00:19:19 Rolf Agather
Yeah. I think one of the side effects of being transparent is like you suggest, we’re conveying information to the marketplace in advance of the trade. So, as index providers it doesn’t impact us as much, but it clearly would impact the index funds.
There’s a number of different dimensions here. I think the first one is in terms of whether there’s front running or not, there are a lot of academic studies that were done, some which show a stronger effect for some indexes than others. The S&P 500 index effect is fairly well known, and then there was some effect around the Russell 2000. I think one of the arguments was better to provide this information broadly than not at all.
So, if you had to err on the side of well, even if we know there’s people that might try and trade advance, the solution isn’t to not disclose that information to anybody, because that just creates even more suspicion around the index. So, I think the argument is, as long as we’re providing material information to the public all at the same time, that’s the most fair approach from an index provider’s point of view.
The other thing I would highlight is obviously the fund managers were aware of this and potentially they could employ different techniques. I’m not aware of those. That might be another topic for discussion with another speaker. But clearly the fund companies, knowing that this potentially could happen, also could develop different sorts of trading strategies and timing.
Vanguard, I think, was one that was very notable for how they traded around the rebalancing. They wouldn’t disclose that. But I think it was all because they were aware of this potential impact.
00:20:41 Rick Ferri
Where I was going with this was that Russell actually changed their methodology because of it. And they went to what’s called a migration strategy, where instead of June 30th, all these stocks now come out of the 1000, go into the 2000, all these stocks in the 2000 go into the 1000 and all this took place that everybody knew, they actually began to do phase ins over time to try to slow down any front running that might occur. Could you explain the phase ins?
00:21:12 Rolf Agather
Yeah. And that’s actually one of the other dimensions and strictly speaking, that was not always done for the reason you suggest. If you look at the evolution of indexes, they started out, as you said, theoretical portfolios trying to be representative of a particular asset class or segment of the market. And if you think about it, you know that would imply that for purposes of really trying to achieve the best representation, you might rebalance really frequently. But as a practical matter, if you’re really trying to run money, you realize well that that creates turnover and turnover is a cost.
So, one of the reasons for the banding or buffering or packeting was also to reduce the amount of turnover. Because as you said, if you’re only rebalancing once a year, using prices as of May, that’s a singular event that can cause a lot of trades going back and forth. So, one of the reasons to introduce this more banding or buffering or migration methodologies is to reduce the turnover. You could argue that companies on either side of that breakpoint really aren’t that different, so there was also a theoretical argument for why you may not necessarily want to move companies wholesale or all at once.
So, it was really more to achieve lower turnover to be easier for asset managers to replicate. That would be another reason to have that buffering or banding type rule.
00:22:21 Rick Ferri
So, there were actually indexes where you could have the same stock in the growth index as you did in the value index, it was just maybe 50% was in growth or 40% was in growth, 60% was in value. And it was split that way. It was slowly migrating in one direction, slowly migrating the other. And these were the non-growth, non-value stocks, but you had to put them somewhere.
00:22:42 Rolf Agather
Right. And yeah, and again and the other motivation there was in order to be a fair benchmark, if you looked at growth and value managers, they would argue that well, there’s some stocks that growth managers might hold and value managers might hold. So, the original growth and value indexes, a company was either all growth or all value and then the indexes were completely exclusive in terms of the companies they held. But when you actually looked at value manager portfolios and growth manager portfolios, you would actually see empirically there was an overlap there.
So, not only did that reduce turnover as you suggest, but it also was more reflective of what asset managers actually held.
00:23:17 Rick Ferri
And your current firm, Morningstar, when they developed their indices, they actually decided it’s not going to be just value and growth. They were going to have a core component as well. And this was Morningstar’s way of catching the middle.
00:23:28 Rolf Agather
That’s right. Yeah. Again, it’s a much more fair representation of growth and value portfolios. The other thing I would highlight is the methodologies have evolved where Russell – at the time all we used was price to book, literally price to book was the sole determinant of growth and value – I think that’s evolved over time where Morningstar actually uses five characteristics for value, five characteristics for growth.
The idea is it’s really hard to capture the growth/value dimension with a single variable. You really need to have a more robust description of what is a value stock and what is a growth stock? And so, at Morningstar we do that for the indexes and the style box. There’s also a consistency if you look at a Morningstar style index, it is constructed very similar to using the same signals as is done for the style box.
