Earnings surprises and guidance: beats, misses and price moves

Four times a year, a public company reports its earnings and the market immediately compares them with a number decided in advance: the consensus estimate, the average forecast of the analysts who cover the stock. Coming in above that number is a beat, below it a miss, and the gap is the earnings surprise. Financial news is organized around it.

The surprise is easy to calculate and easy to over-read. Estimates are not neutral predictions; they are shaped by what the company says in advance, its guidance, and they tend to drift down before the report. So most companies beat most of the time, and the stock price often ignores the beat entirely.

This page shows how surprise and surprise percent are computed using Apple's last five quarters from Finabase's earnings tab, what guidance is and how Apple gives it, and why the next day's price move is a separate question.

Consensus estimates

Sell-side analysts at banks and brokers publish forecasts for the companies they cover: EPS, revenue, margins, often for each quarter and year ahead. Data vendors collect these forecasts and publish an average (sometimes a median) as the consensus. For a heavily covered stock like Apple, dozens of analysts contribute.

Three things are worth knowing about the number. First, it usually forecasts the adjusted EPS the company will report, on the company's own definition, not necessarily GAAP (see GAAP vs non-GAAP earnings: how adjusted numbers are built). Second, it changes until report day as analysts revise. Third, different vendors include different analysts and update at different times, so two sites can show two consensus figures for the same quarter.

Surprise and surprise percent

The absolute value in the denominator matters when an estimate is negative. A company expected to lose $0.20 a share that loses $0.10 has a surprise of +$0.10, or +50%. Without the absolute value the sign would flip. Very small estimates also produce huge percentages (a $0.02 estimate and a $0.05 result is +150%), so surprise percent is most meaningful for companies with stable, positive earnings.

Apple's last five quarters

Each row is the same calculation: for the June 2025 quarter, $1.57 − $1.44 = $0.13 and $0.13 ÷ $1.44 = 9.03%; for March 2026, $2.01 − $1.95 = $0.06 and $0.06 ÷ $1.95 = 3.08%. The average surprise across the five quarters is (9.03 + 6.94 + 6.74 + 3.08 + 6.88) ÷ 5 = 6.53%.

One detail shows why you check the filing. The vendor records $2.85 for the December 2025 quarter; Apple's 10-Q reports diluted EPS of $2.84. A one-cent difference in the "reported" figure moves that quarter's surprise from 6.74% to 6.37%. Vendors sometimes record a rounded, adjusted or early figure; the 10-Q is the authority.

Why most companies beat

If analysts' forecasts were unbiased, companies would beat about half the time. In practice, in most quarters a clear majority of large US companies report EPS above consensus. Apple's five-for-five record is typical rather than exceptional. Several forces push in the same direction:

• Estimates are walked down. Analysts' forecasts for a quarter often start high and are revised lower as the report approaches, partly in response to company guidance and commentary. The final consensus is a lower bar than the early one.
• Companies guide conservatively. A company that sets an outlook it can beat avoids the penalty of a miss. Analysts know this and anchor on the guidance anyway.
• Management has levers. Expense timing, the tax rate, and buybacks that shrink the share count can all move EPS by a few cents in a quarter.

So a "beat" is the expected outcome, and its information content is low. More useful questions: Is the size of the beat larger or smaller than usual for this company? Did revenue also beat, or only EPS? Did the beat come from operations or from a lower tax rate? And what happened to the estimates for the next quarters after the report? Apple's March 2026 quarter, a 3.08% beat, was the smallest of the five; by itself that says nothing, but it is the kind of change worth noticing.

Guidance

Guidance is a company's own forecast of its future results: a revenue range, a margin range, an EPS range, or qualitative direction. US companies are not required to give it. Some give detailed annual EPS ranges, some give only next-quarter revenue, and some give none. Guidance is a forward-looking statement, and companies use the safe harbor of the Private Securities Litigation Reform Act of 1995, which is why releases and calls open with a cautionary statement.

Apple's practice shows where to look. Its Q4 fiscal 2025 earnings release contains results, statements, a dividend announcement and a reconciliation, but no outlook; its forward-looking statements paragraph refers only to the dividend and future business plans. Apple's management discusses the outlook on the quarterly conference call, held that day at 2:00 p.m. Pacific time. To know what Apple expects, you read or listen to the call, not the release; the Earnings calls: how they work and what to listen for page covers how calls are structured.

