The Long View

Rob Arnott: Don't Sleep on Value Investing (Especially Emerging-Markets Value)

Episode Summary

The Research Affiliates founder thinks "value" is cheap, sees shades of the 2000 tech bubble in today's market, and takes issue with multifactor investing, among other matters.

Episode Notes

Our guest on the podcast today is Rob Arnott. Arnott is partner and chairman of the board of Research Affiliates, a firm he established in 2002, following stints at First Quadrant and Salomon Brothers. He also runs several prominent mutual funds, including PIMCO All Asset. In addition to these duties, Arnott is an accomplished thought leader, having published more than 100 articles in professional journals. Among other plaudits for his work, he has received seven Graham and Dodd Scrolls, awarded by the CFA Institute to the top financial analyst journal articles of the year. An innovator, Arnott popularized the concept of fundamental indexation, which some refer to as smart beta. 

Background Information

Related Links

Research Affiliates Forecasts

Research Affiliates Research

Other Research Referenced

About the Podcast: The Long View is a podcast from Morningstar. Each week, hosts Christine Benz and Jeff Ptak conduct an in-depth discussion with a thought leader from the world of investing or personal finance. The podcast is produced by George Castady and Scott Halver.

About the Hosts: Christine Benz and Jeff Ptak have been analysts and commentators on investments and the investment industry for many years. Christine is Morningstar's director of personal finance and senior columnist for Jeff is head of global manager research for Morningstar Research Services, overseeing Morningstar's team of 120 manager research analysts in the U.S. and overseas.

To Share Feedback or a Guest Idea: Write us at

(Disclaimer: This recording is for informational purposes only and should not be considered investment advice. Opinions expressed are as of the date of recording. Such opinions are subject to change. The views and opinions of guests on this program are not necessarily those of Morningstar, Inc. and its affiliates. Morningstar and its affiliates are not affiliated with this guest or his or her business affiliates unless otherwise stated. Morningstar does not guarantee the accuracy, or the completeness of the data presented herein. Jeff Ptak is an employee of Morningstar Research Services LLC. Morningstar Research Services is a subsidiary of Morningstar, Inc. and is registered with and governed by the U.S. Securities and Exchange Commission. Morningstar Research Services shall not be responsible for any trading decisions, damages or other losses resulting from or related to the information, data analysis or opinions or their use. Past performance is not a guarantee of future results. All investments are subject to investment risk, including possible loss of principal. Individuals should seriously consider if an investment is suitable for them by referencing their own financial position, investment objectives and risk profile before making any investment decisions.)

Episode Transcription

Christine Benz: Hi, and welcome to The Long View. I'm Christine Benz, director of personal finance for Morningstar.

Jeffrey Ptak: And I'm Jeff Ptak, global director of manager research for Morningstar Research Services. Our guest this week is Rob Arnott. Rob is partner and chairman of the board of Research Affiliates, a firm he established in 2002, following stints at First Quadrant and Salomon Brothers. He also runs several prominent mutual funds including PIMCO All Asset. In addition to these duties, Rob is an accomplished thought leader, having published more than 100 articles in professional journals. Among other plaudits for his work, he has received seven Graham and Dodd Scrolls, awarded by the CFA Institute, the top financial analyst journal articles of the year. An innovator, Rob popularized the concept of fundamental indexation, which some refer to as smart beta. We're pleased that he can join us today.

Rob, welcome to The Long View.

Rob Arnott: It's a pleasure.

Benz: Thank you for being here. Let's get right into the first question. I'd like to talk about value investing. Though you probably wouldn't identify yourself as a value investor through and through, value is an important input to your approach, and value has really struggled. So, let's talk about its impact on the performance of some of your better-known strategies like PIMCO All Asset. What has the experience of the past decade with value's travails taught you, and how is that reflected in the way that you manage portfolios today?

Arnott: Well, firstly, I'd cheerfully acknowledge that we have a pretty strong value bias. We recognize that historically growth has tended to trade at a premium that is actually larger than subsequent growth would justify. There are exceptions, though. The cycle for growth versus value can be a long one. From 1984 to 2000, value underperformed, with almost all of that underperformance in the last 12 years and about half of it in the last two years.

So, what we find is that you can have protracted periods of time when value underperforms. This most recent span is so far 12 years and rivals the experience of the tech bubble, both in terms of magnitude of underperformance and in terms of duration. But the market does go in cycles; even if they're long cycles, it does go in cycles. And the important thing to note is, Did value underperform because the value stocks themselves performed really badly as companies--as businesses--or because growth itself structurally performed better than it normally does as businesses? In other words, was their underlying growth faster than normal? And the short answer to that is no.

