ResonanceMarket Intelligence · GammaQC
Resonance The Primer · GammaQC
◆ The Primer · Foundations

A beat is not a signal. The drift is.

Post-earnings announcement drift — PEAD — is one of the most durable patterns in markets, and one of the most misread. Understanding it changes what you look at when a company reports.

The Desk · GammaQC Research

Every earnings season runs on the same reflex: a company beats, the headline says "tops estimates," and the assumption is that the stock should go up. Sometimes it does. Often it does the opposite. The reflex is wrong because it fixates on the one number that carries the least information — whether they beat — and ignores the one that carries the most: what the stock does in the days after.

That second thing has a name. It is called post-earnings announcement drift, or PEAD, and it has been studied for over half a century.

What post-earnings drift actually is

PEAD is the tendency of a stock to keep moving in the direction of an earnings surprise for days to weeks after the report — as if the market absorbs the news slowly rather than all at once. A genuinely positive surprise tends to keep drifting up; a negative one tends to keep bleeding. The market, in other words, is not perfectly efficient in the hours around a print. It leaks.

PRICE DAYS AFTER THE PRINT → EARNINGS pre-print: flat, uncertain the drift — where the edge lives what many expect: a fade
Illustrative — the shape of the pattern, not a specific security.

The effect is not folklore. It was first documented by Ball and Brown in 1968 and formalized by Bernard and Thomas in 1989, and it has been replicated across decades and markets. It is one of the few anomalies academics broadly agree is real. What has changed is its size: as the pattern became widely known, it decayed — front-run by faster money and arbitraged down in the most liquid names. It did not disappear. It concentrated.

Why the beat itself tells you almost nothing

Here is the part that breaks most people's intuition. Across a full earnings season, beat rate and post-print direction are close to uncorrelated. In a strong tape, the large majority of companies beat consensus — because consensus is an estimate analysts and companies have every incentive to set beatably low. If almost everyone clears a bar that was placed on the floor, clearing it is not information.

"They all beat" tells you nothing about the move. The surprise that matters is the one the market was not already positioned for.

What actually moves a stock after earnings is the gap between the print and what was expected — and "expected" is rarely the published consensus. Into a crowded name, the market quietly sets a higher, unofficial bar than the estimate on the screen. A company can clear the published consensus, fall short of what the market actually expected, and sell off within minutes. The tape is not reacting to the beat. It is reacting to the surprise relative to positioning.

The three things that happen after a print

Priced-in beat

Beat, then fade

The surprise was already expected and owned. The stock gaps up on the headline and drifts back — the classic "sell the news." Near-zero or negative drift despite a clean beat.

Genuine surprise

Beat, then drift

The result clears a bar the crowd had not positioned for. Under-owned and under-covered names leak the news over days — the textbook upward drift.

Sell-the-beat

Beat, then break

Guidance, margins, or a single line in the call reframes the story. The headline is green; the institutional flow is a persistent negative drift. The beat becomes exit liquidity.

Notice that all three start with the word "beat." The headline is identical. The only way to tell them apart is to stop reading the beat and start measuring the drift — and to know, going in, how much was already priced.

How to actually use it

PEAD is a lens, not a trade. It does not tell you to buy or sell anything, and anyone who sells it as a guaranteed edge is selling you the 1968 version of a pattern the whole market has since read. Used honestly, it changes three habits:

  • Weight the surprise, not the beat. Ask what the market actually expected — not just the published consensus — before you ask whether they cleared it.
  • Watch the first days, not the first minutes. The initial gap is noise and positioning. The drift over the following sessions is where the pattern, if it is there, reveals itself.
  • Define your invalidation before the print. A view on drift is still a view that can be wrong. The level that says you are wrong belongs on the chart before the report, not after — a thesis without a stop is a horoscope.

None of this requires you to predict the number. It requires you to measure the historical drift for a given name, know how the Street, the model, and the crowd are positioned, and size for the pattern rather than bet on the headline. That measurement is exactly what an earnings-intelligence tool is for.

Frequently asked

Does a stock always go up after beating earnings?
No. A company's beat rate and the direction its stock moves after the print are close to uncorrelated — "they all beat" tells you almost nothing about the move. When a beat was already expected, the stock often fades; when the surprise is genuine and under-owned, it tends to drift in the surprise's direction.
How long does post-earnings drift last?
The classic academic window runs up to roughly 60 trading days after the announcement. In modern, heavily-covered large caps the drift is often faster and shallower; in less-covered names it can persist longer. The horizon varies with how much of the surprise was already priced in.
Is PEAD still profitable?
Post-earnings drift is a well-documented anomaly, first identified by Ball and Brown (1968) and formalized by Bernard and Thomas (1989). Like most published anomalies it has decayed as it became widely known, and it persists unevenly — concentrated in genuine surprises and less-followed names. It is a pattern, not a guarantee, and past behavior does not promise a future outcome.
Why does a stock fall after beating earnings?
Because the published consensus is not the real bar. Into a crowded, popular name, the market quietly expects more than the official analyst estimate. A stock can clear the published consensus, fall short of that higher unofficial expectation, and sell off anyway — the beat was already priced in, so it carried no surprise.

See the real drift for any ticker — not the shape, the numbers.

  • Earnings Intelligence measures the actual post-earnings drift for a given name — beat rate versus real five-day move, before-open or after-close timing, and where the Street, the model, and the crowd disagree.
  • Every read carries a mandatory invalidation level — the price that says the thesis is dead — because a view on drift is still a view that can be wrong.
  • A 7-seat executive council pressure-tests each verdict and shows you which seats dissent, and every verdict seals into a tamper-evident, timestamped receipt — including the calls it gets wrong.

Stop reading the beat. Measure the drift.

Run any ticker's real post-earnings drift, timing, and invalidation level. Free, no signup.

Check the drift ▸ gammaqc.com/earnings