The short answer: Not yet. As of October 26, 2026, the evidence points to slowing expected earnings growth and a fading AI-spending tailwind, not a confirmed, broad-based downgrade of 2027 earnings. In fact, consensus 2027 S&P 500 earnings per share stood at $419.32 as of October 2, up 15.3% since February. The real risk is deceleration, and the signal that would change that is weaker 2027 spending guidance from the hyperscalers (the largest cloud-computing companies, such as Microsoft, Alphabet, Amazon and Meta).
Every few weeks, a headline suggests that the artificial intelligence boom is about to unwind and take corporate earnings down with it. I understand why that story gets attention. But when I talk with clients about it, I find that two very different concerns are getting blended into one.
The first is whether Wall Street is actually lowering its forecasts for 2027 S&P 500 earnings. The second is whether the enormous sums being spent on AI capex (capital expenditures, meaning money spent on long-lived assets like chips, servers and data centers) will generate the returns investors expect. Those questions are related. They are not the same, and treating them as one leads to poor decisions.
Key takeaways
Consensus 2027 S&P 500 EPS (earnings per share) has risen, not fallen, through early October 2026.
What is fading is the pace of growth. Research now expects AI spending to contribute less to earnings growth in 2027.
A slowdown in AI spending would hit chipmakers and data-center suppliers first, though it could also improve free cash flow (the cash left after capital spending) for the companies doing the spending.
Rising depreciation on AI infrastructure can weigh on reported profits even if spending stays high.
The biggest risk to the index is a double hit: lower earnings estimates combined with investors paying a lower multiple for those earnings.
What is the actual risk AI spending poses to S&P 500 earnings?
The risk is that companies spend less on AI infrastructure than analysts currently anticipate. That would slow revenue and earnings growth for semiconductor manufacturers, data-center suppliers and other AI-related businesses.
But there is another side to that ledger that rarely makes the headlines. Lower AI spending could also improve free cash flow for the companies that pull back. One company’s reduced order is another company’s stronger balance sheet.
Three questions I ask before drawing any conclusions
When I assess a story like this, I separate it into three distinct questions:
Are 2027 earnings estimates actually declining? Answering that requires comparing current consensus forecasts with prior estimates, rather than reacting to negative headlines.
Are AI capital expenditures actually being cut? A slower growth rate in spending is not the same as an outright reduction in spending.
Are valuations vulnerable? Even if earnings estimates remain positive, stocks can decline if investors decide the projected growth is too optimistic.
What does the latest evidence say?
As of October 26, 2026, the data point to slowing expected growth and a diminishing AI tailwind, rather than a clearly established downward revision to 2027 earnings.
S&P 500 consensus earnings per share
2027 EPS estimate (as of October 2, 2026) $419.32
Revision since February (from $363.54) +15.3%
Source: State Street Investment Management, using FactSet data. These figures establish the direction of revisions through October 2, not the precise change through October 26.
Three developments sit behind that headline number:
Goldman Sachs sees slower growth ahead. Its September 23 research projected S&P 500 EPS of $415 in 2027, or roughly 11% growth. It estimated that a $250 billion surprise in either direction in 2027 hyperscaler capex could shift S&P 500 earnings growth by about six percentage points. That is how much this single variable matters.
Goldman’s October outlook emphasizes a fading tailwind. Its October 16 outlook expects earnings growth to slow in 2027 as AI capex growth decelerates and depreciation expenses rise. It describes its 2027 forecast as broadly in line with bottom-up consensus (the sum of individual analysts’ company-by-company estimates), suggesting the recent run of upward revisions may be ending.
Current results have not shown a broad collapse. FactSet reported on October 23 that third-quarter earnings growth was tracking at 29.5%, although the average earnings surprise was below its five- and ten-year averages. That is evidence of continued strength, but not proof that 2027 estimates are secure.
Why does AI capex matter so much for 2027 earnings?
The concern is not simply whether technology companies will cut spending. It is whether the market has extrapolated an unusually strong period of AI-driven earnings growth too far into the future. Here is how I think about the range of outcomes:
AI capex continues rising rapidly: Supports chipmakers, networking, data centers and related suppliers.
AI capex rises more slowly: Earnings can still grow, but growth among AI infrastructure suppliers may moderate.
