From the FRBSF:
Demographic patterns have a strong historical relationship with equity values in the United States (Liu and Spiegel 2011). In particular, the ratio of those people who are the prime age to invest in stocks to those who are the prime age to sell has historically served as a strong predictor of U.S. equity values as measured by price/earnings (P/E) ratios.
Research suggests one reason for this close relationship is a person’s life-cycle pattern of investing. An individual’s financial needs and attitudes toward risk change over the years. As retirement approaches, individuals become less willing to tolerate investment risks, so they begin to sell off stocks. Thus, the aging of the baby boomers and the broader shift of age distribution in the population should have a negative effect on capital markets (Abel 2001). In theory, global demographic changes may further impact U.S. equity values. For example, Krueger and Ludwig (2007) demonstrate that U.S. returns can import the adverse impact of population aging in other countries.
Since the study by Liu and Spiegel (2011), U.S. stock values have increased markedly. Between 2010, which is the end of their sample, and 2013, the Standard & Poor’s (S&P) 500 Index has increased by 47% and the P/E ratio has increased from around 15 to nearly 17. However, the bearish predictions in Liu and Spiegel (2011), which were based solely on projected aging of the U.S. population, have worsened. Indeed, extending the Liu-Spiegel model’s sample through 2013 suggests that the P/E ratio will decline even more, from about 17 in 2013 to 8.23 in 2025, before recovering to 9.14 in 2030.
In this Economic Letter, we investigate the implications of global population trends for U.S. stock values. Other industrial countries are expected to have similarly aging populations in the next two decades, in some cases even more so than in the United States. Given that equity markets are integrated across industrialized countries (see, for example, Chan et al. 1992), one would think that this change in foreign conditions might also have adverse implications for U.S. stock values. However, our results suggest that the tight historical relationships in the U.S. data are not mirrored in other countries. Thus, global aging is unlikely to add to the demographic headwinds facing the U.S. market.
Demographic trends in industrialized countries
Following Liu and Spiegel (2011), we use Bloomberg’s P/E ratio for the United States, which is the ratio of the end-of-year S&P 500 Index levels and the average earnings per share over the previous 12 months. We measure the age distribution using the ratio of “middle-age” people between 40 and 49 years—the group most likely to buy stocks—to those in the “old-age” group from 60 to 69 years—the prime age to sell. We call this measure the M/O ratio. Liu and Spiegel (2011) showed that this measure of the age distribution has been highly correlated with U.S. P/E ratios, outperforming alternative demographic measures, such as the ratio of middle-age to young adults studied by Geanakopolous et al. (2004).
We extend the Liu-Spiegel study by examining the historical correlations between the M/O ratios and the P/E ratios in other industrialized countries, specifically those in the Group of Seven (G-7). For these countries—Canada, France, Germany, Italy, Japan, and the United Kingdom—we construct the M/O ratios using United Nations (UN) population data and P/E ratios from Global Financial Data.
Figure 1 shows the M/O ratios in G-7 countries from 1954 to 2010, extended to 2013 for the U.S. sample. The figure also shows the projected M/O ratios through 2030 based on the UN population projections. The M/O ratios in most G-7 countries peaked by the mid-2000s and are expected to decline through at least the mid-2020s. For several countries, the declines are expected to be even larger than in the United States, which is projected to decline from 0.76 in 2013 to 0.60 in 2024. For example, the Canadian M/O is projected to decline from 0.82 in 2010 to 0.53 in 2024, and the German M/O is expected to decline from 0.90 to 0.48.
A notable exception is Japan, where the M/O ratio is expected to increase over the next 10 years, driven primarily by expected declines in the size of its old-age cohort. However, this surprising trend is limited to the specific demographic indicator we use in this study. If we broadened the indicator to the ratio of middle-aged to the combined middle- and old-aged, Japan’s pattern would also decline. Moreover, beyond 2024, population aging in Japan is expected to drop rapidly regardless of which demographic indicator is used.
Global demographics and equity values
In theory, the rapid aging of the global population is likely to have additional adverse implications for U.S. stock values. Ang and Maddaloni (2005) find that using the fraction of retired people in the population predicts excess returns in the four largest equity markets outside the United States. Evidence also suggests that U.S. and foreign markets are integrated. This implies that if a tight relationship exists between the M/O ratio and P/E ratios in foreign economies, their demographics are likely to impact U.S. equity values as well.
We examine the specification used in Liu and Spiegel (2011) for the remainder of the G-7 countries. In particular, we estimate the statistical relationship between the log of the P/E ratio and the log of the M/O ratio for each country.