00:24:18 Rick Ferri
It seems when I look across all the index providers, S&P, Morningstar, Russell, CRSP, MSCI, et cetera, that the cuts for large cap, mid cap, small cap are fairly similar.
So, when you look at the performance of all the different small cap indices, there’s a fairly narrow range of return. Same thing perhaps with mid cap and certainly large cap. There’s a very narrow range of return because the cuts for what is large cap, what is mid cap, what is small cap are fairly close.
But when it comes to value and growth investing, the way or the methodologies that are used to determine what’s growth and what’s value are all over the map. As you said, Russell uses a single factor price to book model, and Morningstar uses a multi factor model. All of the other index providers have their own methodology.
Value is in the eyes of the beholder, and the returns, therefore, are much more dispersed among the index providers.
00:25:23 Rolf Agather
Some accounting values can be subject to, I hate to say manipulation, but they can be managed. So, if you think about originally, value might have been defined by price to earnings. But there are definitely concerns about how earnings can be managed using various different accounting techniques.
Book value tended to be more stable, but even book value has issues. When you have companies with maybe large write offs. So, I think the whole idea is coming up with a more robust way of defining value, potentially looking at different variables, taking some sort of an average across those variables and coming up with a signal that you could argue is much more robust as you’re looking at individual companies.
00:25:59 Rick Ferri
Let’s move on to other asset classes like fixed income. So, fixed income creates a challenge. Most of fixed income is not traded on a stock exchange. So, a lot of it is just not traded. And yet you still have to come up with indices. And Lehman Brothers came up with the first aggregate bond market index. In fact, in Episode 45 of Bogleheads on Investing, I interviewed Nick Gendron, who was at Lehman Brothers and worked on the aggregate bond market index and is now doing the same thing at Bloomberg.
Have you done much work with fixed income indexes?
00:26:34 Rolf Agather
I have not. That’s not my background. My background is typically in equity. But obviously at Morningstar we have a full suite of fixed income indexes that would mirror the sorts of exposures that you would see in more of an aggregate, like formerly a Lehman aggregate.
00:26:47 Rick Ferri
It was interesting to talk with him about TIPS, Treasury Inflation Protected Securities, and why they were not added to the aggregate bond market index. And I got the feeling in talking to him where it was Lehman’s original belief that TIPS should have been added to the aggregate bond market, but it was actually the companies that were licensing that index, either as a benchmark or using it to run money against, that back in the early 2000s when TIPS were first coming out, they were the ones who asked that it not be included.
00:27:22 Rolf Agather
Yeah, there’s this tension between having an index that is the best representation of a particular asset class versus something that’s tradable. On the equity side, the analogy might be REITs. I think generally when you look at the small cap indexes, REITs tend to show up in fairly healthy proportions.
There’s some asset managers that either aren’t allowed to invest in REITs or don’t like REITs, and so you would have seen a similar thing on the equity side where there’s people who would have said, hey, can you keep REITs out of the portfolio? So, there’s a similar principle there in terms of what is the real purpose of the index and why you might try to leave certain stocks or securities out.
00:27:56 Rick Ferri
And this gets to my next point, which is nobody was telling Russell back in the 1980s what to include in their Russell 3000 or even in their original growth and value and size indexes. Nobody was telling you that. You were coming up with the benchmark and the asset managers didn’t have any real say in it.
It seems though, things have changed. And that the index providers now are creating indexes, yes that are benchmarks, yes that are used as the basis for index funds, but that the asset managers who are licensing these also have some say in this. In fact, in my interview with David Blitzer, which was Podcast #2, he talks about going out to the major holders of the S&P 500 and discussing these issues with them. Not to say particular companies, but just a more significant part of changes that occur in indexes have at least some participation by the licensees of the index. Would you agree with that?
00:29:01 Rolf Agather
Yes, I think as we look at now the latter part of the index history, it’s more about indexes being created as the basis of an investment product. The last 20 or so years or 30 years. Especially with the advent of ETF. And so yes, I think it’s more about choice. What you were mentioning, it started out much more research driven.
So, Russell and other firms as a result of their research, and again doing things like assessing active managers, would come up with what they felt were the right principles. But as the industry has evolved, it really is now that people are now being given more choices of how they want to invest. The ability to get more discrete granular exposures. And so there’s a lot more input now from clients, from the market, in terms of what sort of additional choices do people want to have and what should they have? It is much more demand driven now than research driven, I would say.