The short answer to that is: If you look at the relative valuation of growth versus value, value has grown cheaper by more than it underperformed. And that's really important. If a style underperforms by 1,000 basis points while getting cheaper by 2,000 basis points, that's a buy, not a sell. And that's sort of the picture of value investing today. It's trading in its cheapest quintile historically. In emerging markets, it's trading in its cheapest decile, cheapest 10% of all of history. And the spread between growth and value is wide enough that just coming back to historic norms, just getting repriced to the normal growth value, valuation spread, would deliver over 2,500 basis points of incremental performance in the U.S. and over 4,000 basis points of incremental performance in emerging markets. So, to me, this looks like a classic long value cycle that has been painful.

In terms of our strategies, we've performed below expectations but way above the medians for value managers.

Ptak: And so, maybe reflecting back on when value began to underperform growth, I mean, hindsight is 2020, of course, but when we look back on that period, and you do so through your analytics and tools, do you see some of the telltales that value was about to enter one of the cyclical periods of underperformance?

Arnott: At the time, I was not bullish on value. At the time, I did not think that value was cheap. 2007, the relative valuation of value stocks versus growth stocks had risen to a point not seen since the 1980s. And we knew what happened after the 80s. This is not to say that I lay claim to press the unseen, saying value is going to have a bad decade. I didn't. But I didn't have the valuation tools we have now. And in fact, this drawdown for value inspired some of the work that we've done.

Back in 2007, relative valuation for value was high. Quantitative strategies, multifactor strategies, the value tilts, and leveraged long-short value strategies were very popular. And we saw billions of dollars flowing every quarter into these kinds of strategies. Well, lo and behold, you had the quant crash in August of 2007. There was one major hedge fund that lost 40% of its value in three days. Oh, my goodness. And that was leveraging long value short growth, long momentum short underperformers and so forth, and doing it on a global basis. So, they thought they had broad diversification across factors and across geography, but they lost 40% in three days. That's the end of the last value cycle.

Now, Fundamental Index was live at that time, and it had its smallest bets ever. That is to say, its tracking error relative to the market was tight. It had about a 1.5% tracking error. Normal is 4%. And so, when value and growth are priced similarly, that is, when the gap between growth and value is small, we're going to have a small value tilt. And that's what we had. By the end of the financial crisis, the spread between growth and value was near record extremes. The value stocks were trading as if they weren't going to survive. And at that point, RAFI wound up taking on a deep, deep value tilt, with the result that just plain-vanilla RAFI, which underperformed by 3% in 2008, outperformed by 15 percentage points in 2009--just huge outperformance.

So, what we find is that our value strategies go deeper value when value is cheap and lighten up on the bet when value is expensive. And that helps enormously over a market cycle. But as we've seen in the recent value cycle, some kinds of market cycles last a long time.

Benz: So, in terms of the current environment, does it remind you of sort of the late-'90s period when some of the high-tech names were so extremely overvalued and value and small value, in particular, were poised for a serious recovery?

Arnott: It reminds me a lot of the peak in 2000. We saw narratives back then of a new paradigm of: Earnings don't matter; what matters is reinvesting for future growth. The standing joke at the time was: We lose money on every sale we make, but we make it up on volume. (laughs) OK. Today, you're seeing similar narratives that these companies are disruptors, that these companies are taking out entire industries--Amazon being a classic example of that, and Google and Facebook just undermining the entire advertising industry. The narrative is: These companies that prosper based on advertising--advertising in these electronic media will not be cyclical. Well, that's not proven. That's just a hypothesis.

Now, it bears notice that disruptors get disrupted. Palm in the year 2000 was briefly priced at greater market capitalization than General Motors. OK ... How many people today still have a Palm? You had BlackBerry priced at lofty valuation multiples, and everybody was going to use a BlackBerry, and then a year or two after BlackBerry sales peaked, the iPhone was introduced. OK, so now you had smartphones, and disruptors got disrupted.

So, compare today with the year 2000. In the year 2000, best of my recollection, four of the top seven market-cap stocks were technology stocks. It was called the tech bubble. Well, as of year-end 2018, you had seven of the eight largest market-cap stocks in the world, all technology-centric companies. I say technology-centric because you have companies like Facebook and Amazon that aren't classified as tech companies. Amazon is consumer discretionary, for instance. But the fact that it's not categorized as technology – everybody knows it's a tech stock, really. The essence of its competitive advantage is technology.

So, when you have seven of the eight largest companies in the world, all drawn from the tech sector, that's a level of concentration that's been seen only once before in history, and it was not the tech bubble--it was the Japan bubble. Peak of the Japan bubble at the end of '89, nine of the 10 largest market-cap stocks in the world were Japanese stocks. So, what happened over the subsequent 10 years? Well, eight out of the 10 disappeared from the list…weren't on the top 10 list. Nine of the 10 underperformed the ACWI Index. So, you have a history in which the utterly dominant sectors--the ones that are ostensibly poised for assured long-term success--those are the ones that are priced for perfection and--when they fail to deliver perfection--perform poorly.

