Capital Economics Warns AI Stock Market Bubble Nearing End

The S&P 500 is expected to fall to 6,500 by the end of 2027, a 21% decline from its current levels, as a range of bubble indicators suggest the AI-driven equity boom is nearing an end [2]. While the index may see an 8% gain through the end of 2026, analysts at Capital Economics warn that medium-term prospects are deteriorating due to extreme market froth [1, 2].

Indicators of a market peak

Capital Economics senior market economist James Reilly identified eight key metrics to track the potential bubble, noting that several have reached levels not seen since the dot-com era [2]. Earnings expectations stand out as the primary warning sign, with long-term earnings-per-share growth forecasts hitting record highs [2]. Because this expected growth is heavily concentrated in the tech sector, any weakness in tech earnings could trigger a significant index-wide correction [2].

Other structural indicators are also flashing red. Index concentration—the degree to which a few large companies drive the market—is at its highest point in half a century, with the five largest firms holding up 30% of the S&P 500 [2, 3]. Additionally, foreign ownership of U.S. stocks has reached a record high, and net equity issuance has turned positive [2]. Historically, booms in IPOs and share sales have often coincided with market peaks, leading analysts to suggest the bubble’s end may be months away rather than years [2].

The cost of the AI build-out

The sustainability of the AI rally faces scrutiny over the massive capital expenditures required to maintain infrastructure. In late 2025, U.S. mega-cap companies were projected to spend $1.1 trillion on AI between 2026 and 2029 [3]. Critics, including JP Morgan CEO Jamie Dimon, have questioned whether this massive investment will yield sufficient returns, noting that while AI is a "real" technology, some current capital deployment will likely be wasted [3].

Financial strain is already visible in the sector. OpenAI, a central player in the AI boom, has been projected to run out of cash by mid-2027, with annual losses expected to continue through 2028 [3]. Furthermore, a recent study from the National Bureau of Economic Research found that 90% of firms have yet to see a measurable impact from AI on workplace productivity, fueling comparisons to the historical "productivity paradox" [3]. While volatility and leverage metrics currently appear less alarming than other indicators, they are trending in a direction that analysts describe as concerning [2].

What to watch

• Tech Earnings Reports: Monitor for any signs of weakness in the tech sector, as the index's heavy concentration makes it uniquely vulnerable to earnings misses among the largest firms [2].
• Equity Issuance: Watch for a new wave of IPOs and share sales, which historically serve as a signal that a market peak is imminent [2].
• Infrastructure Spending: Track the actual return on investment for the $1.6 trillion in total AI spending expected through 2029 to see if it justifies current valuation multiples [3].

The central question remains whether the massive capital outlays by tech giants will translate into tangible profitability or if the current market valuation—trading at 23 times forward earnings—is built on unsustainable hype [3]. With the Bank of England already warning of risks to global market stability, the divergence between AI-driven growth expectations and actual productivity gains remains the primary tension point for investors [3].