There is no stock market bubble

Martin Wolf: It all depends on earnings and interest rates

Are stock markets, especially the US market, in a bubble that is sure to pop? The answer depends on prospects for corporate earnings and interest rates. Provided the former are strong and the latter ultra-low, stock prices look reasonable.

The best-known measure of market value – the "cyclically adjusted price/earnings ratio" of Yale's Nobel laureate, Robert Shiller – is indeed flashing red. One can invert this metric, to show the yield: on the S&P Composite index, this is just 3 per cent today. The only years since 1880 it has been even lower were 1929 and 1999-2000. We all know what happened then.

Another price is also exceptionally low by past levels: interest rates. The short-term nominal interest rate is near zero in the US and other high-income economies. US short-term real interest rates are about minus 1 per cent. Real yields on US 10-year Treasury-inflation-protected securities are minus 1 per cent. In the UK, yields on similar securities are about minus 3 per cent.

Desired returns on equities ought to be related to the returns on such supposedly safe assets. This relationship is known as "the equity risk premium", which is the excess return sought on equities over the expected returns on government debt. This premium cannot be measured directly, since it only exists in investors' minds. But it can be inferred from past experience, as explained in a 2015 paper by Fernando Duarte and Carlo Rosa for the New York Federal Reserve. More recently, in the Credit Suisse Global Investment Returns Yearbook 2020, Elroy Dimson, Paul Marsh and Mike Staunton of the London Business School estimated the excess return on world stocks over bonds at 3.2 percentage points between 1900 and 2020. For the UK, the excess is estimated at 3.6 percentage points; for the US, at 4.4 percentage points.


Are these excess returns in line with what people initially expected? We do not know. But they are a starting point. The premium demanded now might be lower than that sought for much of the past 120 years. Corporate accounting has improved greatly. So, too, has macroeconomic stability – at least by the wretched standards of the first half of the 20th century. Moreover, the ability to hold diversified portfolios is far greater now. Such changes suggest the risk premium, often believed to be excessive, should have fallen.

Real returns

The Credit Suisse study estimates aggregate real returns on stocks and bonds in 23 markets weighted by market capitalisation at the start of each year. It shows, interestingly, that the excess return of equities since 1970 have been very low and since 1990 negative. But this is because of very high real returns on bonds, as inflation and real interest rates collapsed. Looking ahead, it estimates the prospective excess return of equities at 3.3 percentage points. This is the same as the long-run historical average.

Estimates of Shiller’s metric do not exist for such lengthy periods for non-US stock markets. But estimates can be made since the early 2000s. The cyclically adjusted earnings yield is currently 7.6 per cent on the FTSE 100, 5.4 per cent on the Dax 30 and 4 per cent on the Nikkei 225. At current real interest rates on long-term bonds, the implied equity return premium is thus over 10 percentage points in the UK, over 7 percentage points in Germany and 4 percentage points in Japan and the US. The UK market looks extremely cheap today, perhaps because of the Brexit lunacy. Japan and the US look well valued, but not, by historical standards, overvalued.

Further support for the rationality of the US market today is that 55 per cent of the increase in the S&P 500’s market value over the past 12 months is due to gains in the information and technology sector. This makes sense, given US dominance in these areas and the technological shift of 2020. We should also note that real interest rates below zero make future profits more valuable than profits today, in terms of present value. Looking through the short-term impact of Covid-19 makes sense.

Given the interest rates, then, stock markets are not overvalued. The big questions are whether real interest rates will jump, and how soon.

Many believe that ultra-low real rates are the product of loose monetary policies over decades. Yet, if that were right, we would expect to see high inflation by now.

Better hypothesis

A better hypothesis is that there have been big structural shifts in global savings and investment. Indeed, Lukasz Rachel of the Bank of England and Lawrence Summers of Harvard argued in Brookings Papers 2019 that real economic forces have lowered the private sector's neutral real interest rate by 7 percentage points since the 1970s.

Will these structural, decades-long trends towards ultra-low real interest rates reverse? The answer has to be that real interest rates are more likely to rise than fall still further. If so, long-term bonds will be a terrible investment. But it also depends on why real interest rates rise. If they were to do so as a product of higher investment and faster growth, strong corporate earnings might offset the impact of the higher real interest rates on stock prices. If, however, savings rates were to fall, perhaps because of ageing, there would be no such offset, and stock prices might become significantly overvalued.

Some major stock markets, notably the UK’s, do look cheap today. Even US stock prices look reasonable, valued against the returns on safer assets. So will the forces that have made real interest rates negative dissipate and, if so, how soon? These are the big questions. The answers will shape the future. – Copyright The Financial Times Limited 2020