One last observation about the comparisons with tech. I've heard people say the tech bubble wasn't wrong; it was just early. Look what's happened to tech recently. Actually, it's different companies. The 10 largest market-cap stocks in the tech sector in the year 2000--one of them has beat the market since the year 2000. That's Microsoft. How big a margin did it outperform by? 1% a year? That's it. OK, well, that's interesting. What happened to the other nine? Well, they all underperformed, of course. Four of them had positive returns averaging 1% to 2% a year. That's all. Five of them had negative returns. Two of them disappeared altogether. The ones with negative returns had an average return of minus 7% per annum compounded. So, if you look at the entire list, you're looking at a list that, on average, delivered minus 1% or minus 2% a year. So, when people say, "Look, the tech bubble was just early." No, it's different tech stocks. The dominant tech companies today, with one notable exception, weren't on the top 10 list back at the last tech bubble.

Ptak: So, taking your point that there's concentration in technology, and it sounds like, in your view, you see some eerie parallels to what happened around the year 2000. I mean, hasn't concentration in a handful of names at the top of the index been kind of a hallmark of indexing? And it just so happens that while some of those names may fall out over time, the index still proves to be a very formidable opponent to active investors. What would you make of a counterargument like that to some of the observations that you just shared?

Arnott: Firstly, indexing is a formidable opponent to active management. I'll come back to the Achilles' heel of indexing shortly. But by definition, indexers track the market. And if you strip them out of the market, what are you left with? The selfsame portfolio. Active managers divide the rest and collectively they hold essentially the same portfolio--the only difference being the names that are left out of the index. So, active managers collectively can't win, but they can win if other active managers are losing. So, you have to ask the question: What are the characteristics of active managers that underperform and, in the long run, the underperformance of the performance-chasers--those who pour money into stocks and strategies that have performed brilliantly and yanked money out of stocks and strategies that have been disappointing. Which tacitly means that the contrarian investor does have an edge in the long run, and we proved that in one of our recent papers entitled, best of my recollection, "Hiring Winners and Firing Losers in Fund Management." Recent Journal of Portfolio Management article. In any event, what we find is that active management can win as long as there's a losing active manager on the other side of your trade, and that's easier than it sounds.

What we find is that indexing itself has an important Achilles' heel. You're weighting stocks in direct proportion to the price of the stock and to its market capitalization. If a stock doubles in price, its weight in the portfolio doubles, all else equal. And why would you want to do that? Why would you want to have twice as much invested in the stock just because it has recently doubled in price? That makes no intuitive sense and makes no investment sense unless the underlying economics of the company and its perspective growth have improved by a like margin.

And obviously, the market is expecting that that expected future growth will have improved by a margin large enough to justify the price rise, but the market is sometimes wrong. At the very top of the list, we did a paper called "Top Dogs" that looked at the largest market capitalization stocks in each sector, in each country, and the worldwide top dog, which today is Microsoft. What we found is that the top dog in any sector--and this is a global phenomenon; we tested it in eight developed economies around the world--the number-one stock in each sector on average underperforms by 5% per year compounded over the next 10 years relative to its own sector.

The number-one top dog in the world underperforms by over 1,000 basis points per annum on average over the next 10 years. Now, does that mean Microsoft is going to underperform by 1,000 basis points a year for the next 10 years? Of course, we can't make a bold assertion of that magnitude. But that is the historic pattern. The other interesting thing is: There have been eight top dogs, eight global top dogs in the last 40 years. And that's one for every five years. How many of those have outperformed the market since they were a top dog? None of them except Microsoft, which is the newest entry on the list.

So, indexing does have an Achilles' heel. And what we find is that any strategy that doesn't weight according to share price, that strips share price or market capitalization out of the formula for weighting stocks in your portfolio, any such strategy will add 1.5% to 2% per year compounded over long periods of time. And we've tested that all over the world. That's the essential profit engine of Fundamental Index. Fundamental Index doesn't win because of the fundamentals, it wins because it contra-trades against the market's most extravagant bets, which are often wrong.

Ptak: Widening out, let's imagine I'm a novice investor, and I'm trying to cut through everything to figure out what really matters to future stock and bond returns. So, can you walk us through how one might have answered that question, let's say, 30 years ago, and how we'd answer it today?

Arnott: Well, for better or worse, I was in the business 30 years ago. And the notion of growth and value was very much part of the investing parlance; the notion of contrarian investing and performance-chasing was already part of the language of investing. And so, 30 years ago, the assertion would have been if you're a value manager, you have an edge on average over time. If you're a contrarian investor, you have an edge on average over time. If you buy what's newly cheap, on average over time, you're going to win, especially if it's been cheap for a long time. The tired cheap stocks that have been out of favor for a long time historically had really demonstrably superior performance.