Guidance often moves a stock more than the quarter just reported, because the price reflects expectations about the future and the reported quarter is already the past. A beat paired with an outlook below what analysts expected is a common pattern behind a falling share price on a "good" report.

Price reaction is a separate question

Apple releases earnings after the US market closes, so the first full trading-day reaction is the next day's close compared with the release-day close. For the five quarters in the table above:

Five beats; three next-day declines. The smallest beat (+3.08%) was followed by the largest gain (+3.2%), and a +6.88% beat by the largest fall (−7.4%). The three earlier releases in the figure (October 2024 to May 2025) were each followed by a decline, of 0.7% to 3.7%. This data does not say why any of these moves happened. A stock's reaction combines the reported quarter, revenue, margins, the outlook, the call, what was already priced in, and the whole market's move that day. Five observations of one company prove nothing statistically; they illustrate that the EPS surprise alone is not an explanation.

Two more cautions. The finance press often reports "whisper numbers" or buy-side expectations above the published consensus; those are not observable, which makes "the stock fell because the beat wasn't big enough" an untestable story. And a one-day move is noise against a long holding period: AAPL rose from $255.46 at the end of fiscal 2025 to $348.91 on October 29, 2026, through all five of these reactions. The Trading earnings season: implied moves, IV crush and earnings gaps page covers the trading side of earnings dates, including the risk of holding through a report.

Pitfalls

• Mixing bases. Comparing a GAAP result with an adjusted consensus can create a fake miss or beat. Check what the vendor compares.
• Trusting one vendor. Consensus figures differ between vendors; small surprises can flip sign.
• Ignoring revenue. An EPS beat on a lower tax rate or fewer shares is weaker than one from higher sales and margins.
• Reading the beat rate as quality. A company that beats every quarter may simply guide low.
• Inventing causes. Price moves have many inputs. Treat any single-factor explanation of a one-day move with suspicion.

Surprises compared with related measures

How Finabase does it

The Earnings tab lists each quarter with its report date, quarter ending, EPS estimate, EPS reported, surprise and surprise %, all from the data provider; the table above is read from it. A Next earnings card shows the upcoming date and the average estimate: on October 30, 2026 it listed November 2, 2026 for the quarter ending September 2026, with an estimate of $1.99. Finabase does not show revenue estimates or guidance.

The summary page at /symbol/aapl repeats the next earnings date and estimate, and the earnings calendar lists companies reporting from five days back to twenty days ahead, with sector, last price and change. The calendar has no estimates; open a symbol's earnings tab for those. To see reactions like the ones above, open the symbol's chart and look at the closes around each report date.

Frequently asked questions

How is an earnings surprise calculated?

Subtract the consensus EPS estimate from the reported EPS; for the percentage, divide that difference by the absolute value of the estimate. Apple's September 2025 quarter: $1.85 − $1.73 = $0.12, and $0.12 ÷ $1.73 = 6.94%.

Why does a stock fall after beating earnings?

Because the price reflects expectations about the future, not just the last quarter. Guidance, revenue, margins, what investors already expected and the overall market all feed into the move. Apple fell 7.4% the day after a 6.88% beat on July 30, 2026; the data alone does not tell you why.

What is the difference between guidance and the consensus estimate?

Guidance is the company's own forecast; the consensus is the average of outside analysts' forecasts. Analysts usually anchor their estimates on guidance, so the two are closely linked.

Does Apple give earnings guidance?

Apple's earnings release contains no outlook. Management discusses its expectations on the quarterly conference call instead.

Summary

An earnings surprise is reported EPS minus the consensus estimate, and surprise percent divides that by the estimate. Apple beat in each of its five quarters from June 2025 to June 2026, by 3.08% to 9.03%, which is the norm rather than the exception because estimates are guided and revised down before reports. Guidance, the company's own outlook, often matters more than the quarter itself, and Apple gives its outlook on the call rather than in the release. The next day's price move is a separate question: Apple fell after three of those five beats, and nothing in the surprise alone explains why.