Beyond the AI Stock Selloff: U.S. Earnings and Market Breadth Both Hit Record Highs
Claire Weston
Strip out AI and the S&P 500 is at an all-time high. Q2 EPS growth estimates have been revised up from 21% to 36%, and the market is now answering a pointed question: can US stocks rally without the AI trade?
How strong are earnings, exactly?
Morgan Stanley data shows US Q2 2026 EPS growth estimates have climbed from 21% to 36% since reporting season began — and the breadth of beats is the widest in five years.
The S&P 500 blended net profit margin has hit 15.7%, the highest since records began in 2009. Strip out Alphabet — the single largest contributor — and the margin is still 14.4%, a historical second-best.
This means → this is not a one-giant story. Seven of 11 sectors expanded margins year-on-year; eight sit above their five-year averages. Profitability is improving across the board.
Why does breadth actually improve once you strip out AI?
The S&P 500 ex-AI index (SPXXAI) and the equal-weight S&P 500 have both touched all-time highs.
In plain terms = for the past year-plus, a handful of AI mega-caps dragged the index higher. Remove them, and the rest of the market is doing better — capital is rotating into a wider set of industries.
This reflects a "breadth relay": as AI stocks pull back, other sectors are catching the index on the strength of their own earnings.
What does the buyback window reopening mean?
Corporate buyers are exiting their earnings-blackout periods — the windows around results when companies cannot repurchase their own shares — and buyback capital is about to re-enter the market.
This means → buybacks are the largest structural bid in US equities. Over the past decade, corporate self-purchases have exceeded the net buying of any single investor class.
The return of buyback flow is seen as potential fuel for the next leg higher, especially with retail and institutional positioning far from crowded.
What story is the macro data telling?
Goldman Sachs' Global MAP indicator shows July global economic activity data delivered upside surprises.
Morgan Stanley's Richmond Fed ISM-equivalent tracker edged up from 53.2 to 53.3; its July manufacturing PMI tracking estimate was revised up to 53.8 — both firmly in expansion territory.
In plain terms = the macro backdrop is not dragging on the story. If anything, it is adding demand-side evidence to the "strong earnings, broad market" thesis.
Is the market expensive?
UBS argues the S&P 500 forward P/E sits well below its model-implied fair value.
Yardeni Research notes that the S&P 500 PEG ratio — price-to-earnings divided by earnings growth, a measure of how much you pay per unit of growth — has fallen to its lowest level in over thirty years.
This means → from these two firms' vantage points, the index may be at record levels, but relative to the current pace of earnings growth, it is not expensive.
Is all of this enough to sustain the next leg?
Earnings beats, buyback restart, macro improvement, reasonable valuations — four tailwinds arriving simultaneously, holding up a broader market rally as the AI trade retreats.
The unresolved question: have these positives already been fully priced in? If the market is at a "peak good news" threshold, any single data point that disappoints could trigger a pullback.
Put simply = all the good news is already on the table. To keep rising, the market needs new good news. To stall, all it takes is one small disappointment.
Content is for reference only, not financial advice.