Those messages are just as applicable today as they were then. Now, what called that whole thesis into question was the tech bubble. Roll the clock forward 10 years. You had said 30 years ago. If you go back 20 years ago, people were saying, "Well, I have some investors say to me, I never want to talk to another value manager as long as I live." We're in the only business in the global macroeconomy, the only major business, where price doesn't seem to matter to the customer--in fact, where the customer likes things when they're more highly priced. Imagine if Tiffany's put out a banner sign saying, "Special pricing 20% up from last year," and people were bashing down the door to get in. Imagine Cartier across the street had a banner sign saying, "All prices marked down 20%," and people thought, "Oh, something's wrong with Cartier. I'm never going to go in there." That's the way people think about investing. And it's deeply embedded in human nature. Anything that's given us, joy and profit, we want more of it. And anything that's given us pain and losses, we want to get rid of it.

So, something that's given us pain and losses, well, there's no such thing as a bargain that hasn't inflicted pain and losses. So, what do I find when I talk to people about asset classes that are out of favor, unloved, or bargains? People will say, "Yes, but … look at this narrative for why it can't come back." Emerging markets today, why would you want to go there? We've got a trade war. The emerging economies are struggling. The politics is unstable.

We're a species that is deeply wedded to the notion of narratives. I've heard it described that narratives are what differentiates humankind from other animals, an interesting thesis. But you don't get a bargain in the absence of fear. You don't get a bargain in the absence of pain and losses inflicted on the way down, and you don't get fear in the absence of a narrative for why things are going to get worse before they get better. In fact, that narrative is often true. Things do get cheaper until they don't. And nobody knows where that turn is. So, ironically, if you're a value investor, if you're a contrarian investor, you can't succeed unless you're willing to look and feel like an idiot for a period of time, until the turn happens. If you're willing to look and feel like an idiot, you will average in gradually and, at the turn, you'll have big exposure. That's the best way to succeed as a contrarian. And it's very difficult. It goes against human nature.

Ptak: So, one bedrock assumption that time has tested is this idea that we get paid for courting risk. So, I wonder if you can talk about where our understanding of that relationship risk versus reward has evolved the most, and what insights that might yield for how selective we should be about the types of risks we elect to take as investors.

Arnott: There's a mountain of theory built on the notion of a risk premium. And what I find interesting is the whole language associated with it. I think the notion of efficient markets is partly attributable to the language of describing it as a risk premium. I think what's going on in the markets is more of a fear premium. Assets that people are afraid to own carry a price to reflect the demand for incremental return. Assets that people are afraid not to own can have a negative risk premium. So, you can actually have individual stocks that have a negative risk premium, not because they're not risky, but because the fear goes the other direction of fear to miss out, and that happens at the top of the spectrum in cap-weighted portfolios with the most beloved and most expensive stocks. People are afraid not to own Facebook or Tesla or … the list goes on.

And so, if you thought of it as a fear premium, if it was characterized from the very beginning in finance theory as a fear premium, all of a sudden, the value effect makes total sense. The size effect makes total sense. Long horizon mean reversion makes total sense. And so, the inefficiencies that have baffled the academic finance community, the anomalies, most of them would be fully expected if it was characterized as a fear premium, not a risk premium. But in fact, we do observe that risk and fear are correlated and so we do observe that risk premium is evident in markets and in individual assets. The finance theory tends to anchor on: You get rewarded for nondiversifiable risk. This was Sharpe's pioneering observation.

And so that means beta. But beta relative to what? In theory, it would be beta relative to the investable global portfolio, and some would argue even the investable global portfolio plus human capital. Now, that's very hard to model. And that's impossible to replicate with an index fund, but you can replicate segments of the market, the U.S. stock market, the global stock market, the global bond market, using capitalization weighting to capture that theoretical construct. Even Bill Sharpe cheerfully acknowledges that his theory is an approximation of reality. It's not reality. It's predicated on an array of assumptions that you can borrow and lend it at the same risk-free rate, no matter who you are. Well, not so sure about that. The notion that all investors have a shared expectation for risk and return for all assets. Hmm. Not sure about that one either. And the list goes on and on.

So, if you acknowledge that the assumptions are false, then the result--the notion that return is correlated with beta and no other risk measures--becomes suspect. And then all of a sudden, the fact that it doesn't correlate with beta, the fact the capital markets line is a lot flatter than capital asset pricing model would predict, suddenly becomes expected. It's unsurprising. And it represents an inefficiency. So, where this becomes beautiful from an investor's perspective is that inefficiencies abound, and the challenge is finding them and exploiting them.

