The relationship between stock prices and corporate earnings has been a topic of interest for investors and financial analysts for over 150 years. Research conducted by Robert Shiller has revealed a strong correlation between the two variables, which has significant implications for understanding long-term market trends. This article delves into the nuances of this relationship, examining how fluctuations in earnings can impact stock market performance and the limitations of relying solely on correlation for predicting future returns.
Table of Contents:
The significance of earnings in market analysis
Understanding the long-term connection between stock prices and earnings is essential for investors, particularly those with a time horizon exceeding a decade. This extended timeframe is critical for retirement planning and making informed asset allocation decisions. By analyzing historical data, we can better grasp the factors that influence market behavior over prolonged periods.
Assessing the correlation
To investigate this relationship, we utilized monthly averages of the S&P Composite earnings per share alongside the corresponding stock prices from 1871 to. The data consistently demonstrated a high level of correlation across various timeframes.
For instance, the correlation coefficient for the entire dataset stands at a remarkable 0.977, while specific periods such as the last 100 years yield a coefficient of 0.974. Even after the implementation of the 1940 Investors Act, designed to enhance investor protection and standardize accounting practices, the correlation remained strong at 0.973. This indicates that regardless of regulatory changes, the relationship between earnings and stock prices has persisted.
The variability of correlation over time
While the correlation between earnings and stock prices is robust over long periods, it exhibits fluctuations when examined over shorter horizons. The variations are particularly pronounced in five-, ten-, and twenty-year intervals, where factors such as economic crises and inflation have influenced the correlation strength.
Historical fluctuations
During the first half of the 20th century, the lowest recorded rolling 50-year correlation was 0.6, a period marked by significant historical events, including two world wars and the Great Depression. Interestingly, despite these challenges, the correlation did not decline further, highlighting the resilience of the relationship between earnings and stock prices.
Shorter timeframes tend to exhibit greater variability. For example, the rolling 20-year correlations dropped below 0.50 between February 1918 and December 1928, and again in December 1948. Furthermore, the rolling ten-year correlations dipped into negative territory during the aftermath of World War I, World War II, and the high inflation years of the late 1970s and early 1980s.
The predictive value of correlation changes
To determine whether shifts in the earnings-price correlation could serve as indicators for future returns, we conducted regression analyses comparing the correlation levels to subsequent annualized returns. The results revealed a high R² value of 0.95 for the long-term relationship between earnings and stock prices from 1871 to.
However, when we examined shorter rolling timeframes, the predictive power diminished significantly. The rolling ten-year and five-year windows yielded R² values close to zero, indicating a lack of meaningful predictive ability regarding future stock performance. Only the rolling 50-year period maintained a discernible relationship with an R² of 0.53, suggesting that while long-term correlations are informative, they do not necessarily translate into predictive power for shorter horizons.
Ultimately, the evidence supports the notion that while earnings are critical in explaining long-term market behavior, they offer limited assistance in timing market movements. Investors must consider other factors in conjunction with earnings to make informed decisions regarding their investments.
