Q&A with Robert D. Arnott
This prominent financial analyst contends that the ‘cult of equities’ has obscured for many the importance of bonds, which can outperform equities for extended periods of time.
By Robert Stowe England
April 18, 2009
Highlights of Arnott’s Study and the Q&A Below
In an article published in the Journal of Indexes, Rob Arnott reports two key findings from his fresh review of the historical performance data comparing returns on stocks with yields on 20-year Treasuries over the very, very long term, from 1803 to February 2009. The article is titled “Bonds: Why Bother?” and is available on line at this link:
The first significant finding is that there are long periods during which bonds outperform stocks: namely, 1803-1871, 1929-1949 and the period from 1968 to February 2009.
The second notable finding is that while stocks still outperform bonds over the entire review period of 1803 to February 2009, the equity risk premium (see footnote 1) is only half of what people currently accept as the norm: namely, a 2.5 percent premium instead of a 5 percent premium. Indeed, Arnott says, “the much-vaunted 5 percent risk premium for stocks is at best unreliable and is probably little more than an urban legend of the finance community!”
Even so, the equity risk premium, even at 2.5 percent, is not shabby. As Arnott notes in the article, the return on stocks from 1803 to February 2009 would give an investor 4 million times his original investment after 207 years. At the same time an investment in bonds would have returned 27,000 times the original investment. But, the periods in which stocks underperform can be very costly. For example, from the peak in 2000 to year-end 2008, the equity investor lost nearly three-fourths of his wealth, relative to the investor who invested in long-term Treasuries. That could mean eating “cat food instead of caviar” for hapless investors entering their retirement years with 80 percent in stocks in 2007, Arnott quips.
Arnott’s findings suggest that investors should rethink a lot of assumptions about how they invest. Stocks should still be part of any portfolio, but investors need to be sure the price they pay for stocks is sufficiently low so that the dividend yield and the potential from gains in earnings and dividend growth is attractive.
While current stock prices make them generally attractive investments based on dividend yields and anticipated returns, comparisons with bond yields today is complicated by the wide divergence of bond yields. They range from the very low yields of U.S. Treasuries on one hand, and the high yields of investment-grade corporate bonds, convertible bonds and high-yield bonds on the other.
Indeed, while the stock of a given company may be attractive, the yield on the same company’s corporate bonds could be even more attractive, Arnott points out. Further, investors need to differentiate across broad classes. Within the stock market, the deep value stocks are more attractive than growth stocks even though the deep value stocks have risen 25 percent from their lows.
Arnott, born in 1954, is the chairman of Research Affiliates, LLC, of Pasadena, California and served as editor of the CFA Institute’s Financial Analysts Journal from 2002 to 2006. His writing and editing has focused on articles about quantitative investing. He has authored over 100 peer-reviewed articles for financial publications, including Financial Analysts Journal, the Journal of Portfolio Management and the Harvard Business Review. Arnott is on the product advisory board of the Chicago Mercantile Exchange and the Chicago Board of Options Exchange. He was previously chairman of First Quadrant, LP; a global equity strategist at Salomon (now part of Citigroup); and president of TSA Capital Management (now TSA/Analytic). He has also served as Visiting Professor of Finance at the University of California at Los Angeles. He graduated from the University of California at Santa Barbara in 1977. Mr. England caught up with Mr. Arnott at his Pasadena office early Friday afternoon, April 17, 2009.
ENGLAND: I first read about your forthcoming article in Barron’s a few weeks ago. I was fascinated by what was reported. Then Jane Bryant Quinn mentioned it in a column in the Washington Post just about a week ago. So, kudos to you for generating some excitement about the article ahead of its publication date.
ARNOTT: Thank you.
ENGLAND: Can you explain this high level of interest?
ARNOTT: Well, it plows some interesting territory, that’s for sure, so it’s been an exciting project.
ENGLAND: There are two thoughts in this paper that seem to me to be new. One is that bonds outperform stocks for significant periods of time. The other is that the equity premium is only 2 ½ percent instead of 5 percent.
ARNOTT: Correct. That’s exactly right. Basically the thing I have issue with is the widely-held view that stocks should always be our core investment regardless of what you should pay for them. That’s the whole stocks-for-the-long-run thesis. I don’t take issue with the notion that stocks have a risk premium. And I certainly don’t want people to infer from the article that because stocks lost to bonds over 40 years ended February that they are going to lose over the next 40 years. Far from it. But, it’s a sobering reminder if you don’t pay attention to what you pay when you first get in, stocks-for-the-long-run might be [something for] your great grand-kids, not something that you can enjoy. Ultimately the main purpose of the article is to draw people’s attention to the fact that price matters, that stocks are not automatically the best place to invest; that [instead] they’re the best place to invest when they are sensibly priced.
ENGLAND. This flies in the face of what people have been saying for so long. It’s so hard for investors to digest that immediately.
