Analyst Estimate Dispersion

Methodology

Dispersion. For each S&P 500 company, Alpha Vantage's EARNINGS_ESTIMATES endpoint returns the consensus (average) EPS estimate for the current fiscal year, plus the high and low estimates and the number of analysts. Dispersion here is a relative range, (High − Low) ÷ |Average| × 100, which lets stocks with very different EPS levels be compared. It isn't a standard deviation (only the high, low, average and count are available), so treat it as a proxy for spread. Leaderboards and the scatter panel require at least 5 covering analysts. With fewer, a wide high/low spread is usually just one outlier estimate.

Versus Earnings Revisions. Earnings Revisions asks whether the consensus itself is moving, and how fast (Net Revision Ratio and Estimate Drift). This page asks how much analysts disagree with each other at one point in time. A stock can have a tight consensus that's moving fast, or a widely split one that's stable. The scatter panel connects the two: do stocks with more disagreement also see more analysts changing their estimate over the trailing 30 days?

The scatter panel shows Pearson (OLS) and Spearman rank correlation, as on Factor Analysis, so one or two extreme stocks can't carry the result. Sector and company name come from Sector Beeswarm.

Refreshes weekly (Saturday). A current-fiscal-year estimate's high/low spread doesn't change much day to day. The history chart starts from this page's first run and isn't backfilled, because Alpha Vantage's revision counts are trailing windows, not a queryable history.