The long-term relationship between stock prices and corporate earnings reflects a consistent pattern, confirmed by over a century of analysis drawn from Robert Shiller’s data. This investigation delves into the strength of this connection and assesses the utility of the correlation changes between earnings and stock prices in forecasting future market returns.
The research indicates that while earnings influence market behavior over extended periods—defined as over 10 years for effective retirement planning—the fluctuations in this correlation do not offer a reliable basis for predicting returns. The study employed monthly averages of earnings and stock prices from the S&P Composite, spanning from 1871 to 2024. Notably, the correlation values consistently hover around 0.97 for various periods, including those before and after regulatory changes introduced by the 1940 Investors Act.
Analysis revealed that while long-term correlations remained high, shorter time horizons demonstrated greater volatility. For instance, five- and ten-year correlation figures displayed significant fluctuations due to various economic events, such as wars and inflationary periods, suggesting reduced reliability for shorter-term forecasts. The statistical evaluations show minimal predictive power; the R-squared values connecting correlation with subsequent returns were significantly lower for shorter time frames, approaching zero. Conversely, even the 50-year intervals, while demonstrating some correlation, still provided limited predictive insights.
In conclusion, while earnings are a solid long-term indicator of stock performance, they do not serve as effective tools for market timing.
Why this story matters
Key takeaway
Opposing viewpoint