00:29:53 Rick Ferri
I remember when South Korea was going to be taken out of the emerging market index and put into the developed market index, it was a big discussion about this. In fact, one index provider did it and another one didn’t. And I’m not sure where that fits right now, but there was a big discussion about this. How much is it going to impact our portfolios. Do you recall that?
00:30:14 Rolf Agather
Yeah, we’re in that camp as well, where we continue to have South Korea as an emerging market. But this really comes back to the benefit of having a rules-based methodology because we have rules for how companies get classified as developed or emerging, which determines their inclusion. So, the beauty of not having to have some sort of an internal meeting is if someone says, well, why do you have South Korea in your index that’s different than the other guy? We can point to the answer.
Here’s the rules we use. You could anticipate if South Korea were to move. You can see what were the criteria that were used. You can anticipate that. So, you may not agree with us, but at least you can understand why we did what we did. Because as you said, it could be debatable depending on the criteria you’re using for defining who’s developed, who’s emerging, you might get a different answer. So, I think it’s important to be transparent and have rules around why you make those decisions.
00:31:04 Rick Ferri
And I just want to touch on this last point and call it the care and feeding of indexes. There’s some terms that we threw around out there. One of them is called reconstitution and then there’s rebalancing. Talk about the difference between reconstitution and rebalancing of an index.
00:31:19 Rolf Agather
Yes, this one we get into sort of the arcana of the jargon, but strictly speaking reconstitution means we basically reapply all the rules for determining the membership. So, if we’re looking at something, let’s say, like trying to construct size indexes, we would basically go out and look at the market capitalizations of all the companies and reconstitute the indexes in terms of who’s now in the large cap index, who’s now in the small cap index.
We would also then assign weights accordingly. A rebalancing is you don’t necessarily reset all of the membership, but sometimes shares outstanding changes happen for companies. So rebalancing is when you basically reset the company weights back to their targets. So, you aren’t changing the selection of stocks, you’re sticking with the existing set of stocks, but you’re now rebalancing them back to their target weights. And that’s sort of more of in a general way. So, reconstitution means completely choosing or resetting the membership of the indexes. Rebalancing typically means we’re just adjusting the weights of those constituents back to their targets.
00:32:15 Rick Ferri
And Russell did this once a year. How often does Morningstar do this? The reconstitution part.
00:32:21 Rolf Agather
Best practice, generally, is semi-annually. If you’re only rebalancing one time a year, you have to hold on to that now for another year. So, I think semi-annual rebalancing seems to be this happy medium in terms of really making sure you’re representing the market by reconstituting the index every six months and not just holding it stale for an entire year.
00:32:39 Rick Ferri
And how often do you rebalance shares outstanding?
00:32:42 Rolf Agather
Quarterly. Effectively the portfolio is the weights are being reset on a quarterly basis. Two times would be because of the reconstitution, but then off the reconstitution cycle, there’s a rebalance happening.
00:32:53 Rick Ferri
And what about big new issues that come to the market? Are there exceptions for big new issues? How would you include that?
00:33:00 Rolf Agather
Yeah. Generally, IPOs come in through the normal reconstitution process. The goal isn’t to capture these things as quickly as possible. Generally, you want to have some sort of a seasoning. And so, the expectation is IPOs will generally get picked up at least within six months during a reconstitution cycle. I think that’s the other thing to highlight is obviously we have the quarterly cycle, rebalancing and reconstitution, but we’re basically managing a paper portfolio that has to replicate the experience that an investor would have.
So things like splits other corporate actions. We actually are simulating those as we manage the index and that’s on a daily basis. So, spin offs, mergers, acquisitions, anything that’s going to happen to a company in reality, we have to reflect in the index portfolio. And so that’s something we do on a daily basis. Reflecting corporate activity.
00:33:43 Rick Ferri
And I want to talk about the difference between full market replication and free float. Back in the late 90s, there was a problem. Wilshire 5000, if you recall, at the time of a lot of the stocks in there were, you couldn’t really replicate the Wilshire 5000 because so much of the stock was held by insiders. And so eventually index providers changed their methodologies to go to free float. So, could you describe the difference?