One of the problems that quant community has is assuming that the inefficiencies are static over time. No, they're not. Inefficiencies become discovered. They become crowded space. When they become crowded space, they're arbitraged away. They disappear. So, you have to be aware that inefficiencies themselves change. But look for what's popular and beloved, and avoid it. Look for what's feared and loathed, and embrace it. Some of these are value traps. So, don't put all your eggs in one basket. And some of these take a long time to turn. So, patiently average in. These are very, very simple rules that anyone, whether a financial advisor or an individual investor, can follow. If you hear a lot of cocktail chat about thus and such asset having been absolutely brilliant, don't buy it.

Benz: So, to get into a couple of strategies that perhaps might illustrate what you were just talking about, let's talk about two different strategies that you've had really different expectations for: low volatility and what you define as the high dividend universe. So, you've written that you think the low-volatility universe is expensive and will confer poor returns going forward, but high dividend is inexpensive and might offer a better payoff. Can you walk through the specifics on that, and how you would arrive at those conclusions?

Arnott: Sure. Sure. Let's talk first about low volatility. Low volatility was an extremely popular strategy. It had billions a month pouring into it in 2014, 2015, and early 2016. We wrote a paper entitled "How Can Smart Beta Go Horribly Wrong?" back in early 2016. It was, to my surprise, a massively controversial paper that stirred genuine anger among some of our competitors. And what's funny about that is, let's suppose I'd written a paper entitled, "How Can Stock Picking Go Horribly Wrong?" Let's suppose, I advanced the thesis that if you buy a stock, because it's performed brilliantly, and its underlying fundamentals haven't, so it's gotten more and more expensive over time, then its past performance will look fabulous. And if there's any mean reversion in the valuations back towards historic norms, watch out. Past is not prologue. Those wonderful past returns actually create an environment for terrible future returns. If I'd written that about individual stocks, people would have said, "This guy is an idiot. Everybody knows that." But by writing exactly that message about the popular and beloved factor strategies, and other smart-beta strategies that are out there, I was excoriated for the suggestion that changes in valuation multiples for factors and strategies could indeed create an illusion of historical alpha that is negatively correlated with the future results.

The second paper in that series drilled down and looked at individual factors and strategies and suggested that value was unusually cheap, and that most other factors were unusually expensive, with low volatility being an extreme outlier. That was published in June of 2016. And lo and behold over the next six months, value outperformed by about 4% in just six months, and low-volatility strategies--these are live strategies, live mutual funds--underperformed by an average of 800 to 1,000 basis points in just six months. Where are we today?

Today, value is actually a little cheaper than it was in mid-2016, especially in emerging markets. And all of the other popular factors … Momentum today is saying: Chase those FAANG stocks. Quality is saying: Quality is well in the top quartile of historic valuation. Small-cap stocks are trading a little rich relative to large-cap stocks. And low-vol stocks are trading rich relative to the high-vol stocks, the high-beta stocks. And so, everything is trading a little rich except value, and value is trading cheap. What are people doing? They're saying, "Value doesn't work anymore. Get me out of here." Well, relative valuation would suggest that they should probably do the opposite.

The other thing that's happening is multifactor strategies are on a roll. People are pouring money into them. And we offer a multifactor strategy. So, it's a little awkward for me to say this, but many multifactor strategies are trading very rich relative to the market. What's interesting about that is the pitch for multifactor is: "You're tired of value underperforming; take a look at this strategy. It's got value in it. It's one of the sleeves. You also have momentum, you have low volatility, you have quality, you have small cap. So, you have multiple factors, and guess what? They all work. Academia has proven that they all work. And they work at different times. So, it's going to smooth out your ups and downs."

Now, the basic premise there is: Value is in there as a part of the strategy. No, it's not. It's one of the sleeves, but momentum is inherently antivalue. Quality is inherently antivalue. Today, unlike many times in the past, small-cap and low-beta stocks are trading at premium multiples. They're temporarily antivalue.

What that means is that you've got four antivalue sleeves and one value sleeve. The value is completely wiped out. Most multifactor strategies are trading at a premium to the market. So, if you go from value to multifactor, you're embracing an antivalue, a growth-tilted strategy. Is that what you really want when growth has been on a roll for 10 years?

Ptak: Maybe building on those comments, implicit in them is sort of this notion of making an assessment of a given strategy or a factor, whether it's cheap or dear. And I think that you've made a fairly persuasive case that there are fundamental ways in which one can do that. But not every investor is endowed with the sort of tools and analytics that your firm is... But if somebody were just trying to work with what they had and making an assessment of whether a given factor or a strategy was fundamentally attractive or not, I mean, is there a relatively straightforward, accessible way for them to do so, so that they can build a cohesive portfolio that isn't unduly tilting toward something that's really expensive and maybe prone to decline in value?

Arnott: Well, there's a quantitative way to do it and a qualitative way to do it. The quantitative way requires access to data. You could go look at the reports and see what the relative valuation is for a particular fund relative to the S&P 500. And if the fund has beat the market by 5% a year over the last five years, not that there are many funds that have done that, but if it has and it's now priced at a 50% premium to where it was priced five years ago, watch out. That's a sell, not a buy.

