LIES, DAMN LIES, AND BONDS
Just as they have done in virtually every 40 year period in American history, stocks beat bonds over the last 40 years, and by a clear margin of nearly 1% per year (9.0% vs 8.1%). Because of the effect of compounding, this difference is hardly insignificant: each dollar invested in the Lehman Aggregate Bond Index (or equivalent) at the beginning of 1969 would have grown to $23 by the end of 2008, while a dollar invested in the Standard and Poor's 500 Index of large US stocks would have grown nearly 40% more, to $32. Strangely, this was the SMALLEST margin of victory for stocks since the 40 years beginning 1822 and ending 1861. Over the course of American history, an average 40 years saw a stock investor end up with nearly five times as much money as a bond investor. To find the last 40 calendar years when bonds beat stocks, you have to go back to an investment made at the beginning of 1818, when James Monroe was President, the United States was still an emerging market economy, and the difference was not the result of stocks doing poorly (in real terms, they grew by nearly 5.4% per year) but bonds having an extraordinary 40 years, gaining 5.7% per year after inflation, something they have not achieved since the end of the Civil War.
Imagine my surprise, then, to read headlines beginning in early March saying bonds had beaten stocks over the previous 40 years, and talking about there having been many periods in the past when this had occurred. All of this was advanced hype for an article by Rob Arnott that was finally published for all to see in the May/June 2009 edition of the Journal of Indexes (http://www.indexuniverse.com/publications/journalofindexes/articles/149-may-june-2009/5710-bonds-why-bother.html).
Just as I expected, it turned out that the "bonds beat stocks" message was a misrepresentation of the actual results of Arnott's study. Since I feel it borders on malpractice to recommend bonds for an investor with a multi-decade time horizon, given the enormous shortfall of final wealth that has typically resulted from such a point of view and the immense danger of bond losses over long time periods, I'd like to use this month's essay to challenge the "bonds beat stocks over the past 40 years" baloney.
1. IT ISN'T TRUE
While I have enormous respect for Rob Arnott, and we have been on the same side of some battles (such as the case against weighting investments by the market capitalization of companies), my respect doesn't extend to many of the journalists who received advanced copies of his article and misled their readers. I'd prefer to leave the guilty anonymous and just deal with the facts, although it isn't hard to find many of these articles online. The first thing they did was pretend that BONDS beat STOCKS when, in fact, there was only a single bond, the 20 year United States Treasury Bond, that barely edged out a single stock index, the market capitalization weighted S&P 500. Gee, why can't I use Philip Morris as my definition of STOCKS: it multiplied hundreds of times in those 40 years? See, stocks clobbered bonds!
Actually. at 8.9% per year for the 40 years ended 2008, even the long Treasury bond fell short of stocks, according to the respected Ibbotson database, but Arnott chose to use mid-February 2009 for his comparison, benefiting from a weak six weeks in stocks that allowed the 40-year measure of the long T-bond to barely edge out the S&P 500 for a short period of time. Hopefully you didn't blink, though, because you missed it: the 20-year Treasury bond lost 27% from the beginning of the year through June 10, while stocks gained 5% during that time, so that stocks are back in front by a comfortable margin over the long Treasury.
But the whole thing is silly. The common categories of measurement for bonds include short-term Treasury bills, intermediate-term Treasury notes, long-term Treasury bonds, and corporate bonds, while the common categories of stocks include large capitalization US stocks, small capitalization US stocks, and international stocks. Here are the numbers for the 40 years ended 2008 for all the categories, in order of increasing returns:
T-Bills 5.9%
T-Notes 8.1%
Corporate Bonds 8.5%
T-Bonds 8.9%
Large US Stocks 9.0%
International Stocks 9.5%
Small US Stocks 10.6%
Have you got it, folks? Every category of stocks beat every category of bonds, but if you do a day-by-day search, you can find a tiny window of a few weeks when the very best category of bonds barely exceeded the very worst category of stocks. And that becomes the headline "bonds beat stocks over the last 40 years!"
