S&P 500 Profits Surge 50% as Companies Crush Expectations

The S&P 500 is heading for its strongest quarter of profit growth in five years, and it is not close. FactSet's Earnings Insight report of 7 August 2026 puts second quarter earnings growth for the index at 50.4%, a blended figure that combines the results already filed with estimates for the companies still to report. With 88% of the index in, that would be the highest reading since 2021, and 86% of the companies that have reported came in above the per share profit analysts had pencilled in for them, which is also the best showing since 2021.

The size of the beats is stranger than the number of them. Companies are reporting profits far above the estimates they were measured against, revenue is growing at its fastest rate since 2021, and the net profit margin, meaning the share of every sales dollar that survives all the way down to profit, is the widest in a FactSet series that runs back to 2009.

Strategists have taken the quarter at face value. On 10 August, a JPMorgan team led by Dubravko Lakos-Bujas raised the firm's year end target for the index to 8,000 from 7,800, framing the increase as a matter of companies actually delivering more profit rather than of investors agreeing to pay more for each dollar of it.

The bar was raised before the quarter, not lowered

The standard objection to any run of companies beating expectations is that the expectations were quietly lowered first, in the weeks before results are filed, so that the bar is easy to clear by the time anyone jumps. FactSet tested that for this quarter and found the opposite. The profit estimate for the second quarter rose 3.4% between the end of March and the end of June, where the usual pattern over the past decade has been a cut of a few percent, and it was the largest such increase since 2021. The rise was not evenly spread either, with estimates for the energy companies raised sharply while those for health care were cut.

Revenue tells the same story, and revenue is the one line that cannot be lifted by revaluing something. The energy sector sold into an oil price far above last year's, and the chipmakers reported sales more than three quarters higher than a year earlier, both of which are sales made to customers rather than adjustments made in a ledger.

Two companies inside the index number

FactSet publishes the arithmetic for what the index looks like with individual companies removed, and this quarter it is startling. Take out Alphabet and Amazon.com and growth for the whole index falls from 50.4% to 32.0%, because between them those two account for most of the increase in index profits since the end of June. The chart below shows how much each one carries on its own.

The distance between 50.4% and 32.0% is two accounting entries, and neither of them is a sale, a product or a customer.

Alphabet's record beat was a three cent miss

Alphabet's advertising and cloud business did have a good quarter, with operating profit up 30% on a year earlier. What happened below that line was of an entirely different order. A catch all category the company calls other income and expense, which collects everything that is not part of running the business, swung to a gain of roughly $98 billion from under $3 billion a year earlier, and almost all of it was an increase in the value of shares Alphabet holds in other companies. It has not sold those shares, and the release describes the gain as unrealised, meaning it exists on paper at the new market price rather than in cash. Reported earnings came to $9.11 a share, against $2.31 a year earlier.

Alphabet then does the arithmetic itself. In a footnote to that table, inside the filing a company lodges with regulators on results day, it states that the gain on its investments added $6.26 to earnings per share.

FactSet records Alphabet's quarter as $9.11 against the $2.88 that analysts had expected on average, one of the largest positive surprises in the index's history. Take out the $6.26 the company itself attributes to the gain and $2.85 is left, three cents below the estimate that record beat is measured against. The estimate was built by analysts modelling an advertising and cloud business, while the figure it was compared with contains a revaluation of a portfolio of shares.

Amazon's markup, and the business under it

Amazon's quarter has the same shape. Sales grew by a fifth, and below the operating line that same catch all category brought in $53.4 billion against about $1 billion a year earlier, which the release attributes mainly to the company's investment in Anthropic, leaving reported earnings of $5.75 a share.

Amazon does not disclose a per share effect for the gain the way Alphabet does, so what follows is my arithmetic rather than a company disclosure. Applying the quarter's tax rate and share count, that $53.4 billion works out at roughly $3.80 of the $5.75, which would leave something near $1.95 against the $1.82 analysts had expected. What is not an approximation is the 43% rise in operating profit, which came from selling goods and computing power and beat the top of the range the company had guided investors towards.

What the higher target rests on

A price target justified by delivery inherits whatever that delivery is made of, and roughly $118 billion of after tax income across these two companies in a single quarter came from writing up holdings that were never sold. That does not repeat unless the same holdings are written up again by a similar amount, and the entries run in both directions, a point Alphabet makes itself when it notes that swings in the value of its investments could contribute significantly to the volatility of that line in future periods.

The market has applied part of that discount already. Companies reporting positive surprises this quarter rose 0.4% on average in the days around their release, less than half the usual reaction of the past five years, and investors are paying more for each dollar of the past year's profits than they have on average in recent years, on a profit figure that already contains both gains.