Recently, I authored an article that placed the year-end 2011 price to trailing twelve months earnings ratio of the S&P 500 (NYSEARCA:SPY) in a historical context. Multiples of price to trailing earnings as low as 2011's ratio have portended higher returns on average over the next one year. While the sample size of annual returns of the S&P 500 is statistically significant (55 observations 1957 - 2011) and appears to have predictive power, researchers at the Federal Reserve Bank of San Francisco have presented some compelling counterevidence against the hypothesis that today's low earnings multiples will lead to above average forward returns that must be weighed by market participants.
A piece published by Zheng Liu and Mark Spiegel in August of 2011 points to a historical relationship between the age distribution of the nation's population and equity market performance. The article makes the supposition that the transition of the large Baby Boomer generation (born 1946-1964) from the workforce to retirement will necessitate a switch from equities to fixed income instruments that could place downward pressure on equity multiples.
The authors' model is remarkable for both its compelling relationship and its simplicity. Their examination of the relationship between demographics and equity prices studied the time series of the inflation adjusted S&P 500 index divided by trailing earnings versus a ratio of middle-aged (M) Americans (aged 40-49, peak earnings) and an older (O) cohort (aged 60-69, transitioning to retirement). The economists estimated that the ratio of middle aged Americans to older Americans explains 61% of the movements in the price to earnings ratio.
In the authors' analysis, earnings multiples should have reached their peak in 2001 as the ratio of the middle aged population relative to the older population signaled the bulk of the baby boomer generation was peaking in relative economic importance. The period of 1999 - 2001 did see the S&P 500 trade at record multiples. The measure of middle aged Americans to older Americans troughed in 1981, the last year that saw the market trade at an average single digit earnings multiple.
Given the model's ability to time the earnings multiple peak and trough with reasonable precision, it is important to examine what their model predicts prospectively. The ratio of middle aged Americans to older Americans declines each year before bottoming in 2022, well below its 1981 local minima. The nominal increase in stock prices over this period is projected at a meager 25%. Assuming a three percent inflation rate, the next decade could see a 13% decline in real equity prices.
While this study indicates that earnings multiples will decline over the next decade, hampering equity returns, there are plenty of additional counterarguments to the predictive power of this model. Notably, Baby Boomers are entering retirement against the specter of historically low Treasury yields, and long-term bond yields at their largest gap to equity earnings yields in over thirty years. The rotation from stock investment to bond investments presaged by the Liu-Spiegel model could be muted or forestalled due to the current relative valuation of the two equity classes. The lengthening of life expectancies and the reduction in private and public pensions could further smooth out the rotation from equities to fixed income investments as well.
The paper discusses foreign buying of U.S. equities as a potential buoy offsetting the demographic headwinds domestically, but partially discounts this counterbalancing effect due to investor's home bias, or desire to hold stocks in their own country above and beyond what is rational given the benefits of international diversification. Perhaps reduced capital controls in some emerging markets, improved international information dissemination (e.g. Seeking Alpha), or simply better investor awareness of their investment biases could strengthen foreign demand for U.S. equities. Demographic headwinds are not unique to the United States and also permeate the developed economies of Europe and are exacerbated in Japan. On a relative basis, U.S. economic performance and the depth of our capital markets may make the U.S. a relatively superior destination for capital flows as compared to other economies facing similar demographic trends.
Of course, the only certainty in forecasting asset returns over a long horizon is uncertainty. I hope this article and the work of Liu and Spiegel help elucidate a bear case for markets to be weighed against your own baseline.
Disclosure: I am long SPY.