00:34:11 Rolf Agather
Yeah. The academic theory would talk about holding companies in proportion to their total market capitalization, which is their shares outstanding listed on the exchange times their price. That’s the total market capitalization of the company. It’s what we used to reflect the size.
But as you mentioned, what actually happens is not all shares outstanding are necessarily trading in the market. Sometimes they’re locked up. So, you can have situations where a company might have only 50% of their shares are trading. So even though you might see their total market capitalization as one number, in reality only 50% is available. So, if you think about it, if someone is trying to replicate an index with a large amount of money, if they try to achieve that relative weight in some cases, especially for smaller companies, it can have a price effect.
And in fact, I think the reason S&P moved to float back in the early 2000s was because they had a major addition that came in at total market capitalization, all the index fund managers were trying to hold that at the weight, and it created this huge price impact. I can’t remember the stock, but it would show the impact of trying to hold a company based on its total shares outstanding as opposed to its float shares outstanding.
00:35:14 Rick Ferri
A lot of people call this the opportunity set for investors. In other words, the free float is the opportunity set we all have on the outside that we can invest in. The shares that Mark Zuckerberg holds or Jeff Bezos holds of his own company, they’re not part of the free float. So, this is what’s available. And this is how the indexes are created. They all moved or migrated to free float I want to say in the late 1990s and early 2000s. And basically now, with the exception of very few indices, all of them are free float.
00:35:45 Rolf Agather
Yes, and free float really is a proxy for what we would call capacity. So how much of a stock realistically can you hold and before you start to create price impacts. It’s related to liquidity. Some people would say liquidity. But strictly speaking we use the term capacity. But float is really a proxy for how much of this stock is really available for the general investor such that index funds or all investors in fact can get positions in that particular security.
00:36:11 Rick Ferri
We’ve talked about the S&P, and we’ve talked about Russell indices and there of course many other total market indices out there. But one of them is my favorite wonky index. And I know that a lot of people who are listening to this may not agree with, is the NASDAQ 100.
Which is what QQQ – the ETF QQQ – is a benchmark to. And the history behind the NASDAQ 100 is a little bit wonky in itself because the New York Stock Exchange decided the order to get publicity, they were going to create their own New York Stock Exchange Index and it started to get some publicity.
So the NASDAQ said, well, we need our own index. So, they created the NASDAQ 100, which was supposed to be the largest 100 stocks that trade on the NASDAQ. But then they decided that financial stocks were not going to be included even though there were financial stocks trading on the NASDAQ. And then there’s a cap on the value of any particular stock that you could have in the index, and nobody really paid that much attention to it until it became an ETF. At the time it was QQQ’s and now it’s just QQQ. Could you talk about that evolution?
00:37:19 Rolf Agather
I would never want to cast aspersions on other index providers, but the importance for investors is it’s why you need to understand the rules. How is this index being constructed? What are the parameters, and what’s the outcome? And there’s transparency, right?
So, the point is as an investor, I’m not going to categorize it as good or bad, but you should be aware that using an exchange as the criteria for creating a portfolio is somewhat arbitrary. What’s the economic rational in using an exchange? Because as an outcome, what you find is you basically have 60% technology. You can see the exposures. It doesn’t have the same sector profile as the economy. So, you’re making a bet somewhere. So, it’s important for you to be aware what exposure is this index really giving me? Again, exchange criteria don’t reflect, as far as I know, any sort of economic rationale for investing.
00:38:03 Rick Ferri
No, I agree. And when they came out with industry sectors and they came out with the technology industry sector ETFs and index funds, I said, well, if you want technology, why wouldn’t people just buy that? But no. QQQ, just like DIA for the Dow Jones Industrial average, I still see a lot of client portfolios that have this in it. And I kind of scratch my head at times.
I want to continue to get into indexes as a product as opposed to indexes as a benchmark or economic indicator or price indicator. This has become a huge business. As an index provider, you’ve been in the industry, I mean you used to get paid at Russell for using your benchmarks if somebody wanted to use your benchmarks. But now, as a benchmark for ETFs or for index funds, it’s a really lucrative business. Is it the most lucrative part of indexing?
00:38:56 Rolf Agather
Depends on the provider. Clearly it it’s become a bigger and bigger component of most index providers revenue streams. Every provider potentially has a different mix of how much they make from data subscriptions versus licensing for investable products, and then even things like derivatives. Potentially you might consider that a third stream.
