KKR's Ming Lu Warns the Market Has Never Been This Concentrated in One Bet, and AI's Payoff Could Take Decades

Ming Lu, a senior advisory partner at the private equity firm KKR and its former executive chairman for Asia Pacific, has warned that the U.S. stock market has never placed a bet this large on a single technology, and that the real payoff from artificial intelligence is likely decades away rather than the next few years many investors appear to be pricing in. He has described the current level of concentration as a vulnerability of historic proportions.

The numbers behind that warning are striking. The so-called Magnificent Seven technology companies now account for about 35% of the S&P 500's total value, a combined market capitalization of roughly $25 trillion and the highest concentration in the index since 1929. AI-related capital spending has grown to around 2% of U.S. GDP, up from close to zero five years ago, and AI investment alone accounted for 39% of U.S. economic growth in early 2025, a larger share than the dot-com boom ever reached at its peak of 28%. AI companies captured 81% of all global venture funding, about $300 billion, in the first quarter of 2026, and AI-related bonds have grown from 2% to 10% of all investment-grade corporate borrowing since 2023.

Lu draws a direct comparison to the railroad boom of the late 1800s, when railroad stocks made up more than 80% of trading on the New York Stock Exchange in 1885, a concentration that preceded the Panic of 1893 and a depression that lasted roughly four years. He is careful to note that earlier waves of speculative overbuilding, in canals, electrification, and telecommunications, all eventually proved worthwhile despite painful crashes along the way, suggesting he sees AI following a similar arc rather than ending in a permanent bust.

Part of what concerns Lu is how the spending is being financed. Hyperscale cloud companies are planning to spend roughly $800 billion this year alone on AI infrastructure, and a Federal Reserve Bank has flagged the risk of what it calls circular financing, arrangements in which more than half of the investment flowing into AI companies comes from other AI companies rather than from outside capital.

Lu's broader point is less about whether AI eventually pays off, which he does not dispute, and more about how much risk is concentrated in getting there. With wealthier households, the top 20% of earners now account for 60% of U.S. consumer spending, already carrying an outsized share of the economy even before accounting for how much of their wealth sits in AI-driven stock gains, a dependence that could leave both markets and the broader economy exposed if the payoff takes as long as he expects.