Stocks are headed for one last hurrah in the coming year before the AI-fueled rally unravels, according to one research firm.
In recent client notes, Capital Economics said it believed the S&P 500 would likely rally over the next year before running into a steep correction. It sees the benchmark index rising as high as 8,000 in 2026, before falling to 7,000 next year in a drop of about 13% to a “more normal level,” the firm wrote on Monday.
“A more plausible scenario is that the stock market ‘collapses under its own weight’ as investors reassess stretched valuations for major tech firms while still believing in AI’s long-term benefits for the real economy,” Jennifer McKeown and William Jackson, two economists at the firm, wrote, adding that as much of a 30% correction was possible in the worst-case scenario.
There are a few reasons the firm sees a double-digit drop brewing:
1. Valuations become overstretched
Valuations are high but not excessive, given how strong corporate earnings are. Yet, tech stocks are pricey and could keep rising, Thomas Mathews, the head of markets at the firm, wrote on Monday.
Mathews pointed to the price-to-forward 12-month earnings ratio of the US tech sector, one measure of how highly stocks are valued. That ratio briefly rose to its highest level since the dot-com bubble last year, before easing, per Capital Economics’ analysis.
LSEG/Capital Economics
“We suspect that as the rally progresses, a rise in valuations is more likely than not. Indeed, even though they are, in our view, not yet excessive, it wouldn’t be surprising if investors’ enthusiasm for AI tech were to push valuations to a level perhaps above what might be sustainable in the long run,” Mathews said.
2. Tech earnings could slow rapidly
While the backdrop for earnings is still positive, earnings growth in the S&P 500 has largely been concentrated in the tech sector, a key vulnerability to valuations.
“Admittedly, it is, perhaps, possible that a ‘bubble’ could emerge in expectations for earnings, rather than valuations,” Mathews speculated.
“It’s possible that rapid earnings growth at tech companies will, also, eventually come back to earth, even if that doesn’t seem set to happen this year,” he added.
3. The US economy could slow
The AI rally could also be jeopardized if the US economy starts to meaningfully slow, Mathews said, though he noted that economic growth looked healthy at the moment and the risk of a recession looked “remote.”
4. AI demand could falter, face competition from China
There’s also the risk that AI doesn’t deliver as much as investors originally anticipated, or that China could start to edge ahead of the US on the technology frontier.
Mathews pointed to lingering concern about AI capex, a key issue that has hung over the market and triggered multiple sell-offs in the past year.
Concerns about China’s AI innovations, like DeepSeek, also sparked a steep sell-off in early 2025.
“That, at face value, poses a big threat to the big US tech companies,” Mathews said of the threat from China, though he pointed to factors like US companies having a “hardware edge” that could protect the tech sector.
5. The rally faces geopolitical risks
Markets have already gotten a taste this year of how geopolitical tensions can dent the market’s enthusiasm for AI. The firm pointed to recent market volatility stemming from President Donald Trump’s threats on Greenland.
Geopolitical conflict could have a more significant impact on AI trade down the line, considering that Europe is a major market for the US tech sector, it added. The US’s allies and partners in Europe make up around 40% of revenue among tech firms, per Capital Economics’ analysis.
LSEG/Capital Economics
“But our sense is that for those to have a lasting effect on the AI rally, they’d have to either have a sizeable effect on the US economy or on tech companies’ earnings through some other channel,” Mathew added of the risk.