Because as we’ve talked about, we sort of captured the first two uses of indexes. Asset class proxy, a measure of the market, and then benchmarks for active managers. This latter stage of the index world is the use as the basis of an investable product. Since the advent of the ETF in the early 90s, that’s really now where we’ve seen most of the growth.
But as I mentioned, prior to the ETF launch, you would have seen providers like Russell and MSCI would have made the majority of their revenues licensing the index data to active managers because they were using it as a benchmark. S&P was actually making money more around trading of futures and options. So, you kind of had that investable side there. But once the ETF got launched, what you saw is incredible demand now to use indexes as the basis of funds, which is now continued to this day. So, it’s definitely a much bigger component of our revenue streams as it has been. But most of us still make a decent amount of revenue from the data subscriptions.
00:40:05 Rick Ferri
I have a term for the indexes that are created solely for the purpose of creating a product. Special purpose indexes. “Spindexes” for short. You see some crazy indexes. This is an index of CEO’s who are shorter than 5’10” and manage companies West of the Mississippi. I did the backtest on this and you know what? Yeah, I mean it works. It outperforms the market.
But I can’t create an index because if this is my great idea to create this index fund based on this index that I came up with, I have to go to an index provider. And S&P, whomever, and I say, how much money will you charge me to create and run the index for me that I can then license it from you? This is a big business.
00:40:57 Rolf Agather
Yeah. Obviously as a product provider, I’m going to talk about choice. It’s a free market. If there’s someone willing to pay for it, potentially there’s a provider willing to build it. But I think what you’re really touching on is the key thing for investors to be aware of, you have to be really intentional on what do you believe, what’s your conviction, right? What are you saying about this particular group of companies? What do you think is the rationale behind investing in them?
As an index provider, I’m going to give you the best exposure I can. But I’m not the one telling you that these companies are necessarily going to do better. The outcome I’m promising you is some exposure to some characteristics. So, I think it’s important for the investors to say, why do I think these companies are going to do better? What’s the reason I want to get this exposure? It comes back to what do I know that others don’t know? It’s all about information. Presumably efficient markets would tell me that all companies have been priced equally. Therefore, if I’m willing to buy one of these more narrowly prescribed indexes, I must think I know something that the market doesn’t. Those principles still apply as we think about all these new products.
00:41:54 Rick Ferri
But a lot of the indexes that are created that way, these special purpose indexes that are just created for the purpose of creating a product; they don’t last. A lot of them don’t last. And when the product dies and closes, the index goes away. So, we see a lot of that. We see a lot of creation of indexes, then indexes disappearing and so forth.
But the core classic traditional indexes continue, of course. There’s the S&P 500, and then there’s the total markets. Vanguard uses the CRSP for the total market. You’ve got the Russell 3000. S&P has total market, and Morningstar has broad market indexes as well.
00:42:31 Rolf Agather
And style indexes, yes.
00:42:33 Rick Ferri
Style indexes. When we’re talking about more than 50% of mutual fund money is now in index funds, I think it’s safe to say that most of that money are in these core products. Would you agree?
00:42:46 Rolf Agather
Yeah, the last time I looked at this, I think over 70% of AUM and flows are attributable to the more broad market products across a variety of vendors. Even though a lot of the press is around what you call the “spindexes”, most of the flows and AUM that you hear reported is still going into the broad beta products. That is a reality.
00:43:06 Rick Ferri
Now I want to get into one more aspect of indexing. Obviously, back in the day, and this is during the early 2000s. It’s not that way anymore. But in order to launch an ETF, you had to have an index. You had to follow an index. There were no actively managed ETFs.
If somebody wanted to launch an ETF, they had to have the index created first. So I remember, I think it was 2004, a new index was created called an Intellidex index. And it was created by the American Stock Exchange for the benefit of PowerShares. PowerShares was a new ETF company that launched ETFs tracking these Intellidex indexes.
And when I first looked at those Intellidex indexes, I said there’s no way the SEC is going to allow this to be called an index. But I was wrong. Because these Intellidex indexes were quantitative fund selection, alternative weighting, they weren’t cap weighted, they were weighted by other methodologies using fundamentals. And they were pure active management.
