Track direction and magnitude of 30-day consensus EPS/revenue estimate revisions
Answers which way analysts are moving their forecasts and how forcefully. It reads consensus estimate revisions, weighting 30-day EPS and revenue revision magnitude and direction into a 0-100 score labeled positive, neutral, or negative revision. It shows which way the earnings outlook is shifting and how strongly. Studies have found revision momentum associated with subsequent returns on average in their samples, with wide dispersion around that relationship, so a reading describes the direction of analysts' forecasts rather than predicting the stock's price.
Estimate revisions - a screen with genuine academic support behind it, and a precise account of what that support does and does not say.
THE MEASUREMENT. The direction and magnitude of 30-day changes in consensus EPS and revenue estimates.
THE EVIDENCE. Two related findings underpin this. Ball and Brown (1968) documented POST-ANNOUNCEMENT RETURN CONTINUATION - prices kept moving in the direction of an earnings surprise rather than adjusting at once. Be precise about their design, since it is routinely overstated: they used ANNUAL earnings and MONTHLY returns, and the modern quarterly evidence for post-earnings-announcement drift over subsequent weeks and months came from later work. It is among the oldest and most replicated anomalies in the literature. Separately, Chan, Jegadeesh and Lakonishok (1996) found that momentum in analyst estimate REVISIONS predicts returns, and that it is partly distinct from price momentum. So 'estimates are being revised up' is a screen with real support, not folklore.
THE EXPLANATION USUALLY OFFERED is under-reaction: information diffuses gradually, analysts adjust in steps rather than jumping to a new view, and the anchoring is visible in the data as a sequence of revisions in the same direction. A caveat comes with it - documented anomalies frequently weaken after publication, whether because they are traded away or because the original result was partly a product of the sample.
WHAT CONSENSUS ACTUALLY IS. A vendor-defined mean or median of the estimates that vendor includes, governed by its own contributor, staleness and exclusion rules. It is not a survey of the market's belief, and figures from two vendors are not interchangeable. Sell-side estimates also carry a documented optimism bias that tends to decline as the reporting date approaches, so the LEVEL of consensus is less reliable than its DIRECTION - which is a good reason to measure revisions rather than levels, and is what this analysis does.
A DEPENDENCE WORTH NOTICING. Revisions and price are not independent. Analysts revise after price-moving news, so revision momentum overlaps with price momentum, and reading both as separate confirmation double-counts one event.
THE WORDING TO BE CAREFUL WITH. 'Rising estimates often lead price' is a statement about an average historical relationship in a studied sample, with wide dispersion around it. It is not a mechanism, and it is not a claim about any individual name.
The 30-day window, the 0-100 score and the positive/neutral/negative bands are our own convention.
Published reports are readable without an account. The newest is Ether (ETH) and XRP extend a month-long slide as Bitcoin falls below $75,000: trend, momentum and news (Nov 9, 2026). Create an account to run research on the stocks and crypto you choose.