If a strategy has underperformed by 2% per year while getting cheaper by 20 or 30 percentage points, that's a buy, not a sell. And so, you can quantify this. And just as a last check before buying a fund--look at its relative valuation and compare it with the past. Very, very simple thing to do. You need access to the data.

The qualitative way to do this is much simpler. If a strategy is something you're thinking about buying, ask the question: Is it popular, beloved? Are people talking a lot about it, and is money pouring into it? If the recent flow has been massively positive? If so, don't buy it. Look for something that's out of favor.

One of the studies that we did that I thought was just a lot of fun is: One of the best predictors for how a fund will perform over the next three years is how it performed over the last three years with the wrong sign. Meaning, that if you're looking at funds that have been in the bottom decile trailing three-year basis, they tend to, on average, beat the market by a little over 1% a year over the next three years. Well, that's pretty nice. But buying the bottom-decile funds, the funds that have had horrific performance over the last three years, you're going to be buying the funds that have terrible past returns that have massive outflows. The massive outflows are a benefit in the sense that whatever the manager likes to own, they are being forced to sell it. So, they're being forced to sell in order to meet redemptions, the very assets that they think are best. And those assets may be newly cheap now if only because of those outflows. So, that's an interesting, but simple way to do it.

Now, I can't imagine any FA, a financial advisor, listening to this podcast and saying, "Great, I'm going to go to my clients and tell them, I randomly select funds out of the bottom decile over the last three years' performance," but it would work.

Benz: So, you've doubted the U.S. 60% equity, 40% bond portfolio for quite a while. But despite that, it's delivered outstanding risk-adjusted returns for investors. What do you think your original forecasts got wrong, misconstrued? And to what extent do those errors or does that experience inform the way that you've adjusted your forecast for the future?

Arnott: Well, firstly, I'd push back a little bit. I was very wary of U.S. stocks in '07 and '08. And I loved them in early '09, but I also loved a lot of other asset classes that were even more out of favor--high-yield bonds, for instance, and convertibles. So, basically, I've been called a permabear because I've had a cautious view on the U.S. for rather too long. But I'm not a permabear. I'm cautious on things that are fully priced, and I love things that are cheap. And historically, things that are cheap tend to perform better for the patient investor.

Now, the challenge we run into is something known as maverick risk, or that we like to call maverick risk, which is, if you're doing something different from your peers, you may feel very uncomfortable with it. Suppose stocks are up 30%, as they were in 2013, and you're invested in an asset class that's up 10%. You're going to feel like you did terribly when, in fact, a 10% return was just fine.

So, what we find going back historically is: At the tail end of a bull market, when a bull market has been running for a goodly while--and this one's been running over a decade--when a bull market has been running for a long time, people hate diversification, because every time they've ratcheted up their diversification, they've been hurt by it. They take money out of stocks, they put it into diversifiers, and then the stocks go on to new highs. So, one of the lessons from the past is that maverick risk is painful. Doing something different from your peers is painful when the conventional markets are doing exceptionally well.

So, one of the things that I think is really important today is when we look at markets today, the temptation is to say, "Well, let's ride this a little longer." OK, who's going to sound the bell to tell you it's time to get out? If you don't have an answer to that, you need diversifiers. You need assets that fare better when U.S. stocks falter. Much of our work in asset allocation is done using strategies that are deliberately designed to diversify away from mainstream stocks and bonds. And as such, when you're in a roaring bull market for U.S. stocks and when 60-40 is tough to beat, we're going to look like we're Johnny-come-lately--producing modest returns instead of the lofty returns of 60-40. But that's the nature of diversification. The reciprocal is: When 60-40 falters badly, the diversifiers are reasonably likely to add considerable value.

Ptak: On your website, you provide a really impressive raft of tools to help investors unpack and forecast asset-class factor and strategy returns. We recently spoke to James Montier from GMO on this podcast, and he described how they feel markets can be micro-efficient but macro-inefficient, implying that mispricings happen at a broader asset-class level. Assuming there's something to what James and his colleagues at GMO are asserting, why have you made the business decision to give this information away for free? And again, it's laudable that you provide the tools that you do, but just juxtaposing that from some of the previous comments, I'd be curious, your perspectives.

Arnott: Thanks. And those who are listening to the podcast, if you'd Google "asset allocation interactive," you'll be taken straight to the website we have that forecasts asset-class returns. If you'd Google "smart-beta interactive," you'll be taken right to the website that forecasts factor and strategy returns. And these websites are free because we have always been alarmed by people's reliance on past returns for making investment choices. Past returns are in the past. They have nothing to do with what the future has in store. In fact, they are somewhat mildly negatively correlated. What's done well in the past often is what struggles into the future.