Now, I've been preaching thorough diversification of equity portfolios since day one. Even if we ignore the beneficial effects of including commodity futures and REITs in the equity category, a global stock portfolio was in no danger of falling below bonds as a whole OR long term Treasury bonds, even when measured on the magical best day in the middle of February 2009.
So the first thing to know about the "bonds beat stocks" story is that it isn't true.
2. IT ISN'T RELEVANT
Even before I discovered that the news wasn't accurate, I was underwhelmed by the headline, because it didn't make a case for buying bonds. It was no accident that these arguments were being made now, as we recently endured a worldwide panic that brought down virtually every investment EXCEPT Treasuries, which might have experienced their biggest bubble in history (at one point, T-bills were being bought for a yield of ZERO). The same 20-year bond that was the centerpiece of Arnott's article and all the excited headlines has since had its biggest drop in history, which will hopefully shut the bond salesmen up soon enough, but at its present yield, it still offers a pathetic return that will only make sense if we experience major deflation. Please let me know when the price of a first class postage stamp goes down.
Any lessons of history need to be based on more than a cherry-picked time period that was extraordinary. Arnott and others have suggested that the great stock returns often cited over the past century were the result of a once-in-a-lifetime expansion of price-earnings ratios: the problem is that the returns in the century before that were virtually identical. Since I dealt with this topic in a commentary that turned out to be at the bottom of the last bear market in 2002, I'll just refer you to it (http://www.andrewtobias.com/bkoldcolumns/020910.html). More broadly, based on historical evidence, there are two plausible views one can take on what the last decades portend for the next:
(a) The stock market is a random walk - If that's the case, then a globally diversified stock portfolio should be expected to clobber bonds for the simple reason that they virtually always clobber bonds. A study of 16 different countries with a full record for the 20th century for both stocks and bonds had stocks winning in every one, and for those who complain about such records not including failed countries, it is worth noting that failed countries have failed monetary systems, and their bonds typically return minus 100% per cent in such circumstances. I wonder how many people know that, while hyperinflation wiped out German bank accounts and bonds in the decade of the 1920s, German stocks rose 75% in real terms, because they represented real businesses creating real wealth.
US Stock returns after inflation have averaged around 6.5% per year for the past 50 years, the past 100 years, the past 150 years, and the entire history of the American stock market. Stock returns in other countries with long stock market histories are close enough to this number to suggest it is the market's reward for ownership, and I can't explain it any more than I can explain 98.6 degrees being the normal body temperature of people throughout the world and apparently through time. Bond returns are less certain in real terms, because they are based on guesses about inflation that may turn out to be too high or too low as well as guesses about the safety of the bond that may be unduly optimistic or pessimistic: the simplest explanation for why long-term bonds did so well that they ALMOST reached the level of stocks over the past 40 years is that inflation expectations were high in 1969 and low in 2008, and for those who are wondering why bonds did so well in the 40 year periods that ended in the 1840s and 1850s, it was because the country had central banking at the start of these periods which was abolished during the Andrew Jackson administration, leading to an era of the freest banking in US history (until Lincoln ended it during the Civil War). I confess that I will be far more optimistic about bonds should the Federal Reserve System be abolished in the same way the central Bank of the United States was abolished in the 1830s, as a return to honest money will make such claims more reliable and free of inflationary expectations. Still, if I make that average 6.5% real return in stocks and have to watch others making 7% in bonds because of a free market revolution, I can live with the disappointment. It is important to remember that, whatever possibility might exist for bonds to beat stocks in rare circumstances, stocks have been profitable in every time period of 20 years or more in American history, and I would expect a global stock portfolio to be at least as reliable in the future, barring a societal collapse that will leave ALL investments in a state of chaos. You can't hedge the world.
(b) The stock market reverts to the mean - There is certainly some evidence to suggest that above average periods for stocks of 3 or more years are more likely than usual to be followed by below average periods, and vice versa. The poor performance of the stock market over the past decade would, in such a case, be a good omen for the upcoming years. Since we know that even the periods that are bad for stocks rarely lead to bonds beating them, facing a stock market that is reverting to its mean probably provides a vanishingly small chance that bonds will be the right place to be in the coming decades. So if stock markets revert, the case for stocks over bonds is even stronger.