The only difference was they were created first in an index and then the fund manager PowerShares followed within 80%. Because that’s the rule, right? If you’re a fund manager, you have to do 80% of what the index is to be able to call yourself an index fund. But they were calling themselves a passive index fund provider. And I almost fell off my chair when I saw this. I said, where is the SEC on this? But this has now become the norm. Active management can get away with being called passive index investing if they created some sort of a rules based index first.
And not only that, it used to be you had to go out and you had to hire an indexing company, now you can do self-indexing. So, me as the fund company, I can create my own index. And then I can follow the index in a fund or an ETF, and I can call myself a passive index fund manager. How do you feel about that? Because you could see I have issues with that.
00:45:11 Rolf Agather
Yeah, there was a huge turmoil in the industry. Not only the Intellidex, but I think you saw with the fundamental index a few years later. Because again, most people’s concept of an index was it has to be broad market, market cap weighted. But clearly something like an Intellidex was obviously an active strategy. I mean I could make an argument that even when you start to break up the market by large, small, growth, and value, that implies you’re making an active decision somewhere. But most people didn’t think like that.
So, there’s been this incredible blurring now of what’s active, what’s passive. And I think that just highlights, as you say, you can now have much more active strategies doing a variety of different things, but positioned as an index for purposes of an ETF. For me, it really begs the question then, what is the distinction now? And I think it does come back to rules-based and transparent. It’s no longer this academic definition of active and passive, because everything’s really active now.
Once you go past the broad market portfolio, you’re now in some sort of active space. So really the distinction now is, is it truly rules-based (i.e. is there any discretion) and is it transparent? Because obviously the whole concept behind an active strategy is, I must know something the market doesn’t, I’m not going to share that because I’ll lose that edge. So, in theory if I’m making something transparent, it’s not active.
That’s sort of reverse logic, but the litmus test I’m applying is if they are making the rules transparent, it’s still what we would call an index.
00:46:31 Rick Ferri
It’s interesting, the pivot these active managers took over the years. The way they responded to Jack Bogle back in 1976 was, indexing is unamerican. Well, I think we’re beyond that.
Indexing is not only here to stay, but it’s probably the preferred strategy. So, if you’re an active manager, you’ve got to do something else. You’ve got to muddy the water. And you’ve got to try to convince people that what you’re doing is better than traditional broad cap weighted indexes.
We were at an indexing conference back in 2007 when the S&P 500 was 50 years old, and Research Affiliates ran a full page ad in the Wall Street Journal congratulating Standard and Poor’s for 50 years and then said, “and now we’re the indexes for the next 50 years.” And just because they’re trying to muddy the water, they’re trying to confuse people as to what the John Bogle type of indexing (which is the basic traditional indexing) they’re trying to say, oh, this is just as good, if not better. And a lot of people believe that. And I see these in portfolios, and they’re always more expensive, a lot of them don’t work, but they exist.
00:47:33 Rolf Agather
Well, the comment I would make is, that’s as American as you can get. The free market providing lots of choices to people. But I think for me, it’s also a very Darwinian process. As you suggested earlier, the fund companies will launch these products, but not all of them survive.
In theory, the ones that survive are the ones that get flows and AUM. So, there must be some demand from there. To me, this is a sign that the free markets are still potentially working. It just sometimes takes time to weed out the ones that don’t make sense for anybody.
00:47:57 Rick Ferri
OK, fair enough. So, I don’t know if Morningstar has done this, but you at Morningstar, I don’t know how involved you are in this, but you do your benchmarking studies every year where you look at all the active managers and you say how are they outperforming your indices or are they not outperforming? Are you involved in that study at all?
00:48:14 Rolf Agather
I’m not involved. I know what you’re speaking of you. I definitely review that when it comes out, yes.
00:48:19 Rick Ferri
So let me just ask the question to you – and I’m kind of leading you here – I mean, has the data really changed? Because I know when S&P started coming out with the SPIVA study, and even before that there were other studies, academic studies. So, I’ve been looking at these studies, reading these studies, and have done my own studies for more than 25 years. And the data always seem to come out the same. If you’re just using these traditional low-cost total market index funds or broad market index funds, active managers have a very difficult time outperforming. Is that what you saw at Russell? Have the numbers changed very much?
00:48:51 Rolf Agather
Not really. You might argue that it’s getting a bit worse. And I’m not sure if there’s been attribution there. But I think what’s really important for people to understand is this whole distinction between or debate around active and passive management generally focuses on are markets efficient or not? That’s always been the traditional belief you have to have as to whether active management works.
