Barclays retains equities as its top asset-class preference for Q3, saying an earnings cycle driven by AI investment in the U.S. is likely to dominate worries about high market valuations.
The bank increased its 2026 S&P 500 earnings-per-share forecast to $337 from $321 and raised its price target to 7,800, citing clearer visibility on technology earnings, reflationary forces that should support nominal revenue growth, and a comparatively constructive backdrop for industry activity. Barclays also projects S&P 500 earnings of $389 in 2027, which the bank notes would represent an additional 15% rise.
"Expensive is not the same thing as wrong," strategists led by Ajay Rajadhyaksha wrote. "Equities can be fully valued and still outperform if the earnings trajectory holds." They added that the only circumstance in which equities should be sold is a collapse in earnings growth, and they see no plausible catalyst for such a collapse in the current quarter.
Sectors with the clearest leverage to the AI investment cycle are identified by Barclays as technology, media and telecoms, industrials and utilities. At the same time, the bank downgraded financials to neutral, citing concerns around private credit and potential disruption from AI. Healthcare was upgraded to neutral, while consumer stocks remain a negative exposure in Barclays' view because of persisting pressures on spending power that could intensify in the second half of the year.
Looking beyond the U.S., Barclays singled out Japan as the best risk-adjusted way for investors to gain AI exposure. The bank points to the Nikkei 225's broader sector composition relative to markets such as Korea and Taiwan, and to Japan's ongoing structural reforms, which include rising dividends and the unwinding of cross-shareholdings.
More guarded stance on sovereign debt
The strategists expressed a significantly more cautious view on government bonds. They note sovereign issuance across OECD countries is on track to set a new record this year, that the buyer base has shifted away from central banks toward more price-sensitive investors, and that the term premium on long-dated U.S. Treasuries has already climbed to its highest level since 2011.
On that basis, Barclays now expects 10-year U.S. Treasury yields to rise to 4.65% over the next year and recommends investors remain underweight long-duration government bonds globally.
The bank also highlighted higher potential macro volatility as a theme following the Federal Reserve's first meeting under Chair Warsh, which Barclays said revealed "a more inflation-focused reaction function" and a clearer intent to reduce forward guidance. That shift, the strategists argue, makes monetary policy less predictable and more reactive to individual data releases.
"We believe markets are just starting to price this transition. A wider distribution of policy outcomes, greater event risk around FOMC meetings and uncertainty around the Fed's balance sheet review should all contribute to structurally higher rate volatility," the strategists wrote.
Market context and related notes
The bank's views come alongside data points and market tickers referenced around U.S. equities, Japanese equities and U.S. Treasury yields. Barclays' sector calls and rate views form the basis of its recommendation to favor equities with specific sector tilts while reducing exposure to long-duration sovereign obligations.
Promotional note included in the source material: The text referenced an AI-focused investment product and performance claims tied to a 'Tech Titans' strategy, which was described as doubling the S&P 500 within 18 months and noted individual winners cited in the original material.
Barclays' combination of upgraded earnings forecasts, targeted sector preferences, a recommendation to underweight long-dated government bonds and a warning about structurally higher rate volatility outlines a view that favors equity exposure tied to AI investment while urging caution on fixed-income duration risk.