The dynamics between corporate earnings and stock prices have always piqued the interest of investors and analysts alike. Historical data spanning over a century, particularly from the renowned economist Robert Shiller, reveals that these two elements are inherently linked over extended periods. This article delves into the nature of this relationship and examines the extent to which fluctuations in earnings correlate with stock market returns.
Understanding this correlation is vital for investors, especially when planning for milestones such as retirement.
A long-term investment horizon, generally considered to be over ten years, provides a clearer perspective on market behavior, particularly regarding asset allocation decisions. Let’s explore this relationship in detail.
Table of Contents:
The correlation between earnings and stock prices
Analyzing data from the S&P Composite Index, which includes earnings-per-share and stock price averages, the findings reveal a consistently high correlation between earnings and prices. An examination of various timeframes confirms this trend:
- Overall dataset (01/1871 – 12/): 0.977
- Last 100 years (01/1925 – 12/): 0.974
- Post-1940 Investors Act (08/1940 – 04/): 0.973
- Last 50 years (01/1975 – 12/): 0.963
These figures illustrate a robust relationship that has persisted, with only slight differences observed after the implementation of investor protections and standardized accounting practices introduced in the 1940s.
Fluctuations in correlation
It’s important to note that the correlation between earnings and stock prices does not remain static. Over shorter timeframes, such as five, ten, and twenty years, we observe notable fluctuations. For example, the lowest recorded correlation in the rolling 50-year average occurred during the first half of the 20th century, where it dipped to 0.6 amidst global turmoil, including two world wars and the Great Depression.
As the analysis shifts to shorter periods, variability becomes more pronounced. The rolling 20-year correlations dropped below 0.50 for an entire decade from 1918 to 1928, with additional declines observed in late 1948. The ten-year rolling correlations even dipped below zero during significant historical events: the aftermath of World War I, the conclusion of World War II, and the inflation crisis of the late 1970s and early 1980s. The five-year periods exhibited the most instability, demonstrating frequent fluctuations and deeper declines, reflecting the immediate economic climate.
Predictive capabilities of correlation changes
A critical question arises: do variations in the earnings-price correlation serve as reliable indicators for predicting future stock market returns? To assess this, regression analyses were conducted to compare correlation levels against subsequent annualized returns.
The correlation coefficient (R²) between earnings and prices from 1871 to stands impressively at 0.95, highlighting the strong long-term relationship between these factors. However, when examining how fluctuations in correlation might predict future returns, the results were less promising. The R² value for rolling ten-year and five-year windows approached zero, suggesting negligible predictive capacity. In contrast, the rolling fifty-year period displayed a slightly stronger connection, with an R² of 0.53.
Implications for market timing
The findings suggest that while earnings have a significant role in explaining long-term market movements, they do not provide a solid foundation for market timing. The evidence indicates that other factors beyond the earnings-price relationship influence changes in annualized returns, particularly in the shorter term. Investors seeking to time the market based on correlation shifts may find little success.
Conclusion
In summary, the historical data underscores a strong correlation between earnings and stock prices over long periods. The insights gleaned from this analysis can help investors better understand market behavior, particularly in the context of long-term investment strategies. However, the lack of predictive power in correlation changes emphasizes the need for investors to consider a broader range of factors when planning their investment approaches.
