What is the average decline in the bear market

By Robert J. Shiller

The American stock market is currently in a phase of strong earnings growth and very little volatility. Does that mean the United States is moving in the direction of a "bear market," that is, steadily falling prices? To answer this question, one needs to look back at the past and define what constitutes a bear market.

Today's media describe a "classic" or "traditional" bear market as a 20 percent decline in stock prices. Where this definition comes from is unclear; possibly it can be traced back to October 19, 1987, when the stock market plummeted a little more than 20 percent in one day. In any case, this 20 percent mark is now widely recognized as an indication of a bear market.

There seems to be less consensus on the timeframe within which this decline will take place. No period was previously mentioned in newspaper reports; Journalists who have written on the subject apparently thought that it did not detract from its accuracy.

In researching America's bear markets, I used the traditional 20 percent mark and added my own time frame. By my definition, the peak before a bear market was the last 12-month high of a year, before falling 20 percent in any month the following year. If there was a continuous series of peak months, I used the last one for my investigation.

Using monthly data from the S&P Composite and similar information, I found that the United States has had only 13 bear markets since 1871. The peak months prior to these bear markets occurred in 1892, 1895, 1902, 1906, 1916, 1929, 1934, 1937, 1946, 1961, 1987, 2000, and 2007. Some notorious stock market crashes - 1968 to 1970 and 1973/74 - are not on the list because they were longer and gradual developments.

After identifying the bear markets, it was time to pre-evaluate the stock valuations, using an indicator my Harvard colleague John Y. Campbell and I developed in 1988 to predict long-term stock price gains. This cyclically adjusted price / earnings ratio (cape) is determined by dividing the real (inflation-adjusted) stock index by the average profit over ten years, with cape values ​​above the average meaning below average returns. Our research showed that the Cape value is a reasonably effective means of predicting real gains over a ten-year period.

A look into the past might be reassuring, but it's not that simple

This October, the US Cape is just over 30. That is a high number. In fact, the average Cape value between 1881 and today was only 16.8. In addition, it exceeded the value of 30 only twice during this period: in 1929 and in the years 1997 to 2002. However, that does not mean that high Cape values ​​are not associated with bear markets. On the contrary: In the peak months leading up to past bear markets, the Cape average value of 22.1 was higher than the overall average, which indicates that this value tends to increase before a bear market. In addition, the three sub-Cape bear markets occurred in 1916 (during World War I), 1934 (during the Great Depression) and 1946 (during the post-World War II recession). Thus, a high Cape value implies a potential bear market vulnerability, although it is by no means a perfect predictor.

But there seems to be some promising information out there. According to my data, real S&P Composite earnings have increased an average of 1.8 percent annually since 1881. In contrast, earnings growth from the second quarter of 2016 to the second quarter of 2017 was 13.2 percent, which is considerably higher than the historical annual value.

However, this strong growth does not reduce the likelihood of a bear market. In fact, the peak months leading up to past bear markets also tended to show high real earnings growth: an average of 13.3 percent annually for all 13 bear markets. In addition, real profit growth at the market peak before the largest price fall ever recorded between 1929 and 1932 was 18.3 percent.

Another supposedly good news is that the average share price volatility - calculated by determining the standard deviation of the monthly percentage changes in real share prices in the previous year - is an extremely low 1.2 percent. Between 1872 and 2017, the volatility was almost three times as high at 3.5 percent.

But even that does not mean that a bear market is not approaching. In fact, stock price volatility was below average for the year preceding the peak months leading up to the 13 US bear markets of the past, although the current value is lower than the 3.1 percent average for those periods. In the peak month before the crash of 1929, volatility was just 2.8 percent.

Bottom Line: Today's US stock market is very reminiscent of the peak periods leading up to most of the 13 bear markets of the past. That is not to say that a bear market is certain to come - such developments are difficult to predict and the next bear market may be a long time coming. And even if a bear market does occur, those who do not buy at peak times and sell at lows tend to lose less than 20 percent.

However, my analysis is intended to serve as a warning. Investors who take excessive equity risk today because of misperceptions about history could suffer significant losses.