And so, what we wanted to do was to put together a website that recognizes that return for any asset class consists of the yield on the investments, plus any growth in earnings and dividends. Or if it's a bond, growth in coupon--well, it's fixed income, so it doesn't grow, but it can have default, so the growth can be negative. So, you have yield, you have growth and income. And the last, the very important piece, is changes in valuation multiples. If a strategy is becoming more expensive, that bolsters the return.

And so, how do you forecast these? Well, the yield is what it is. The growth in income--we use history. We ask the question, How fast have earnings and dividends grown in U.S. stocks? Well, they'd grown about 1%, 1.5% faster than inflation over the last century. So, let's use that as a base-case assumption. Then changes in valuations, which are going to dominate even for a 10-year return, changes in valuations--we just assume that you go halfway back to historic norms. Why halfway? Well, because you might get mean reversion or you might not. Maybe things are different this time. So, if the U.S. market is at 30 times its 10-year smoothed earnings. That's called a Shiller PE ratio. If there's mean reversion, it's going to hurt you. If it's a new normal and it stays at 30 10 years from now, that's fine. You're going to get your yield plus growth. So, if it mean reverts, historic norm is 18 times. Going from 30 times to 18 times and taking a whole decade to get there, that's going to cost you 6% a year.

So, we assume you go halfway, and it costs you just under 3% a year. Well, the yield on U.S. stocks is just under 2%. If you use 1% to 1.5% real growth in earnings and dividends, that gets you to about 3.25% as a real return for U.S. stocks. If you take away 3 from mean reversion back towards historic norms, taking you only halfway back there, you take away 3 and you're left with a real return of zero or 0.25% positive over the next 10 years. That's our return estimate for U.S. stocks.

We did the same kind of thing for factors and strategies. We asked the question, Are they trading rich or cheap relative to history? And what's their historical excess return once you net out the effects of changing valuations? And netting out the effects of changing valuations, what you find is that value works pretty well, momentum works pretty well, quality works OK, low volatility works OK, and the list goes on.

Well, if you take into account that value is cheap and the others aren't, you're going to ratchet up your expectation for value a little bit and ratchet down your expectations for the others a little bit. If you take away trading costs, so each of these have a certain measure of turnover--value and quality have low turnover; low-volatility strategies and momentum strategies have high turnover--take out trading costs, and you are left with a world in which, right now, value is the place to be and other factors look within rounding error of neutral results. So, we're expecting most multifactor strategies to struggle to add value. Value strategies we're looking to perform well.

Benz: You've been bullish on emerging markets for a while, and that raises the question of time horizons. If someone believes that emerging markets will outperform in the future, and they decide to tilt their portfolio toward them and the expectation that they'll earn some positive excess returns over time, how long should they be prepared to hold? What's an appropriate sort of holding period?

Arnott: Well, the holding period varies depending what the markets do. If a strategy has been highly successful, it's probably time to start averaging out of it; if a strategy has been unsuccessful, you probably want to continue averaging into it. We were mildly bullish on emerging-markets stocks in 2014. Became, I'll say, wildly bullish by early 2016. Well, what happened in January of 2016? The Shiller PE ratio, price relative to sustainable 10-year earnings, for emerging-markets stocks fell below 10 times earnings. That's extraordinary. The value stocks in emerging markets were trading at about 7 times earnings. Fundamental index in emerging markets was trading at around 6 times earnings. So, you could get half the world's GDP fundamentally weighted for 6 times earnings. That's extraordinary. That's as cheap as the U.S. was in May of 1932.

So, I looked at that, and I thought, "This is phenomenal." Well, as it happens, RAFI in emerging markets gave over an 80-percentage-point return over the next 12 months. So, it really did pay off handily. As emerging markets then went and struggled again, we're back at a point where valuations are again pretty cheap. So, RAFI in emerging markets is now in the 8.5 range. It's not 6, but it's 8.5. That's cool. That's really very cheap. So, if you average in as markets go against you and take your profits very slowly, let your profits run, but take them slowly, gradually, averaging out, you can actually do pretty well.

Ptak: I wanted to ask you about another strand of research that's not necessarily quantitative in nature, but it's this notion of the paradox of skill. This idea that the skill gradient between professional investors and novice investors--that perhaps it's been narrowing. I suppose we could also plot people on a behavioral continuum of sorts, and perhaps there has also been a narrowing there just as we've become more aware of some of the ticks that we're prone to as investors and human beings for that matter. What do you make of that research? And what if any implications do you think it has on those that maybe would approach investing from a more quantitative factor-oriented perspective?

