Artificial intelligence-related equities have been the dominant force behind 2026 market returns, contributing more than 80% of the S&P 500's gains this year. That concentration has prompted debate about whether the surge can continue, but strategists at Jefferies argue the advance is supported by fundamentals rather than by multiple expansion.
Jefferies' quantitative strategy team examined a defined basket of AI-linked names and concluded that the recent performance reflects underlying earnings momentum. The team noted that if AI is removed entirely from the index, the S&P 500 would be up only 2% year-to-date, underscoring how concentrated the rally has been.
Forward-looking profit expectations for the AI basket have increased substantially: 2026 forward earnings estimates have risen by more than 30% since mid-2025. On a consensus basis, analysts project an earnings-per-share compound annual growth rate of 38.5% for 2026-27 for the AI group, compared with a projected 11.9% CAGR for non-AI sectors.
Despite rapid forecast upgrades, valuation measures do not appear stretched across the basket. Jefferies reports the group trades at roughly 25 times forward earnings, a level below its own one standard deviation mark, and the basket carries a price-to-earnings-growth ratio of about 0.6 times. "AI is the cheapest sector to own in the U.S.," the team led by Desh Peramunetilleke wrote, referring to PEG metrics.
Performance within the AI theme has been uneven. Sub-themes that have outperformed this year include AI servers, optical components, and memory. By contrast, hyperscalers and chip designers have lagged behind. Jefferies' screening on PEG valuation shows memory and compute stocks as the most attractively valued sub-themes, while semiconductor capital equipment and chip design names appear relatively expensive.
Corporate results in the first-quarter earnings cycle provided further support for the view that earnings are the engine of returns. According to Jefferies, 86% of companies beat earnings expectations in the quarter, a post-COVID high and an increase from 75% in the prior quarter. Sales beats were also notable, reaching 82% of companies reporting.
However, the strategists cautioned that positive surprises were not uniformly rewarded across the market. Outside of AI and a limited number of other sectors, companies that reported beats generally did not see meaningful outperformance following their results, while misses tended to be punished more severely. The team highlighted that, on a brighter note, beats tended to be followed by analyst upgrades, which suggests earnings downside risk is limited despite elevated geopolitical uncertainty.
Jefferies also examined tone and sentiment in corporate communications. Using the AlphaSense platform to analyze roughly 330 earnings calls, the team found management optimism registered at 95%, and analyst sentiment became more positive year-on-year, with 58% of calls displaying a positive tone versus 48% in the fourth quarter of 2025.
Geopolitical developments emerged as a recurring concern. The U.S.-Iran conflict was cited as a negative factor by 44% of companies analyzed, and firms flagged supply chain disruptions and weakening consumer sentiment among their specific worries.
Outside the AI and commodity sectors, the broader revision picture is more muted. Aggregate S&P 500 earnings revisions over the past three months amount to 6%, but that aggregate change falls to just 0.3% when AI and commodity sectors are excluded, indicating that most of the upward revision activity has been concentrated in those areas.
In sum, Jefferies' quant team interprets the 2026 AI-driven rally as being largely earnings-based, with pockets of attractive valuation on a PEG basis and an uneven distribution of performance across sub-themes. The strategists flag geopolitical risk, supply-chain vulnerability, and consumer weakness as recurring concerns cited by company managements, while also noting that beats have generally been followed by upgrades—pointing to limited near-term earnings risk.