Analyst & Spread
Blended Consensus
An estimate from two or three analysts, mixed with our own trend in proportion to the analyst count; a lone analyst is not counted.
Definition
One analyst's forecast is an opinion, not a consensus. Where fewer than four analysts cover a year, every estimate they give that we also forecast ourselves (revenue, EPS, net income, free cash flow, operating and gross profit, EBITDA, the dividend, a bank's pre-provision profit and net interest margin, a REIT's FFO per share) is their figure mixed with our own for that year: one analyst counts for nothing (the cell shows our own figure), two for a third, three for two thirds, four or more for all of it. Our side is built from our own trends only, never from the analysts' lines, so they are not counted twice; EPS and net income move together, so EPS is still net income over the share count. The mix is on growth (logarithms) for positive figures, a straight average otherwise. The fair value weighs thin consensus years by the same weights.
Formula
w = (analysts − 1) / 3, between 0 and 1; shown = consensus^w × ours^(1 − w) (positive figures)
Why it matters
A single analyst can be far off, and a thinly covered estimate should not move the table as much as a broad consensus does. The blend keeps the analysts' view in proportion to how many of them hold it.
How Standard Equity uses it
The estimate cell's confidence tooltip says so ("2 analysts, blended with our trend"; a lone analyst: "not counted"), and its interval mixes the analysts' and our own by the same weights. EPS is blended only where the analysts' EPS is on our reported basis. A line we have no figure of our own for (a holding company's, the year after next) is shown as is, and the analysts' own figures still set the price-target upside, the analyst ranges and the basis checks.
The metric
| Category | Analyst & Spread |
|---|---|
| Unit | same as the row |