|
The Beat Rate Says More About the Estimates Than About the Companies
At the close of trading on Friday, May 29, the S&P 500 sat at 7,580.06, per S&P Dow Jones Indices. The Nasdaq finished at 26,972.62 and the Dow closed at 51,032.46. All three were records. Oil fell 17% for the month, the largest monthly decline since April 2025. The 10-year Treasury yield settled at 4.443%.
Behind those numbers is a first-quarter earnings season that, by every standard metric, looks spectacular. The problem is in the standard.
What an 85% beat rate actually means. FactSet tracks the share of S&P 500 companies that report earnings above the consensus analyst estimate. The five-year average is 78%. This quarter the number is 85%. The aggregate surprise, the margin by which reported earnings exceed forecasts, is 16.7%. The five-year average is 7.3%. On the surface, Corporate America is crushing it.
Look one layer deeper. Analysts set their estimates. Companies guide those estimates lower during the quarter. Then companies beat the lowered bar. The beat rate measures the distance between the final consensus and the actual print. It does not measure whether the economy is strong. It measures whether the consensus was conservative. At 16.7%, it was roughly twice as conservative as usual. That is not strength. That is low expectations.
Why this matters for the next two quarters. When companies beat by this margin, analysts revise their forward estimates upward. That is already happening. FactSet now projects S&P 500 earnings growth of 21% for calendar 2026, up from 17% at the start of the year. The beat rate guaranteed itself a higher bar. The companies that cleared 16.7% above consensus in Q1 now face a Q2 estimate that has been revised higher. The easy beats are behind them.
Dell is a case study. The company reported $4.86 EPS against a $2.93 consensus, a 66% surprise. AI server revenue came in at $16.1 billion, up 757% year-over-year. The stock rose 33% in a single session. Now the Street expects Dell to sustain something close to that trajectory. The Q3 estimate has already been revised upward. If Dell merely meets the new bar, the stock does not rise 33% again. That dynamic applies across the entire index.
The breadth problem underneath the beat rate. AI stocks now represent roughly 45% of S&P 500 market capitalization, per GoTrade. Strip them out and the index has been essentially flat since February. Goldman Sachs projects 6% S&P returns for the remainder of 2026. That number assumes the AI names keep delivering and the other 55% of the index does not contract. The 85% beat rate masks how narrow the actual earnings acceleration is. Dell, Snowflake, Micron, and Broadcom accounted for a disproportionate share of the aggregate surprise. The median S&P 500 company beat by roughly 4%, per FactSet. That is barely above the five-year average.
The base case here is not bearish. It is that the market is priced for earnings growth that the beat rate made look easy. It will not be. The estimates have been revised higher and the beat rate will normalize. The 85% headline will not repeat in Q2. If the AI buildout continues at pace, the narrow group of companies driving the index can sustain it. If it slows even modestly, the revised estimates become a ceiling, not a floor.
My read: the nine-week rally was earned by real earnings. But the earnings looked better than they were because the bar was set too low. June is the month where the bar resets. Broadcom on Wednesday and payrolls on Friday are the first two tests. The question is no longer whether companies can beat. The question is whether they can beat a bar that just moved.
Eighty-five percent of companies beat estimates. The five-year average is 78%. The difference is not that companies got better. The difference is that estimates got worse. The rally was real. The bar it cleared was not.
|