GREATER risk means greater reward, right? Wrong, at least over the last 25 years. As the graph shows, Treasury bonds have actually outperformed riskier asset classes over the last quarter century. That is despite the long equity bull-market from 1982 to 2000.
Treasury bonds have understandably beaten equities this year, when the financial sector has been in crisis and the economy headed towards recession. The government bond-market always performs well at times of crisis. But the last 25 years have been, by and large, a pretty good time for global economies, marked by the “great moderation” in inflation and growth.
But the last quarter century has been positive for all asset classes, with government debt, corporate bonds and Treasuries all returning an average of around 10% a year in nominal terms. So why have Treasuries done so well in relative terms? The explanation, as Jim Reid of Deutsche Bank explains, lies in reversion to the mean.
Asset classes can go through long periods when they underperform, leaving them cheap and ripe for revaluation. That happened to Treasury bonds, which suffered four consecutive decades of negative real returns from the 1940s through the 1970s. At that point, Paul Volcker, then the Federal Reserve chairman and now an advisor to President-elect Barack Obama, successfully brought down inflation, allowing investors to lock in double-digit Treasury yields. It was one of the great historical buying opportunities.
Such a period is extremely unusual. Since 1900, the average annual return from Treasuries has been 4.6%, or 1.5% after allowing for inflation. In contrast, American equities have delivered 9.3%, or 6% in real terms.
The current poor performance of stockmarkets reflects, of course, a reversion to the mean after the excesses seen during the dotcom bubble, when the rolling 25-year annual return of US equities reached a remarkable 16%. On a 10-year basis, the return from equities has now slumped to minus 3.5% in real terms.
So we may be getting to the stage when reversion to the mean starts working against Treasuries and in favour of equities and corporate debt. For example, Mr Reid calculates that the likely nominal return from American equities over the next 10 years will be 10.6%, more than a percentage point above the historical average. Corporate bonds could earn 9.3%, almost four percentage points above their historical return. In contrast, unless the economy slumps into deflation, Treasuries look likely to provide low or negative returns in real terms.
There are two important caveats. First, this kind of analysis is no use at all in predicting short-term movements. Second, markets spend very little of the time at fair value (on Mr Reid’s calculations), tending to veer wildly from one extreme to another. All three asset classes might still be overvalued; after all, the figures show returns over the last quarter century have been well above average.
Postscript: A couple of months ago, this column featured the stockmarket indicator for the presidential election. If the market rose in September and October, then the incumbent party is usually re-elected; if the market falls, the opposition triumphs. This year, the theory was provided right again with the bears ushering in President-elect Obama.