For over a century, the relationship between stock prices and corporate earnings has intrigued investors and economists alike. The analysis conducted by Robert Shiller reveals that, over extended periods, these two factors often move in tandem, indicating a robust connection that can help explain market dynamics. This article delves into the nuances of this relationship, exploring whether variations in earnings-price correlation can serve as indicators for future stock market performance.
One key objective of this analysis is to provide a framework for understanding stock market behavior over long investment horizons, particularly those exceeding 10 years. Such insights are vital for effective retirement planning and strategic asset allocation decisions. Additionally, the investigation into correlation changes over time seeks to establish whether these fluctuations can predict subsequent market returns.
Table of Contents:
The correlation of earnings and stock prices
To conduct this analysis, monthly averages of the S&P Composite earnings per share alongside their corresponding stock prices were examined, utilizing data ranging from 1871 to December . The consistently high correlations observed across multiple time frames suggest a persistent relationship between corporate earnings and stock prices.
Several specific periods were analyzed, including the decade following the 1940 Investors Act, which aimed to enhance investor protections and establish standardized accounting practices. Interestingly, the results showed minimal variation in earnings-price relationships before and after these regulations were implemented.
Fluctuations in correlation over time
Despite the overall strong correlation, variations do occur, particularly within shorter time frames such as five, ten, and twenty years. The analysis found that during the first half of the 20th century, the rolling 50-year correlation dipped to a low of 0.6, coinciding with significant global events like the two world wars and the Great Depression. This context is essential in understanding the historical market fluctuations.
As the time frame shortens, the variability of these correlations increases. For instance, the rolling 20-year correlation fell below 0.50 for an entire decade from February 1918 to December 1928, with similar drops occurring in December 1948. This pattern of volatility is further pronounced in the rolling 10-year correlations, which even dipped below zero during tumultuous periods such as the aftermath of both world wars and the inflation crisis of the late 1970s to early 1980s.
Examining predictive capabilities
To determine if the shifts in earnings-price correlation could predict stock market returns, regression analyses were conducted. Remarkably, the R² value between S&P Composite earnings and prices from 1871 to stood at an impressive 0.95, highlighting the strong long-term relationship. However, when examining the R² values for correlation changes against subsequent annualized returns across various rolling windows, the results were less promising.
For the rolling 10-year and five-year periods, the R² values approached zero, indicating a lack of predictive capability in these shorter horizons. In contrast, the rolling 50-year window displayed the strongest correlation at 0.53, albeit still offering limited insight. As the time frames decreased, the R² values continued to reflect increased variability, with the 20-year window at 0.24 and the 10-year window dropping to just 0.06.
Ultimately, the rolling five-year periods demonstrated no consistent correlation, with the R² value nearly reaching zero. This outcome suggests that while earnings are a significant factor in explaining long-term market behavior, they do not provide reliable guidance for market timing.
Conclusion and implications
In summary, the comprehensive analysis of over 150 years of data affirms that earnings and stock prices maintain a strong correlation over extended periods. However, the findings also indicate that fluctuations in this correlation do not serve as effective predictors for future returns, particularly over shorter investment horizons.
For financial advisors and investors alike, understanding this relationship can aid in framing long-term market behaviors for clients. While earnings play a crucial role in elucidating stock price movements, they should not be relied upon to time the market accurately.
