Are US equities undervalued? It seems like an odd question, given that the cyclically adjusted price/earnings ratio is now so high: at 29.7 it is more than 50 per cent above its post-1949 average of 19.4. From another perspective, however, the market is indeed cheap.
There’s a huge problem with the idea that the S&P 500 is overvalued merely because the cyclically adjusted price/earnings ratio is high. It’s that this ratio has been above its long-term average since 2010 and yet US share prices have more than doubled during this time. An apparently expensive market has become even more expensive.
Of course, this could be because of irrational exuberance. One fact, however, speaks against this. It’s that one measure of investor sentiment – which has for years predicted equity returns well – tells us that investors haven’t been unusually exuberant recently. It is that non-US investors have been record net sellers of US shares in the past 12 months. That’s a sign of pessimism.
Instead, there’s another reason for the high cyclically-adjusted price/earnings ratio. It’s that this is justified by the extraordinary ability of the US economy to deliver non-inflationary growth.
My chart below shows the point. It shows the so-called misery index, which is simply the sum of consumer price index (CPI) inflation and unemployment rates. Since 1970, this has been massively well correlated with the cyclically adjusted earnings yield*. A high misery index in 1974, 1980 and 2008 was accompanied by high earnings yields – low equity valuations – and the low misery index in the late 1990s saw record-low earnings yields.
Common sense gives us several reasons why this should be. The same strong economy that gives us low unemployment also encourages investors to buy risky assets. Also, low inflation causes low real interest rates, which raises share prices by reducing the discount rate applied to future earnings. And, thirdly, because the Fed has little need to raise interest rates to choke off inflation, investors have more faith that growth is sustainable, which supports earnings expectations and hence share prices.
The misery index is now at its lowest level since 1956: CPI inflation is just 1.8 per cent and the unemployment rate just 3.6 per cent, near a 50-year low. Past relationships therefore tell us that earnings yields should be low, and the cyclically adjusted PE ratio therefore high.
In fact, the post-1970 relationship between the misery index and the earnings yield suggests that equities are actually cheap. It predicts that the cyclically-adjusted PE ratio should be 36.8, not the 29.7 it now is, which implies that shares are almost 20 per cent undervalued.
Which raises the question: why might this inference be wrong?
One possibility is that the correlation could break down. The fact that it did not exist before around 1970 warns us of this possibility, although I suspect that lack of correlation was for reasons that don’t apply today**.
Another possibility is that the gap between the misery index and equity valuations might be closed not by share prices rising but by the misery index doing so. It would only require a rise in it of 0.7 percentage points to be consistent with the market’s current cyclically adjusted PE. This would only take the index back to October’s levels.
Granted, neither the Federal Reserve nor the economists surveyed by the Philadelphia Fed expect such a rise: they foresee unemployment and inflation more or less flatlining well into next year.
But, of course, economic forecasts are not wholly accurate. In particular, they fail to see recessions coming. Investors should not rely upon them.
But this is not my point. Instead it is that while US equity valuations do look stretched by historic standards, they are in fact justified by economic conditions. What investors have to fear (other than shorter-term noise) is not that the market is overvalued, but that these conditions will deteriorate. Such a worsening is of course inevitable sometime – economic cycles have not been abolished. The only question is: when?
*It is also highly correlated with the unadjusted earnings yield and with the dividend yield.
**This is because investors didn’t worry much about inflation before the late 1960s because history suggested it was only a temporary phenomenon. And vivid memories of the 1929 crash and subsequent depression meant that investors were scared to hold equities in the 1950s even though economic conditions were great.
