Q3 Earnings Season Opens With Goldman Beat and Rare JPMorgan Miss

A strong season with weaker beats

Third-quarter earnings season started with a contradiction. Most companies are beating forecasts, but by much less than usual. Analysts had raised their numbers going into the season, and the early reports have mostly met that higher bar rather than cleared it with room to spare. The big banks went first and reported very different quarters. Goldman Sachs beat estimates by a wide margin, while JPMorgan Chase missed, which it rarely does. Taiwan Semiconductor Manufacturing Co. (TSMC) then posted another record quarter on AI chip demand.

What matters most is how narrow the strength is. Headline profit growth for the S&P 500 is near 30%, but the typical company is growing far more slowly.

The aggregate picture: more beats, smaller beats

FactSet's October 16 update calls the start of the season mixed. About 9% of S&P 500 companies have reported, and 84% of them beat earnings estimates, above the five-year average of 78% and the ten-year average of 77%.12 The size of those beats is small, though. In aggregate, companies are coming in 2.6% above estimates, compared with averages of 7.5% over five years and 8.0% over ten.12

The blended growth rate has slipped as a result. FactSet puts expected third-quarter earnings growth at 29.0%, down from 29.4% a week earlier and 29.2% at the end of September. Downward estimate revisions in Health Care were the biggest drag.12 Even at 29.0%, it would be the third straight quarter above 25% growth and the eighth straight quarter of double-digit growth.12 Revenue is holding up a little better. Blended revenue growth has risen to 12.6% from 12.2% at quarter-end, although only 67% of companies have beaten revenue estimates, below historical norms.12

The soft start runs against what forecasters expected a week earlier. On October 9, FactSet noted that actual growth has beaten quarter-end estimates in 37 of the past 40 quarters. Applying the average historical upside, it projected that third-quarter growth would likely finish above 35%.11 That forecast depends on beats of normal size. If surprises stay near 2.6% instead of the 7–8% average, the usual late-season rise in the growth rate will be smaller.

Why the bar was so high

The main reason for the small beats is that analysts raised estimates during the quarter, which they rarely do. Ameriprise's Anthony Saglimbene noted that analysts lifted third-quarter S&P 500 EPS estimates by 1.4% during the quarter. Over the past five years, estimates have fallen an average of 2.2% over the same stretch. That leaves companies less room to surprise.19 Companies also guided unusually high. Of 116 companies that issued EPS guidance, 72 guided positively, according to an EXANTE summary of FactSet data. That is the highest count since FactSet began tracking in 2006.17

State Street pointed out that if the quarter comes in as expected, it would be the first quarter since Q3 2008 in FactSet's data to complete both three straight quarters of 25%-plus growth and eight straight quarters of double-digit growth.14 Expectations have been rising steadily all year. Consensus EPS for calendar 2026 rose from $313.72 at the end of February to $362.65 by October 2, a gain of 15.6%.14

The sources don't fully agree on the size of the expected growth. One planning firm noted that Zacks projected roughly 22–24% growth in August and September, well below FactSet's figure. The reported consensus depends on which provider and method you use.15 FactSet's own StreetAccount preview noted that the Street has also cited a 27% consensus that excludes some other income.13

Growth concentrated in AI chips

The 29% figure overstates how most companies are doing. StreetAccount's preview expected semiconductor earnings to grow about 130%, contributing more than 40% of the entire index's earnings growth. Previews also noted that more than half of S&P 500 earnings growth comes from AI infrastructure companies.13 One independent analysis cited Goldman Sachs estimates that Micron and Nvidia alone account for about 34% of expected growth. The same estimates put the median company's growth at about 9%, down from 14% in the second quarter.20

Other sectors show the same pattern. Communication Services is expected to grow earnings about 51%, but that falls to roughly 12% without its two largest contributors, Meta Platforms and EchoStar.1619 Energy is expected to lead all sectors. Tight fuel supplies have widened refining margins, and refiners are expected to grow earnings more than 400%.1319

