From these data, I created a Valuation Confidence Index, which is the percent of respondents who think the domestic stock market is not overvalued; a Crash Confidence Index, which is the percent who think
that a 1929- or 1987-style crash in the next six months is highly unlikely; and a One-Year Confidence Index, which is the percent who think the stock market will go up in the next year.
The CAPE ratio has successfully explained about a third of the variation in real 10-year stock market returns in United States history since 1881.
The high fraction of our survey respondents who think
that the stock market is unlikely to fall in the next year may simply reflect a failure of imagination about how a Trump bull market could suddenly end.
The current level of CAPE suggests a dim outlook for the American stock market over the next 10 years or so,
but it does not tell us for sure nor does it say when to expect a decline.
In years when CAPE was lower than that, subsequent 10-year returns for the stock market tended to be good.
In 1988, John Y. Campbell (now at Harvard) and I showed in a joint article
that such a ratio has, since 1881, forecast returns somewhat well in the stock market.
But in both cases, during the initial election campaigns the economy was in recession
and the CAPE ratio was extremely low — around 5 in 1920 and 9 in 1980.
One possible interpretation might be that respondents perceive a stock market bubble: They think valuations are high
and there is a non-negligible probability of a crash.
This is a bit technical: It is real, or inflation-adjusted, stock price divided by a 10-year average of real earnings.