Everyone agrees earnings are booming. Here is the strongest case that they aren't.
S&P 500 companies grew earnings more than 50% in the second quarter of 2026 and 86% beat expectations. Both facts are true. Neither means what the headline suggests.
- What happened
- The blended earnings growth rate for S&P 500 companies in the second quarter of 2026 was 50.4%, with 86% of companies beating estimates — well above the five-year average of 78%.
- Why
- Excluding Alphabet and Amazon, whose results included substantial investment gains, blended growth falls to 32.0%. And a high beat rate partly reflects how estimates are set, not only how companies performed.
- The numbers that matter
- 50.4%Blended Q2 EPS growth32.0%Excluding two companies86%Beat rate20.0×Forward P/E19.0×10-year average
- Why it matters
- Earnings growth is the single most cited justification for current equity valuations. If the number is flattered, so is the justification.
- What happens nextas we saw it on 16 Aug 2026
- Q3 earnings growth was estimated at 27.4% on August 7, and full-year 2026 at 30.0%. The first real tests arrive with the retail results due in mid-August.
Two companies account for roughly 18 percentage points of the index's reported earnings growth.
| As reported | 50.4% |
|---|---|
| Excluding Alphabet and Amazon | 32.0% |
This series takes a narrative that almost everyone accepts and assembles the strongest available evidence against it. Not for the sake of disagreement — because a consensus that has never been stress-tested is a fragile thing to own.
The consensus in August 2026 is that corporate earnings are extraordinary, and that this justifies equity prices. The supporting data is real.
The bull case, stated fairly
Per FactSet's August 7 update: the blended earnings growth rate for the S&P 500 in Q2 2026 was 50.4%. Eighty-six percent of companies reported EPS above estimates, against a five-year average of 78% and a ten-year average of 76%. Q3 growth is estimated at 27.4% and calendar-2026 growth at 30.0%.
That is a genuinely strong set of numbers, and anyone dismissing it is not paying attention. Now here is the case against.
Objection 1: two companies are carrying more than a third of the growth
FactSet's own note points out that excluding Alphabet and Amazon — whose results included substantial investment gains — blended earnings growth for the quarter falls from 50.4% to 32.0%.
50.4% − 32.0% = 18.4 percentage points of index-level earnings growth attributable to two companies, driven materially by gains on investments rather than by operations.
32% growth is still an excellent quarter — FactSet notes it would be the second consecutive quarter above 25%. The point is not that the underlying result is bad. It is that the number in the headline is not the number describing the operating economy.
This matters beyond the index level. Investment gains are marked to market. They can reverse. An earnings base built partly on asset appreciation is more cyclical than an earnings base built on selling things to customers — and it is most flattering precisely when markets are highest.
Objection 2: a high beat rate is partly a measurement artifact
Eighty-six percent of companies beat expectations. It is worth asking what would have to be true for that to be evidence of exceptional performance.
- Analyst estimates are anchored to company guidance, which management issues and then works to exceed.
- Estimates are typically revised downward in the weeks before reporting season, lowering the bar being cleared.
- The long-run average beat rate is already 76%. A world in which three-quarters of companies routinely exceed expectations is not a world where 'expectations' means what the word implies.
The beat rate is closer to a measure of how conservatively companies guide than of how well they performed. It is still informative at the extremes — a collapse in the beat rate would be a genuine warning — but 86% against a 76% average is a modest deviation being reported as a triumph.
Objection 3: the price already reflects it
The S&P 500's forward 12-month P/E was 20.0 in early August, against a five-year average of 19.9 and a ten-year average of 19.0.
The index is not at an extreme valuation. But it is above both averages while earnings growth is at a level that historically does not persist. Paying an above-average multiple for peak-rate growth is the specific combination that has historically produced disappointment — not because the earnings were fake, but because the growth rate normalized while the multiple did not.
Above average, but not extreme. The risk is in what happens to the multiple if earnings growth normalizes.
| Current | 20.0× |
|---|---|
| 5-year average | 19.9× |
| 10-year average | 19.0× |
Objection 4: the economy underneath is not producing this
Corporate earnings grew more than 30% on an adjusted basis in a summer during which July payrolls were first reported at −23,000 (later revised), prior months were revised down by 103,000, retail sales fell 0.6%, and labor force participation dropped to 61.4%.
Profits and employment can diverge for good reasons — margin expansion, automation, international revenue, cost discipline. But a wide and persistent gap between corporate profitability and household conditions is the sort of thing that eventually resolves, and the resolution runs through demand. Companies sell to households. This is an interpretation of the relationship, not a prediction of its timing.
Where this argument is weakest
An honest bear case has to state its own vulnerabilities, so here are three:
- 32% growth is excellent. The adjustment removes the exaggeration, not the strength.
- A forward P/E of 20.0 against a ten-year average of 19.0 is a 5% premium. That is not a bubble by any reasonable definition of the word.
- Investment gains are real profits. Excluding them is analytically useful, but a company that owns appreciating assets genuinely is worth more than one that does not.
Sources
- S&P 500 Earnings Season Update: August 7, 2026FactSet Earnings Insight · Aug 7, 2026 · Secondary source
- The Employment Situation — July 2026U.S. Bureau of Labor Statistics · Aug 7, 2026 · Primary source
- Advance Monthly Retail Trade Report — July 2026U.S. Census Bureau · Aug 14, 2026 · Primary source
Related analysis
What a $5.45 trillion price tag actually assumes about NVIDIA
The most common argument about NVIDIA is whether it is expensive. That is the wrong question. Here is the arithmetic of what the market was paying for on August 15, 2026 — with every assumption disclosed.
The economy is sending two signals at once. The Fed is listening to the louder one.
Payrolls fell. Retail sales fell. Inflation held at 3.4%. And three Federal Reserve officials voted to raise interest rates. Here is how those facts fit together.
Why small caps hit a record on the day the S&P 500 fell
On Friday, August 14, 2026, the Russell 2000 closed at an all-time high while the three big indices finished lower. The explanation is not that investors turned bullish on small companies.