ARNOTT: I think that’s right. And the reason that it’s so hard for people to swallow the idea is that it has been drummed into their heads again and again and again that if you are patient, the stocks will win. And that is simply not always true.
ENGLAND: So, what does this say about some other investment strategies, such as constantly buying stocks a bit at a time because you never know how to time the market? How should you invest? Should you look for a benchmark, like price to equity? How do you know when to buy a stock?
ARNOTT: I think people need to pay attention to yield. I think they need to compare yield on stocks to yields available on bonds. If the yields available on stocks are materially below the yields on bonds, they need to ask hard questions about how plausible it is that earnings and dividend growth can make up the difference. For instance today [April 17], if you look at the stock market yielding three and long bonds yielding three and a half, well what kind of growth do you need to beat the bonds? You need a half of a percent growth. How plausible is that? It’s more than plausible. It’s pretty easy. But, if you compare stocks with their own bonds. Investment-grade [corporate] bonds yield about seven and so [the equities] have to have four percent growth to beat their own bonds.
ENGLAND: Four percent being the difference between the yield on stocks and the company’s own corporate bonds.
ARNOTT: Yes, because that’s the yield difference. What does history tell us about that? Well, if you go back over the last century, average earnings and dividend growth has been four and a half percent. That’s not much of a premium to reward you for risk-bearing. So, I think right now, stocks are utterly compelling relative to Treasuries. They weren’t 40 years ago. That’s why bonds won for the next 40 years [with a higher return over stocks from 1968 to February 2009.] And stocks are not so compelling relative to their own corporate bonds.
ENGLAND: This means you will be throwing a lot of other conventional wisdom out the door.
ENGLAND: For example, people believe today that individuals should be determining their asset allocation based on age. As they approach retirement, they are supposed to transfer more and more wealth from stocks to bonds. Your findings would do what to the age-related investment theories?
ARNOTT: It would rejigger them. Certainly one should have more preference for stocks when they are young and have that long horizon ahead of them and more preference for bonds when they are old and can’t bear the downside risk. I feel really horrible about retirees who might have been duped into putting 80 percent of their money in stocks two years ago. And now have to have cat food instead of caviar.
I think it’s still true that the longer your [time] horizon the more risk tolerant you should be. This does change that picture a little bit, in that risk tolerance doesn’t necessarily mean willingness to buy regardless of price. I think those who bought stocks in preference to long Treasuries at the top of the [high-tech] bubble in the year 2000 will not live long enough to see their stocks ahead of those bonds. And I know that’s a pretty powerful statement. They’ve already lost 75 percent of their wealth relative to those who put it in bonds.
ENGLAND: You could make an even stronger statement in that regard if they had put their money into high tech stocks instead of the broad stock market.
ARNOTT: Oh, absolutely. Absolutely.
ENGLAND: A person could invest in stock at any point if they thought the return was worth it and it is priced right.
ARNOTT: Exactly. The article was misconstrued by some that people shouldn’t like stocks. That’s not the point at all. The point is that stocks are great if you buy them when they are cheap and they are terrible if you buy them when they are expensive. Right now you have the bizarre situation where bonds are both the most overpriced and the most underpriced assets, depending on which part of the bond market you’re looking at. High yield bonds are priced for Armageddon. So, if you think Armageddon isn’t coming, high yield is a bargain. At the other end of the spectrum, Treasuries are priced at extraordinarily low yields. So, unless you think there’s no risk of inflation in the years ahead, Treasuries are a terrible investment.
ENGLAND: So, an investor who wants to be in bonds is facing a strange dichotomy. Could both stocks and bonds be right as a good investment?
ARNOTT: Oh, absolutely. You could find that 10 years ahead that your best performing major asset class is high-yield bonds and your worst performing is Treasury bonds and stocks are somewhere in the middle.
ENGLAND: Unlike some comparisons of data, you look back beyond the Depression back to 1803. Why is that people have been looking only at the period since the Great Depression to compare stocks to bond and to come up with a 5 percent equity premium?
ARNOTT: Roger Ibbotson and Rex Sinquefield did an enormous project in the 1970s, assembling long-term numbers back to 1926 (see footnote 2). That has been the Bible of the investment community for gauging long-term investment expectations ever since. And, of course they’ve updated their data ever since on a regular basis. Now, the problem is most folks have generally been too lazy to explore other data. Bob Shiller (see footnote 3) went and gathered the data from the Cowles Commission (see footnote 4), the 1930s commission that looked at long-term returns, back to 1872, attached that data to the Ibbotson data, and wound up giving us a 130-year history. That was wonderful. Then Bill Schwert (see footnote 5) at the University of Rochester dug deep into old data from the 1800s.
