William J. Bernstein
Stocks? For the Long Run?
"We don't invest where we can't drink the water." -annual report, Tweedy Browne
Most of you have heard the apocryphal story of the financier who prayed all of his life for a peek, just a little peek, at the next day's Wall Street Journal. Finally granted his wish, the first item he comes across is . . . . his own obituary. (Probably the best variation on this theme is "A Special Story," told in the inimical style of Barton Biggs. You can find it in Classics, an Investor's Anthology, ed. Charles D. Ellis.)
Well, if you made me that offer I wouldn't want tomorrow's WSJ. Nope, I would have too much trouble finding enough leverage to exploit the juiciest 24 hour option play, and besides, I'd get an industrial grade case of sweaty palms. What I'd really want is a fast look at a copy of that venerable rag dated about 5 years after I plan to retire. The real gold mine (Remember, the dude in the white turban gives me only a brief look.) lies in the long term returns from the world's major national stock, bond, and commodity markets.
Problem is, until relatively recently we didn't even know the answer for past global stock returns. Oh, sure, the return of the EAFE and the S&P from 1969 were both about 12.5%. However, you may recall that shortly prior the human saga had a few military and economic bumps. Some cynics have even suggested that, like history in general, financial history is also written by the winners. Long term equity returns of 10-13%? Maybe if you were lucky to be living in the right place at the right time, but not in Lima, Delhi, or Budapest. In fact, like the detritus of some ancient terrestrial asteroid encounter, there was a mass extinction of entire equity markets from the 1930s to the 1960s. Stocks for the long run? Not if you lived someplace where they took Karl Marx more seriously than Groucho Marx. Not if you found yourself in the immediate proximity of a self proclaimed military and racial genius.
Since the founding of the New York Stock Exchange under a buttonwood tree in 1792, inflation adjusted total returns of domestic equity have been in the 6%-7% range. However, not until this century did this fact become known, and even then not popularly appreciated until a decade or so ago. Consider the implications of a 7% real return. If you really could invest $1 for 200 years at a 7% real rate, then you have a real $752,000 after 200 years. Invest $1 at a 7% real rate at the birth of Christ and you will have real $6 x 1058. Does that seem like a lot of money to you? It is - it's the value of a solid gold sphere 43 light years in radius.
Clearly, something is fishy here. Returns of global equity (or even debt) simply cannot be that high. A recently published article by Philippe Jorion (UC Irvine) and William Goetzmann (Yale) , as well as some data from Bryan Taylor (Global Financial Data) shed some much needed light on the topic.
It turns out that the good old USA was the winner in the global equity sweepstakes in this century. Unfortunately, many of the other horses ran badly behind, and some even broke their legs and were put down. Heard much about the Cairo Stock Exchange lately? In the 1920s it was one of the world's largest. It was done in not by war or revolution, but by a colonel who should have paid better attention in Econ 101 at U. Egypt.
The article does have a "Picture Worth a Thousand Words." Here it is:
At first glance, things look downright ghastly - the very best returns are no higher than 4%, and many markets seem to have negative returns. In reality, things aren't that grim. It has to be realized that the y axis plots inflation adjusted, capital only, returns. In other words, inflation is already adjusted into the prices, and dividends are not included. (This is because for most of the markets reliable dividend information was not available for the whole period.) The authors found about a 4% long term dividend rate for those markets for which reliable information was available, so in reality real total rates of return were positive for all but the hard luck cases - the Philippines, Poland, Columbia, Argentina, Peru, and Greece. Even here, nominal total returns in US dollars were positive.
Just as important, there is a modestly positive slope to the data - the longer a market has been around, the higher tend to be its returns. This is a good demonstration of "survivorship bias," a prominent characteristic of mutual funds and rock musicians - the bad ones are quickly taken out and shot, so the ones that are left give an overly favorable representation of the genre.
The authors immediately point out that this data presents an "equity risk premium puzzle" - i.e., why are investors in some nations not rewarded for bearing the risk of stock ownership? In classical Ibbotsonian terms the "equity risk premium" is defined as the excess return of stocks over t-bills. Since short duration bond returns can be difficult to come by in many markets, long bond rates may be used for comparison instead. The answer, according to Bryan Taylor of Global Financial Data, is that in nations with low stock returns, bond returns were even worse. So even in the nations with low stock returns there is, indeed, an equity risk premium. Hence no puzzle. This provides scant comfort to emerging markets bond investors; it's well to recall that for nearly a century Latin American nations defaulted on sovereign debt with near clock like regularity.