TSMC shows clearly where the profits are going. The chipmaker reported third-quarter net profit of NT$763.1 billion, up 68.7% from a year earlier and its sixth straight record quarter. Both revenue and profit beat LSEG SmartEstimates.6 Revenue rose 50.9% year over year to NT$1.494 trillion, above the company's own guidance range. Gross margin reached 66.4%, slightly below the prior quarter because early 2-nanometer production diluted margins.7 Operating margin of 58.7% came in above the top of guidance, which CNBC cited as evidence of TSMC's pricing power.6

Banks: Goldman's blowout, JPMorgan's miss

The big banks matter more than their small expected growth would suggest. Financials were forecast to grow earnings only about 3% this quarter.19 The Federal Reserve raised its target range by a quarter point on September 16, to 3.75%–4%.28 With rates rising, the bank reports showed how Wall Street's businesses are holding up.

Goldman was the standout. It earned $17.01 per share on net revenue of $18.68 billion. Equities trading revenue jumped 72% to $6.44 billion, and investment banking fees rose 21%.1 Net earnings climbed 29% to $5.29 billion, and the firm said quarterly revenue was its second highest ever.25 Advisory revenue of $1.65 billion was a company record, although CNBC noted that Goldman's deal pipeline shrank.21

The size of the beat depends on which consensus you use. Goldman beat Barron's FactSet-based estimate of $12.86 by about 32%, LSEG's figure of $13.02, and the $13.09 cited by Investing.com.215 On any of these measures, the beat is far larger than most companies are posting this season.

JPMorgan missed. It reported $5.57 per share against a $5.90 estimate, and revenue of $50.91 billion against $51.22 billion expected.1 Excluding $658 million in securities losses, EPS would have been $5.75, still below consensus.1 Fixed-income trading came in about $550 million below estimates, and expenses grew faster than revenue. Revenue growth slowed to 8% from 27% in the second quarter.1 Several outlets called it JPMorgan's first miss in nearly three years.2529 Coverage also differs on JPMorgan's revenue: some outlets cite managed revenue of $50.9 billion and others reported revenue of about $50.2 billion, which accounts for gaps between reports.26

Most of the other big banks beat. Citigroup earned $2.85 per share on $23.85 billion in revenue, beating estimates on both, and reported its best third-quarter revenue in a decade.1 Accounts of Citi's profit growth vary, from 17% in the Financial Times to 40% elsewhere, probably because they use different baselines.2227 Morgan Stanley earned $3.03 per share on $20.34 billion in revenue, helped by a 41% jump in equities trading. Fixed income and investment banking fell short of expectations.4 Wells Fargo beat on EPS at $1.87 but missed on revenue.10 One tally of the six largest banks had Bank of America and JPMorgan as the only EPS misses.24

The trading results split along one line. Equities were strong across the industry and bond trading was weak. Oppenheimer analysts called the gap between JPMorgan's equities and fixed-income results "odd," since the two usually move together.22 The broader dealmaking environment is weaker. Dealogic data show global investment banking revenue fell 6.6% year over year in the third quarter.1 JPMorgan's CFO said deals are being "delayed."1

Beats didn't move the stocks

Investors did not reward the beats. Reports of Goldman's stock move vary by timing. It rose nearly 4% before the open1, but one end-of-day tally found all six big banks closed lower on their earnings days. That included a 0.2% dip for Goldman and a 5.6% drop for Citigroup.24 Morgan Stanley fell about 3% despite beating, and Wells Fargo analyst Mike Mayo called its quarter "good, but not great."4 A Truist survey cited in one report found institutional support for bank stocks has fallen from 82% last December to 35%.

The takeaway

In my reading, this season is mainly about a narrow group of companies. The index is on track for its eighth straight quarter of double-digit profit growth. Much of that growth comes from chipmakers and AI infrastructure, while the median company grows in single digits.1220 Analysts raised estimates going into the season, and so far companies have beaten them by less than usual. That is why markets are focusing on guidance and how well the growth holds up rather than on beat rates. Analysts still expect 28.6% growth in the fourth quarter.12 If the AI chip cycle slows, few other sectors are growing fast enough to make up the difference.