Stats › Valuation › Earnings surprises
Earnings surprises and the beat rate
Every quarter, companies report their earnings and the financial press tells you how many of them beat expectations. It sounds like a scoreboard. Beat is good, miss is bad, and a quarter where most companies beat must have been a strong one.
It doesn't work like that, and once you see why, a lot of quarterly reporting stops making sense in a useful way. Beating is the normal outcome and has been for years. Roughly three quarters of S&P 500 companies clear their estimate in any given quarter, which tells you far more about how the estimates get set than about how the companies performed. Once you know the baseline, a headline saying 'most companies beat' stops being news and starts being weather.
This page has the current beat rates, their five and ten year averages, the same numbers for revenue, and a good example of how a single company can drag an aggregate figure around by more than twenty points on its own. I put it together because the beat rate is one of the most repeated numbers in financial media and one of the most consistently misread, and the fix is just knowing what normal looks like. Everything here comes from one source, published on a stated date and linked at the bottom. It's a free PDF, so if a number surprises you, the original is a click away.
86% of S&P 500 companies beat their earnings estimates in Q2 2026, against a five year average of 78% and a ten year average of 76%. Beating is the normal outcome, not the exception, which tells you something about how estimates are set rather than how companies perform. The aggregate surprise was 39.3% above estimates, but that figure is distorted by a single company: excluding Alphabet it falls to 12.6%, still well above the ten year average of 7.4%.
Members only
The rest of this page is for members
Below this point there are 7 sections, 1 chart and 1 named source, roughly 1200 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.