Sales Nowcast

What is a whisper number in equity research?

A desk calendar next to a cup of coffee and a small succulent plant.

Every earnings season, somebody on the desk says "the whisper number is higher than the Street" the day before a retail print. If you've only ever worked with the published consensus estimate, here's what the whisper number is, where it comes from, and why it matters more for retail names than for most other sectors.

The whisper number, defined

A whisper number is the EPS (or sometimes comp-sales) figure that circulates informally in the days before a print, as opposed to the official consensus estimate sitting on your terminal. It isn't published anywhere. It moves through sell-side sales desks, buy-side chat, and the kind of conversation that happens on an earnings-eve call. The premise is that published consensus lags what the smart money actually expects, because sell-side analysts only revise formally a handful of times a quarter and some leave stale numbers in the system long after they've mentally moved on from them.

For a retailer specifically, the whisper number often reflects channel checks that haven't made it into anyone's published model yet: a credit card panel, a footfall estimate, chatter from store visits. If three or four well-regarded buy-side analysts have quietly concluded comps are tracking ahead of the Street model, that collective view becomes "the whisper," even though no single estimate ever gets printed as such.

Whisper number vs. consensus estimate

Consensus is simple to define: the average or median of published sell-side EPS and revenue estimates, aggregated by a data provider into one clean number. It's transparent, it's citable, and it moves slowly, since most sell-side analysts only formally revise a few times per quarter.

The whisper number is the opposite on every count. Informal, unpublished, and able to shift day to day as new data points surface. It also only exists where there's enough trading interest to generate one; nobody whispers about a thinly covered small cap.

The gap matters because stock reaction on the print usually tracks actual results against the whisper. The comparison against consensus carries much less weight. A retailer can beat published consensus EPS by three cents and still sell off hard if the whisper was running five cents higher. A "miss" against consensus that still clears the whisper can trade up. If you've watched a print beat the headline number and sell off anyway, that's usually the whisper talking.

Buy-side estimate vs. sell-side estimate

The whisper number sits closer to the buy-side world, which is worth separating out from the sell side. Sell-side estimates are the published ones: an analyst at a bank puts out a model, a price target, and a rating, and that estimate feeds the consensus average. Sell-side analysts have an incentive to stay reasonably close to consensus, partly because being a wrong outlier looks worse than being wrong alongside everyone else.

Buy-side estimates are different. A portfolio manager or buy-side analyst builds an internal model to size a position, and that model never gets published or averaged into anything. It's built for one reader. Buy-side models tend to react faster to new information, since there's no editorial process and no reputational cost to revising weekly. When a batch of buy-side models converges on a number different from consensus, that convergence is often where a whisper number comes from in the first place.

Where this leaves you ahead of a print

None of this tells you the actual number. It tells you that if you're relying only on last quarter's management commentary and the published consensus, you're working with the slowest-moving inputs in the room. Much of the buy-side convergence that produces a whisper starts from the same ground-level signal: foot traffic at the stores themselves.

That's the gap a daily parking-lot count across a chain's locations is built to close. Sales Nowcast turns vehicle counts from satellite imagery into an indexed, seasonally adjusted time series, so you can watch a traffic trend build weeks before the call instead of hearing about it secondhand once it's already baked into somebody's whisper.

If you'd rather track the traffic yourself than wait for the whisper to catch up, that's what Sales Nowcast is built for.

Get started

← Back to the blog