But so much for academic quibbling. What does this data mean to the average Josephine, scanning the library copy of Forbes for the Honor Roll list of foreign mutual funds? That all depends on two issues, more philosophical than financial:
1. Is history progressive or cyclical? The cataclysmic events of this century devastated entire nations, races, and social systems. Were they a singular occurrence, the likes of which we shall never again encounter? Or were they merely a depressing human commonplace, an intrinsic vicissitude of the species? (My favorite quote from the paper: "Had the outcome of the Second World War been different . . . . the beta of the U.S. on the world index would likely have been different." Yep, suffering a global cataclysm is bad enough, but even worse, it really screws up your portfolio parameters.)
2. Even a cursory look at the above graph demonstrates that the winners and losers segregate on the basis of their avoidance of the twin scourges of Armageddon and Marx. What is the likelihood that the century's winners (the U.S., Canada, Sweden, and Switzerland) will also escape the next march of folly? For those not current on the state of the art in strategic weaponry, the correct answer is "not very."
As to portfolio risk from the next generation of Marxist star gazers, the issues are more subtle. In 1900 who would have predicted the state takeover of the means of production in much of the globe? The key point here is that at the beginning of the century investors, blissfully unaware of the havoc Juan Peron and Gammal Nasser would wreak on their nations, probably demanded very similar risk premiums from the US, Egypt, and Argentina. These particular horses are now long out of the barn; the risks of investing where you dare not drink the water are baldly obvious. Which is why valuations in Asia and Latin America in general are currently about half of what they are in the US and Europe. In other words, the manifest risks of emerging markets investing are already priced into the markets in a way they were not at the turn of the century. It is thus quite likely that the emerging markets investor will be appropriately compensated in the coming decades. There's no Midas Muffler guarantee on this one of course. The compensation of risk with reward in the capital markets is true only in a statistical sense, and the lot of the individual financial statistic is often disagreeable.
It also has to be admitted that like generals fighting the last war, we tend to prepare ourselves for troubles which do not occur, and are woefully unprepared for those which do. It's quite likely that the forces which obliterate nations and capital in the next century will look nothing like those of the past.
Finally, Jorion and Goetzmann evaluate the long term returns of a global investment strategy by putting together an index of all of their national indexes, weighted by GDP. Even accounting for the markets which vaporized, the return of this global index was 4.04%, versus 4.32% for the US for the 1921-96 study period. (Remember, this is inflation adjusted, without dividends included.) While the authors correctly point out that the 0.28% gap makes a very big difference when compounded over 76 years, they also demonstrate that their global portfolio was a good deal less risky, with a standard deviation of returns of just 11.05%, versus 15.8% for the US market. Even more interesting, the non-US index, including the deadsters, returned 3.39% with an even lower SD (risk) of 9.96%.
Those of you familiar with these pages know what's coming -- if one applies standard optimization techniques to the Jorion/Goetzmann data, does foreign equity belong in an efficient portfolio for the 1921-96 period? Does it pay to rebalance assets with a 1% return difference compounded over 76 years? To answer this, I assumed a 1% domestic stock advantage, a 0.5 correlation between US and foreign markets, and a zero return/zero SD for t-bills, and fed the data into MvoPlus, a multiperiod optimizer produced by Efficient Solutions. A screenshot of the output is reproduced below:
The above plot shows the "efficient frontier" for these 3 assets; in the low risk region (SD < 0.1) the preferred foreign/domestic ratio was about 70/30. Only in the high risk/high return upper right corner are US stocks more strongly preferred. Secondly, and somewhat surprisingly, you were better off not rebalancing annually in the low risk/return (SD<0.1) region, but better off rebalancing in the high return/risk (SD>0.1) region. In the 60/40 stock/bond world that most of us inhabit, it really doesn't matter.
This data will be widely discussed in the coming years. Many will look at the first graph and conclude that it's better to stay at home. And, as I hope I've shown, they'd be wrong. The fairest appraisal of the data comes from the authors themselves - "Based on these long term series, the main benefit of going international appears to be risk reduction rather than increased returns." Amen.