So, people used the data that is easiest for them to get hold of. And, for a long time the data that was easy to get hold of was 1926 to date. The data that is easy to get hold of is also U.S. [data.] Remember World War I and World War II were for the most part not fought on U.S. soil.
The U.S. has what is statistically referred to as survivorship bias. It did not have the experience of German, Italian and Japanese stock and bond market devastation in World Wars I and II. It didn’t have the nationalization of all capital markets that occurred in Russia and China and Argentina and Egypt in the last century.
In the U.S., we've had the benign experience of seeing how markets perform during a period of political stability for our country. And people use that data as if these markets are entirely normal. To some extent they are normal. Political stability has been, and remains the norm in the U.S. and probably will remain the norm. We shouldn’t ignore the fact that our data has survivorship data and that it covers a very benign span in which yields were generally falling, boosting stock market returns over the course of many decades and boosting the perceived advantage of holding stocks.
ENGLAND: So you have the bias in the modern period, but what did the 19th century U.S. data bring to the picture and why was the period from 1803 to 1871, when bonds outperformed stocks for 69 years, so different?
ARNOTT: That was an environment in which equities were largely unregulated and whoever held the majority stake in the stock could scoop the primary benefits. Dividends were paid out just sufficient to keep the minority shareholders content, but leading to returns that were not competitive with bonds. An interesting result in the early 1800s is that U.S. went through a difficult period of time from the War of 1812, several depressions and then to the Civil War. Ultimately wars are usually the result not of religion or ideology but of economics. The economic depression we went through in the 1840s and 1850s sowed the seeds for the Civil War and savaged the equity investor. Note also that our data does not include CSA [Confederate States of America] assets, which went to zero.
ENGLAND. The bonds you were using for comparison in the early 19th century was still the 20-year Treasuries?
ARNOTT: You know the bond data back in the 1800s was terribly scattered. There was even a period of time during the Andrew Jackson (see footnote 6) Administration when there were no government bonds when the U.S. national debt was zero. That’s something people overlook when we are accustomed to seeing multi-trillion-dollar national debt. But we had no national debt during the latter part of [Andrew Jackson’s] Administration. So, what we used was historical yields for long-government bonds, when they were available, and railroad and canal bonds, which were generally considered the ultimate blue chip bonds, when there were no Treasuries. And we converted those yields into the equivalent of a 20-year bond and a total return history as if there were 20-year bonds available at the time.
ENGLAND: And this was something you did?
ARNOTT: Yes. So, I would say that the data before 1871 has to be taken with something of a grain of salt. It is not terribly reliable data, but it wouldn’t be more than a percent or two per year off from what would have been achieved if you actually had the historical data base on the bonds that we’re looking at.
ENGLAND: And I guess you relied on the work of the people you mentioned who built up a record of the historical data.
ARNOTT: Exactly. Schwartz’s work has been vetted by lots of folks and is considered to be the definitive pre-1871 history for stocks and bonds. And it also mirrors the results that you see in the U.K. if you go back that far. In the U.K., after the Napoleonic wars, they had national debt totaling 250 percent of [gross domestic product]. They had a crippling debt burden. And, as a consequence of that, the economy struggled for decades to pay that down and the yields on government bonds were juicy.
ENGLAND: It was hard for stocks to beat that.
ARNOTT: Exactly. And so, back in the 1800s you had parallel experience in the U.K., which adds additional support to the legitimacy of these results.
ENGLAND: The other period that is interesting to me, at least, is 1968 to February 2009, which is . . . .
ARNOTT: 41 years of bonds beating stocks.
ENGLAND: Right. And during that time there was a huge variance in the comparative performance of stocks and bonds, but still at the end you were better off if you had been in bonds.
ARNOTT: Not only better off with bonds, but better off with plain vanilla long Treasuries, just about the dullest long-bonds you can imagine.
ENGLAND: For any body that is at any period in their lifetime of accumulating assets, this has to be something that should grab their attention.
ARNOTT: You’d think so; because we have had such an equity-centric view of the world that I think it’s fair to characterize it as a "cult of equities." People believed in stocks as a core investment without examining that belief at all.
ENGLAND: Just about every thing we do is built on that assumption.
ARNOTT: Right. And it’s just an assumption. It’s almost a religious view: Stocks are good. Stocks are powerful. Stocks are right. If you roll back the clock to the peak of the bubble in 2000, stocks were yielding one percent and inflation-linked government bonds were yielding four [percent]. Now, for stocks to beat those bonds, stocks would have had to have had earnings and dividend growth of at least 3 percent a year above inflation. There’s no historical support for that. The only time stocks have seen earnings and dividend growth faster than 3 percent real [growth] is coming off of deep Depression bottoms, not off the peak of a bubble. That’s why I say people who bought stocks in preference to bonds in the year 2000 probably won’t live long enough to see their stocks win.