But to me an even probably more important principle that Sharpe defined was the zero-sum game. His point was the sum of all investor portfolios equals the broad market, which is just a big average, right? So, by definition, passive managers are going to get the market experience by investing in the market portfolio. So, his whole point was it’s a zero-sum game, and so half the people have to outperform by definition, half the people have to underperform.
So, even going into these studies, your expectation should be that about half of anyone trying to be active is going to underperform the market average. We all can’t be above average. To me that’s a really important principle for any investor to think about. Because it comes back to, do I know something or do I think I can find someone who knows more than the market? Because half of the time I’m going to be right and half the time I’m going to be wrong. So, to me the whole zero-sum game principle, the arithmetic of active management, is the more powerful concept to think about as an investor.
00:50:08 Rick Ferri
And it’s active management, less fees. Morningstar does a study every year where they analyze the fees based on capitalization weight, and the last one I saw showed that active equity funds, although they’ve come down a lot, are charging still about 0.6% on average, whereas index funds are much lower like 0.06%. And that difference, that 55 or 60 basis points difference, is meaningful. I mean you have to get over that hurdle rate before you even start to outperform.
And the bonds, it’s about 40 basis points difference between the active bond funds and the index funds. Now, that’s just one asset class. You’re looking at US stocks, international stocks, bonds and they all come out to about the same over the long term. But I did a study with Alex Benke who used to be at Betterment. We did a study back about 10 years ago, and we made a case for index fund portfolios. But because if you’re looking at all these things just in isolation, you could say, well index funds outperform active funds in the US stock market 75% of the time over a five year period of time, and in bonds that does this, and international that does that.
But what if you were going to have a diversified portfolio of US stocks, international stocks, bonds, maybe real estate. Maybe just those four asset classes together in a portfolio? What if you just used all index funds? Well, actually the performance of that portfolio goes up.
Because, yes, maybe one of them you may have, if you were trying to pick active funds, you may have picked (out of those four) you may have picked an active fund that did better. But the other active funds you picked maybe did worse. When you put them together, the dynamics of putting a portfolio together of a lot of traditional low-cost index funds is that that portfolio outperforms a portfolio of active funds almost 90 something percent of the time.
So, in my view, we’ve got products out there like balanced index funds and target date retirement funds that follow indices. I think these are wonderful products. How do you feel about those?
00:52:08 Rolf Agather
Yeah, people just assume that investing is hard and therefore you need to spend a lot of time and energy and hire all these smart people. But I think the reality that you’re highlighting is there are possible ways for you to achieve your goals with fairly simple portfolios and simple products. I think there’s no question for that.
But I think there’s human behavior that becomes part of it. Every once in a while, I might go to a slot machine. I’m smart enough to know that I’m probably going to lose all my money, but to be honest, some of it can be entertainment. I think some people like to brag at cocktail parties. So, it’s definitely a fair point. And I think I would also reemphasize maybe something that I haven’t said directly. The only true passive investor is holding the broadest portfolio, market portfolio possible.
I think the point I made earlier, even if you’re starting to hold large cap, small cap, growth, value in these broad exposures, there’s an active decision in there somewhere. So, I think as much as we highlight that active managers, it appears that they’re not able to be their benchmarks, I’d say anyone who’s starting to do anything active, even if it’s just managing those more discrete exposures, it’s the same principle.
Again, what do you think you know more than the broad market portfolio? It even simplifies things more. In essence, if you truly wanted to be a passive investor, you would just hold the broad market. You don’t even need to hold the broad market components like large, small, growth and value.
00:53:22 Rick Ferri
We have something called the three-fund portfolio which does exactly that. So, with that, I couldn’t ask for a better ending. Rolf, it’s been wonderful having you on, and great to catch up with you again. And thank you so much for being on the Bogleheads on Investing podcast.
00:53:36 Rolf Agather
Yeah. Thank you, Rick. This is a great opportunity. I hope this was useful for everyone.
00:53:40 Rick Ferri
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 boglecenter.net, bogleheads.org, the Bogleheads Wiki, Bogleheads Twitter, the Bogleheads YouTube channel, Bogleheads Facebook, Bogleheads Reddit, join one of your local Bogleheads chapters and get others to join. Thanks for listening.