Arnott: Well, firstly, professional investors are human beings. They are subject to the same errors that any human being would be prone to make. They are subject to performance-chasing. The better institutional investors have trained themselves to not make those mistakes as often or on as larger scale as individual investors. That's the extent of their advantage. As I said earlier, active managers can only win if there is another active manager willing to take the other side of their trade. And one thing that I've never been asked in a finance presentation for an institutional allocation of money that I think should be part of any enquiry into an active manager is, Who is the loser on the other side of your trade, and why are they willing to lose? If you don't have a good answer to that, you shouldn't be in the investment management business, you shouldn't be an active manager.

So, when we look at institutional investors, they make the same performance-chasing mistakes that retail investors make. They just, in general, will do it a little slower, a little bit more carefully and get hurt a little less by it. Russ Kinnel's work on Mind the Gap, I think, has been seminal in this area, and what it generally tends to show is that the dollar-weighted return in mutual funds is a couple of percent lower than the time-weighted return, which means that if you owned all the mutual funds equally, you get a return of x, and if you own them the way the investors do own them, the dollar-weighted average holdings, the return is x minus 2% a year. The results vary from one year to the next as the updates that work, but that's the general picture.

All right. Losing 2% a year is daunting. How does that happen? After a fund has had wonderful results, if they don't pivot and move into newly cheap assets, well, then they're now stuck in newly expensive assets. And what's happening at the retail market? People are noticing the performance. They're pouring money into the strategy. We saw this firsthand. Our asset-allocation funds were seeing inflows that were measured in billions a month back in 2013 at a time when I was telling people that stocks are fully priced, bonds are fully priced, the yields on bonds are terrible, and diversifying markets--which we characterize as third pillar of investing--third-pillar markets were very fully priced, but money was pouring in. Early 2016, when third-pillar markets--the diversifying markets--were really cheap, money was pouring out.

So, it's a phenomenon that we see firsthand as fund managers that when we've had brilliant results, money pours in. I was applauded as a genius in 2012 and 2013 and then three years later, people were wondering how many stupid pills I had taken. All that happened really was that our diversifying asset classes performed brilliantly and then went into a bear market. So, it is interesting watching human behavior, but there's not as much difference between institutional investors and retail investors as most people would like to think.

Benz: I'm trying to remember if you've closed strategies in the past. But can you talk about whether you in addition to sort of warning people that, "Oh, we don't think our near-term expectations for various asset classes are that attractive." Have you taken more proactive measures in terms of trying to kind of shake people off from investing at what you think is maybe not the best possible time?

Arnott: We're unusual. We're probably the only $180 billion asset manager in the world that manages no assets. We license our ideas through others. We subadvise. And so, that kind of decision is made by our distribution partners. We have had strategies close, but it's comparatively rare. One of the things that we do because we're so focused on creating strategies and products that distribution powerhouses, the likes of PIMCO, Invesco, SSGA, BlackRock, Nomura, FTSE, these distribution powerhouses can take our ideas to market and distribute them much better than we can and service the clients much better than we can. If our model is redistribution through others, if these distribution powerhouses do not want illiquid strategies that have to be closed after they hit $1 billion, then we'd better be focusing on highly liquid markets with low turnover strategies, and that's exactly what we do.

Benz: Well, Rob, this has been a fascinating discussion. We really appreciate you taking the time out of your schedule to talk with us. It's been truly great to hear your insights. Thank you for being here.

Arnott: Thank you so much for the invitation. This has been a lot of fun. I appreciate it. And thanks for

Benz: Thank you.

Ptak: Thanks again, Rob.

Benz: Thanks for joining us on The Long View. If you liked what you heard, please subscribe to and rate The Long View from Morningstar on iTunes, Google Play, Spotify, or wherever you get your podcasts. You can follow us on Twitter @Christine_Benz.

Ptak:And @syouth1, which is S-Y-O-U-T-H and the number 1.

Benz: Finally, we'd love to get your feedback. If you have a comment or a guest idea, please email us at Until next time, thanks for joining us.

(Disclaimer: This recording is for informational purposes only and should not be considered investment advice. Opinions expressed are as of the date of recording. Such opinions are subject to change. The views and opinions of guests on this program are not necessarily those of Morningstar, Inc. and its affiliates. Morningstar and its affiliates are not affiliated with this guest or his or her business affiliates unless otherwise stated. Morningstar does not guarantee the accuracy, or the completeness of the data presented herein. Jeff Ptak is an employee of Morningstar Research Services LLC. Morningstar Research Services is a subsidiary of Morningstar, Inc. and is registered with and governed by the U.S. Securities and Exchange Commission. Morningstar Research Services shall not be responsible for any trading decisions, damages or other losses resulting from or related to the information, data analysis or opinions or their use. Past performance is not a guarantee of future results. All investments are subject to investment risk, including possible loss of principal. Individuals should seriously consider if an investment is suitable for them by referencing their own financial position, investment objectives and risk profile before making any investment decisions.)