ENGLAND: As I mentioned, this should complicate the decisions people make when they decide how to allocate or re-allocate their assets. As this point, what is the exercise they should do? Should they do as you suggest; that is, look at what the stock yield is compared to stock, just for a generic asset class decision?
ARNOTT: Yes, I think that’s right. The comparison I like to make on the Treasury side is with TIPS [which are Treasury Inflation-Protected Securities](see footnote 7) where the coupon rate grows with inflation. So, if stocks yield more than TIPS, that will typically mean that stocks are more attractive than Treasuries because earnings and dividends tend to grow with inflation plus one to two percent.
ENGLAND: Stocks should be compared to 20-year TIPS?
ARNOTT: Yes, I’d say so. So, 20-year TIPS right now yield 2.2 percent. Stocks yield 2.8 percent. That would mean that stocks are a little more attractive than Treasury bonds.
ENGLAND: You could proceed from there to make your investment decisions, even if you are just an index investor.
ARNOTT: Yes, except there is an added complication in that bonds themselves are less of a monolithic asset class than they have ever been historically, because the yield spreads are wider than they’ve ever been in history. Convertible spreads and high-yield spreads never reached these levels, even in the Great Depression. Which, in turn, means that when you compare stocks with their own bonds, you have to ask the question: Will the company see earnings and dividend growth fast enough to make up the difference? There the picture is much more ambiguous and in many cases I think the answer will be no.
I’ve heard the expression ‘bonds are the new stocks’ and in a very real sense they are because bonds, especially on the high-yield side are bets on the survival of the company, as are stocks – and give you the reward for that survival in form of yield instead of in the form of capital appreciation. The yields are double digit. Well, that gets to be pretty interesting.
ENGLAND: It’s hard to argue with double digit yields.
ARNOTT: Unless you’re expecting massive defaults every year for the coming six or eight years.
ENGLAND: Which could also possibly be a worry for some companies, while not necessarily on a massive scale?
ENGLAND: For the average investors, this means picking the right bond fund rather than trying to pick a specific corporate, convertible or high-yield bond.
ARNOTT: Right. From that perspective, I rather like investment grade corporate bond funds, convertible bond funds, and high-yield bonds more than I like broad stock market portfolios. One exception I would make is that the deep value end of the stock market is cheap.
ENGLAND. The deep value end is attracting a lot of attention, especially the financial stocks.
ARNOTT: Right, exactly. The deep value stocks were hammered beyond recognition and even though they have rebounded sharply in the last three weeks, I think they still represent the low hanging fruit within stocks. And I think growth stocks are darned expensive and very vulnerable. Health care is one example. The health care stocks have performed very nicely as if the nationalization of the U.S. health system isn’t going to affect their profitability.
ENGLAND: We don’t know what’s going to happen with health care and there is the risk of any number of outcomes.
ARNOTT: That’s exactly right.
ENGLAND: Thanks for taking time to talk about your article.
ARNOTT: You’re welcome. All the best.
END OF INTERVIEW
Copyright 2009 by Robert Stowe England
1. In the article “Bonds: Why Bother?” Robert D. Arnott states that he uses the term “risk premium” in the article advisedly. The paper states the following in a footnote: The “risk premium” is the forward-looking difference in expected returns. Differences in observed, realized returns should more properly be called “excess return.” Many people in the finance community use “risk premium” for both purposes, which creates a serious risk of confusion. I use the term here, wrongly but deliberately, to draw attention to the fact that the much-vaunted 5 percent risk premium for stocks is at best unreliable and is probably little more than an urban legend in the finance community!
2. Roger G. Ibbotson and Rex A. Sinquefield, “Stocks, Bonds, Bills, and Inflation: Year-By-Year Historical Returns (1926-1974),” The Journal of Business 49, No. 1 (January 1976), pp. 11-47.
3. Robert J. Shiller, professor of economics, Yale University.
4. Alfred Cowles, Common Stock Indices, Principia Press, Bloomington, 1939.
5. G. William Schwert, professor of finance and statistics and the University of Rochester’s William E. Simon Graduate of Business Administration.
6. Andrew Jackson was the seventh President of the United States from 1829 to 1837. He is the only president in U. S. history to have paid off the national debt. However, this accomplishment was short lived. A severe depression from 1837 to 1844 caused a ten-fold increase in national debt within its first year.
7. Treasury Inflation-Protected Securities or TIPS provide investors protection against inflation. The principal of a TIPS increase or decreases with inflation or deflation, as measured by the Consumer Price Index. When the TIPS reaches maturity, the investor is paid the adjusted principal or original principle, whichever is greater. TIPS also pay interest two times a year at a fixed rate. Interest payments are paid against the adjusted principal so that interest payments rise with inflation and fall with deflation. You can find more information about TIPS from TreasuryDirect® at this link: http://www.savingsbonds.gov/indiv/products/prod_tips_glance.htm.
Copyright 2009© by Robert Stowe England