The higher a stock soars, the more likely it is to plunge.
The U.S. stock market is likely to be a below-average performer over the next 24 months, according to an obscure sentiment index - as are 25 high-flying, popular U.S. stocks including Nvidia $(NVDA)$, Super Micro Computer $(SMCI)$ and Broadcom $(AVGO)$.
The market indicator, known as the "U.S. Crash Confidence Index," was created 40 years ago by Yale University's Robert Shiller and is now maintained by the Yale School of Management. Each month a group of individual and institutional investors are asked, "What do you think is the probability of a catastrophic stock market crash in the U. S.?" As with other sentiment measures, this one has contrarian significance: It's a bad sign when investors believe the probability of a crash is particularly low.
It's worth noting that Yale reports the index in terms of the percentage of respondents who think the probability of a crash is less than 10%, so higher readings in the chart mean investors think the likelihood of a crash is lower - and vice versa. The six-month moving average of the institutional-investor index recently reached its highest level of the past 15 years, and the individual-investor index reached its second-highest level.
I analyzed the data back to 2001, which is when Yale began updating the index on a monthly basis. On average, the monthly readings are inversely correlated with the S&P 500's SPX total return over the subsequent 24 months - meaning that higher index readings are correlated with below-average returns, and vice versa. The correlation is significant at the 95% confidence level for the institutional index and the 93% confidence level for the individual index.
Why the Crash Confidence Index is a contrarian indicator
The U.S. Crash Confidence Index is a valuable contrarian indicator because stock-market investors become more or less bullish for reasons totally unrelated to the stock market. Since irrational bullishness (as opposed to rational bullishness) will cause stocks on average to be overvalued, it follows that the market tends to struggle in the wake of a widespread belief that a crash is unlikely.
The role of emotions in the U.S. Crash Confidence Index has been documented in a new study by Shiller, Yale professor William Goetzmann, and Dasol Kim of the U.S. Treasury Department's Office of Financial Research. Entitled "Emotions and Subjective Crash Beliefs," the study began circulating in June in academic circles.
The researchers found that a survey respondent's belief in a crash's probability was heavily influenced by "rare, extreme events" having nothing to do with the stock market. When an earthquake had recently occurred in the respondent's immediate vicinity, for example, belief in a stock-market crash was more likely than for another respondent who had not experienced a nearby earthquake. Similarly, if a winning lottery ticket had been sold from a nearby store, the respondent was less likely to believe a crash was probable than another person for whom that was not the case.
Emotions and beliefs
The probability of a stock's crash increases along with the extent to which trailing two-year return is ahead of the S&P 500.
The contrarian indicator based on the Crash Confidence Index is not itself forecasting that a crash will occur, but simply that the next two years will experience below-average returns. The index reflects investors' emotions and subjective beliefs, which are far different than the objective factors that have a statistically significant ability to forecast a heightened probability of a crash. As I discussed in a recent column, a recent study of those objective factors implies that the odds of a market-wide crash are below average right now.
But certain individual stocks are still vulnerable to crashing, defined as a 40% drop over the next two years. A recent study found that the probability of a stock's crash increases along with the extent to which the trailing two-year return is ahead of the S&P 500. When this margin of outperformance is 100 percentage points, for example, the probability of a subsequent crash is close to 50%. When it's 150 points or more, a crash becomes nearly certain.
The table below lists the 25 stocks in the S&P 1500 index that have outperformed the S&P 500 over the past two years by more than 200 percentage points, according to FactSet data. Not all of them will crash, of course, but it's a good bet that an above-average number of them will lose at least 40% over the next two years.