As we are all aware, the US stock market has experienced the longest bull market in its history. The result has been historically expensive equity valuation multiples, such as the cyclically adjusted price-to-earnings (CAPE) ratio,1 which had topped 30 prior to the abrupt market crash. The CAPE reached this level only twice before in history, in 1929, and during the 1999–2000 tech bubble. These high valuations are correlated with low subsequent returns as valuations revert to lower historical averages as stock prices fall, driven by the reduced risk appetites of investors.
Although the coronavirus caught everyone by surprise as the catalyst of the market’s downturn, we have had low expectations for US equity performance for quite some time. Over the last five years, our 10-year real return (net of inflation) expectation for US equities has hovered around 1% a year.
Our models, of course, do not include a pandemic factor. They are based on market fundamentals that can be observed and on an assumption that certain market variables, such as valuation multiples, cannot rise or fall forever; in other words, we believe that the largest and most persistent investment opportunity is long-horizon mean reversion.
For long-term forecasting, taking advantage of these principles makes good economic sense, but our expected return forecasts also come with a warning label: Long-term expected returns, unto themselves, are not sufficient for short-term decision making. Ignoring this warning will most likely lead to impaired wealth.
Ten-year return forecasts offer valuable guidance to a buy-and-hold investor about the return they are likely to earn over the next decade. They provide no information, however, about when to buy or sell and do not identify a market top or bottom.
When a market corrects dramatically, say, 30%, long-term expected returns do not rise by the same 30%. To illustrate this point, let’s use the CAPE for the US market. The CAPE uses an average of real earnings per share over a rolling 10-year window, so we can assume the realized real EPS at the end of December 2019 is close to today’s value.2