3. IT ISN'T SAFE
This is, was, and will always be my main reason for hating bonds. People who say that stocks are risky speak the truth. Certainly, over short periods of time, bonds just don't crash as stocks do (although the 27% decline in Treasury bonds this year shows that we cannot even count on that). But over periods of a decade or more, bonds have proven to be much, much riskier than stocks, especially once returns take inflation into account, and especially for those long term Treasuries. It was a virtual tie, but let's concede that long-term Treasury bonds DID barely beat the S&P 500 over the 40 years ended mid-February or so in 2009. The risk taken for that extraordinarily rare success was monstrous.
Long Treasuries lost money in 21% of all 20-year time periods over the slightly more than 2 centuries of American history for which info is available, with the worst occurring in the period ended 1981: they lost 46% in inflation-adjusted terms, virtually cutting the bond investor's wealth in half. A 30-year T-bond investor ended up with a loss for their troubles in 12% of all such periods: the 30 years ended 1980 saw 45% disappear. Only 8% of T-bond investors lost money over 40 years, but the 40 years ended 1980 destroyed 60% of the bond investor's wealth. And the 20, 30, or 40 year investor in American stocks? Not one loss. Ever. In the worst 40 years, the period ended 1920, the hapless stock investor suffered through a gain of 303%. So $100 dropped to $40 in the worst 40 years for bonds, but rose to $403 in the worst 40 years for stocks.
The returns for 2008 represent the short run: nobody is arguing that stocks are safer than bonds over short periods of time. But US stocks were extremely profitable over the past 40 years: doubling, then doubling again, then again, and again, and yet a fifth time, ending up 32 times as valuable as at the start. 38 times for a global investor. The big success being touted about bonds was that an investor in them wouldn't have lost nearly as much as normal relative to stocks, although they still would have had inferior results, only reaching 23 times the starting value, or 31 for someone insane enough to keep all their money in a single long-term bond whose worst 40 years saw a 60% decline. If that's the case for bonds, you can have them.
This essay is getting too long. I do want to add that my data on the 1800s is not nearly as favorable for bonds as those of Arnott. For instance, he states that it took stocks until 1871 to reach parity with the performance of bonds from a starting date of 1803, while my own data (courtesy of Jeremy Siegel) indicates that stocks returned 5 times as much as bonds during those 68 years, an incredible discrepancy. There is a great deal of debate about the accuracy of all return data prior to 1871, the year the respected Cowles Commission database begins, and if the case for Treasury bonds rests on the returns of an emerging market nation (not to mention the use of speculative railroad and canal bonds in the 1840s and maturities of 10 years rather than 20 in the early data, as Arnott mentions in footnotes), then it is extremely suspect. And it assumes an investor bet on the right side of the War Between the States, as confederate bonds ended up being worthless.
At best, then, there is a possibility that government bonds bought at a time when the founding fathers were still alive and repayment was in gold might have done so well as to be competitive with stocks for a while. But the 5% per year advantage of stocks over fixed income securities which Arnott denigrates in his article is not a myth: it has been the reality AT LEAST since John Wilkes Booth interrupted a performance of My American Cousin at the Ford Theatre in Washington DC nearly 150 years ago, which is a very long time, and while nobody can predict the future with confidence, anyone looking at the lessons of history should recognize that bonds are exposed to an enormous risk of massive losses in periods of rising inflation, and have either never or almost never rewarded the long-suffering investor with a return better than stocks in exchange for that risk.
Stocks are risky. Bonds are risky. Life is risky. Most people are not so wealthy that they can enjoy a secure retirement with the low returns available in bond investments, nor accept the significant possibility that their bond investments will earn negative real returns in their upcoming decades. Are equity investments in businesses providing the world's goods and services sure to do better? No, but as a newspaper columnist once wrote, “The race is not always to the swift nor the battle to the strong—but that’s the way to bet.”
