Most investors believe that only by accepting greater risk can you achieve higher returns. This was the basis for the capital asset pricing model (CAPM) taught as gospel to two generations of finance students. The theory made perfect sense; a rational investor should only accept greater risk if the market will, on average, compensate him for the uncertainty; and it does make a good rule of thumb. Given enough time, stocks return more than bonds, corporate bonds pay more than Treasury bonds, and long-term bonds earn more than short-term notes.
As with any rule of thumb, the interesting cases are the ones where it breaks down. Here we start to see that the market is not governed by mechanical rules. For instance, the smallest stocks should in theory outperform the largest, precisely because so many small companies fail. The winners should more than make up for the losers, but actually microcap stocks are by far the worst performing segment by market cap. No successful boiler room operator believes in the efficient markets hypothesis. He knows that suckers will overpay for what are essentially lottery tickets.
The theory also breaks down if you try to apply it at the portfolio level. A portfolio of stocks from around the world, for instance, is inherently less risky than a geographically concentrated one, even if the expected long-term return is the same. Any one country can suffer from a longer and deeper bear market than an index of several countries. Witness the 30 years of stagnation that Japan has experienced, or the era from the 1970s to the late 1990s, when foreign markets outperformed the US.
This brings us to a wider point – much of what investors trust to be true, if it is true at all, is only borne out in the very long term. Let’s look at bonds versus stocks, and see how long you have to wait to be assured of higher returns from stocks.
If you had invested in stocks starting in 1930, you would have had to wait two decades to pull ahead of bonds, which delivered much smoother returns.
Ok, that was cheating – of course bonds beat stocks during the Great Depression. But what if I told you that bonds beat stocks from 1985 to 2012? Where was the promised equity risk premium then? Stocks may have been a great buy in the ‘80s, but long-term bonds were perhaps an even better one. By the way, it would only take a 30% decline in stocks from here to put bonds ahead again, three decades on.
While we’re at it, let’s take a closer look at bonds. There’s something funny in the chart above – government guaranteed Treasuries seem to be beating corporate bonds over this stretch, even though we know corporates pay better interest. It seems that even a third of a century is not enough time for the theory to bear out (although over a full 100 years, corporates do come out ahead).
You might object that our corporate bond index has a somewhat shorter average duration than the 30-year Treasury. Interest rate risk comes with a premium too, after all. In the next chart we’ll examine that theory.
Over a full century, the relationship is plain – the steady, can’t-lose Treasury bills don’t deliver the same return as bonds. But aside from college endowments, who invests for a century? We can see that as rates rose in the 1970s, bills actually caught up to bonds after 60 years. Remember that bond prices fall as rates rise, but that the investor who holds bills can keep up by reinvesting at higher rates. Looking at this chart, it doesn’t take much imagination to picture the gap between the two lines narrowing again in some inflationary future.
As we can see, it’s not only at the portfolio level where the theory of risk and return breaks down, and we haven’t even discussed more sophisticated risk management techniques that tamper the downside without sacrificing upside. Clearly, when accepting more risk in the hope of higher returns, the long-term may be longer than we assume. In real life, where investment horizons are finite, the consistency of returns is at least as important as the value of the portfolio at some distant date. Downside volatility has real costs, such as falling short of intermediate-term goals, or crying uncle and bailing out in a drawdown, thus failing to meet long-term goals.
Diversification delivers consistency
Portfolios that use a wide spectrum of asset classes produce more consistent returns. Any one asset class, be it stocks, bonds, real estate, or even cash equivalents like bills and savings accounts can (especially after inflation), lose value for a decade or more. Taken as a group, however, it is hard to find even a 5-year period in the last century when such a portfolio would have lost value.
In the chart on the next page you can see the inflation-adjusted history of two rudimentary global portfolios in comparison to the S&P 500 and a bond portfolio. Using data that goes back to the 1920s, we tested an equal-weighted mix of US largecap stocks, smallcap stocks, foreign stocks, Treasury bonds, corporate bonds, and gold (unfortunately, REITs only came to the fore in the early 1970s, so we have excluded them here).
Rules for the momentum portfolio: when an asset (US largecap stocks, smallcap stocks, foreign stocks, Treasury bonds, corporate bonds, and gold) closes above its 12-month simple moving average (i.e., has positive momentum), the portfolio is invested in that asset for the next month. When it closes below the SMA (i.e., has negative momentum) the portfolio invests in 10-year Treasury notes until such time as the asset regains positive momentum.
As you can see, stocks have had the highest long-term return, but there have been long stretches when they have gone nowhere. Ironically, it has been during those stretches of low returns when their volatility has been highest. Also note that bonds are not always a good refuge at such times, as in the 1970s and early 1980s. The clear and simple answer to the problem of uncertainty is diversification, and we can see that even these rudimentary global portfolios have chugged right along through the decades. Clearly, diversification is the solution when what happens in the next 10 years is as important as what happens in the next 40.