Money Matters

We’re in bubble territory again

November 13, 2017
By Martin Wolf

Asset prices are, some argue, in an “omnibubble”: prices of every important asset in the high-income countries — stocks, housing and bonds — are at such exalted levels that a devastating crash is inevitable. Moreover, at these prices, prospective returns are too low, which is plainly unfair to wealth owners. It is also clear who is to blame: the central banks. Off with their wretched heads, is the cry.

Is there much merit in these points?

Assets are indeed expensive by historical standards. Consider stocks. The Nobel-Laureate, Robert Shiller, developed the cyclically-adjusted price/earnings ratio, the so-called CAPE, to assess whether stocks are likely to be over- or under-valued. It is possible to invert this measure to obtain a cyclically-adjusted earnings yield which allows one to measure prospective real returns. If one does this, the answer for the US is that the cyclically-adjusted return is now down to 3.4 per cent. The only times it has been still lower were in 1929 and between 1997 and 2001, the two biggest stock market bubbles since 1880. We know now what happened then. Is it going to be different this time?

Yet the US is not the only market in the world, even if it is the most important. An estimate of CAPE for Germany and the UK gives cyclically-adjusted real earnings yields at 5.1 and 6.2 per cent, respectively. While the figure for the US is two-thirds of its average since 1983, that of Germany is at only 89 per cent, while the UK’s is 8 per cent above its average over this period. On this basis, the other two markets are not so highly valued. Japan’s cyclically-adjusted earnings yield of 4.1 per cent is 42 per cent above its average since 1983. So, the US looks the exception, not the rule.

What about that other significant class of assets: housing? Here the story is somewhat different. Real prices in the UK are close to their pre-crisis peaks. US house prices are 29 per cent above their post-crisis low, but also 16 per cent below their pre-crisis peak. In Italy and Spain, real house prices are well below peak levels. So UK house prices look most stratospheric.

This leaves us with bonds, a class of assets that is not only important in itself, but, to some extent, is the anchor for the rest. The crucial point is the long-term decline in real and nominal yields on safe bonds. The real yield on UK index-linked gilts has collapsed progressively, from 4 per cent in the 1980s to negative levels since 2011. Twenty-year US Treasury inflation protected securities are now yielding less than 0.5 per cent. America can also currently borrow for 30 years at nominal yields of 2.8 per cent, the UK at 1.8 per cent, France at 1.7 per cent and Germany at 1.1 per cent. This is very cheap money.

So US stocks look expensive and so do almost everybody’s government bonds. But are they unsustainably expensive?

One way of answering the question is by reference to current conditions. With real interest rates on safe securities so low, asset prices should be high. That is basic economics. Maybe, they should not be as high as they now are. But it is far from obvious they are in extreme bubble territory. The biggest exception, even given current low real interest rates, may well be US stocks.

So the question is whether current conditions — low real interest rates and low and stable inflation — will last.

One perspective is to note that real long-term interest rates have been in an extended period of decline. Furthermore, it is easy to think of long-term — or “secular” — reasons why this should be so. One might refer to a “savings glut” or low productivity growth. Inflation, too, has remained low in the big high-income economies. For these reasons, markets might reasonably expect short- and long-term interest rates to remain low for the indefinite future, even if not as low as today.

Another perspective is to blame low long-term real and nominal rates on central bank manipulation. When monetary policy changes, it is suggested, the bubbles will burst and asset prices duly collapse. Yet this argument is largely unpersuasive. Central banks cannot on their own determine real rates over very long periods. These rates must largely reflect conditions in the real economy.

A final perspective is to insist, nevertheless, that ultra-low real interest rates will have to rise, in the end, even if not right away. That may be true. But nobody knows when. Rates may be not far from present levels for decades. If so, asset prices will still enjoy their support.

What might all this mean for investors and policymakers? The answer for the former is that prospective real returns in many asset classes are modest and could be negative over quite lengthy periods. At the same time it is hard to argue that most assets are vastly too expensive, because it is possible that things are a bit different this time.

Meanwhile, the test for policy is not whether asset prices fall: policymakers are not responsible for delivering any particular level of asset price. The question rather is whether asset prices can adjust without bringing down either financial system or economy. It is on this that policymakers have to be judged.