AI capex declines outright: Creates a greater risk of downward earnings revisions for semiconductor and infrastructure companies.
AI investment continues, but returns disappoint: Could pressure technology valuations and, eventually, broader corporate earnings.
The overlooked factor: depreciation
Companies investing heavily in servers, chips and data centers do not expense those costs all at once. They recognize them gradually through depreciation (the accounting charge that spreads an asset’s cost over its useful life). That means today’s spending becomes tomorrow’s expense. Even if AI spending stays high, rising depreciation can weigh on future reported profits.
My assessment: a growth slowdown, not (yet) a downgrade
I would characterize the current outlook as a risk of earnings-growth deceleration, rather than confirmed, broad-based downward revisions to 2027 earnings. That distinction matters for three reasons.
The bar is high. Analysts have already raised their 2027 estimates substantially during 2026. Even a simple stabilization in estimates could make it harder for the S&P 500 to sustain its advance without further earnings surprises.
The risk is concentrated, at least initially. A slowdown in AI infrastructure spending would most directly affect the companies supplying the buildout. The broader index impact would depend on how large the slowdown is, the resulting revisions, and whether AI productivity gains offset weaker infrastructure profits.
Valuation could amplify the impact. If earnings estimates decline while investors also reduce the P/E multiple (price-to-earnings ratio, what investors pay for each dollar of earnings) they are willing to pay, the index faces pressure from both directions.
What a “double hit” looks like in practice
Consider a simple illustration. If forward earnings estimates fell 5% and the market’s forward P/E contracted from 20 to 18, the combined mathematical effect would be a decline of approximately 14.5% in the index, assuming both changes occurred together and all other factors remained constant. This is an illustrative stress scenario, not a forecast. But it shows why modest-sounding changes in two variables can compound into a meaningful drawdown.
What I am monitoring over the coming weeks
Hyperscaler earnings calls and their 2027 capital-spending guidance.
Monthly changes in FactSet’s bottom-up 2027 S&P 500 EPS estimate.
Earnings revisions for semiconductors, hardware and communication services.
Evidence that AI revenue and productivity gains are keeping pace with infrastructure investment.
Whether earnings growth broadens to financials, industrials, health care and consumer companies.
The bottom line
AI capex remains a major swing factor for 2027 earnings. Current research suggests its contribution to earnings growth is likely to fade, but that does not mean aggregate S&P 500 estimates are already being revised downward. The next critical signal will be whether hyperscaler spending guidance weakens enough to turn slower earnings growth into actual negative revisions. Until then, I would treat the headlines with discipline and the data with respect.
Frequently asked questions
Are analysts cutting their 2027 S&P 500 earnings forecasts?
Not broadly, based on the latest available data. Consensus 2027 EPS was $419.32 as of October 2, 2026, up 15.3% from $363.54 in February, according to State Street using FactSet data. What appears to be slowing is the pace of upward revisions.
Is AI spending being cut?
The current research points to AI capex growth decelerating, which is different from an outright cut. A slower rate of increase still means more spending, just less of a boost to supplier earnings than in the recent past.
How much could AI spending move S&P 500 earnings?
Goldman Sachs estimated in September that a $250 billion surprise in 2027 hyperscaler capex, in either direction, could shift S&P 500 earnings growth by about six percentage points.
Why does depreciation matter for AI stocks?
The cost of chips, servers and data centers is recognized over time as depreciation. As those assets come online, rising depreciation expense can reduce reported profits, even if spending stays high.
Can the S&P 500 fall even if earnings keep growing?
Yes. Stocks can decline if investors conclude projected growth is too optimistic and pay a lower P/E multiple. In an illustrative scenario, a 5% drop in forward earnings combined with a P/E contraction from 20 to 18 would equal roughly a 14.5% index decline.
Disclaimer: This article is provided for informational and educational purposes only and does not constitute investment, tax, or legal advice, or a recommendation to buy or sell any security. The views expressed are those of the author as of the date noted and are subject to change without notice. Figures cited are drawn from third-party sources believed to be reliable but are not guaranteed. Past performance is not indicative of future results, and all investing involves risk, including the possible loss of principal. Hypothetical scenarios are illustrative only and are not forecasts. Please consult a qualified financial professional regarding your individual circumstances.