A subscriber asked about the lead-lag relationship between S&P 500 earnings and S&P 500 Index returns. To investigate, we relate actual aggregate S&P 500 operating and as-reported earnings to S&P 500 Index returns at both quarterly and annual frequencies. Earnings forecasts are available well in advance of returns. Actual earnings releases for a quarter occur throughout the next quarter. Using quarterly S&P 500 earnings and index levels during March 1988 through June 2020 and September 2020, respectively, we find that:
The following chart summarizes Pearson correlations between actual quarterly S&P 500 operating and as-reported earnings and quarterly S&P 500 Index return for relationships ranging from return leads earnings by eight quarters (-8) to earnings lead return by eight quarters (8). Notable points are:
- The earnings-return relationship is very noisy. In other words, aggregate earnings do not dominate as a driver of stock market return.
- The strongest indication is that return leads earnings by one quarter (-1). The coincident relationship (0) is also relatively strong.
- As-reported earnings may be more useful for predicting near-term stock index performance than operating earnings. However, as noted above, actual earnings for a given quarter become public throughout the next quarter (1) so that any earnings lead is not fully exploitable.
In case lead-lag indications may emerge over longer intervals, we look at annual lead-lag.
The next chart summarizes Pearson correlations between actual annual S&P 500 operating and as-reported earnings and annual S&P 500 Index return for relationships ranging from return leads earnings by four years (-4) to earnings lead return by four years (4). Notable points are:
- The coincident relationship (0) is relatively strong.
- There may be an earnings cycle of roughly four years.
The sample period (31 years) is very short for this kind of analysis.
In summary, available evidence suggests that the relationship between stock index earnings and index return is very noisy and may not be exploitable.
Cautions regarding findings include:
- As noted, the sample period is short for annual analysis.
- As noted, the relationship between index earnings and index return is noisy, undermining confidence in inferences.
See also “Stock Market Valuation Ratio Trends” and “FactSet S&P 500 Earnings Growth Estimate Evolutions”.
