A new survey from Bank of America reveals a deepening split on Wall Street over artificial intelligence. A record 82% of global fund managers now describe AI as the market's most crowded trade, yet roughly half of those same managers say they do not believe the sector is in a bubble. The divide sets the stage for a tense earnings season, as companies begin reporting second-quarter results that could either validate or undermine the massive bets placed on AI.
The survey, which polls institutional investors managing hundreds of billions of dollars, captures a rare moment of collective unease. Crowded trades are not inherently dangerous, but they become risky when too many portfolios are positioned the same way. When a trade is this popular, even a small disappointment can trigger a wave of selling as funds scramble to reduce risk.
What the AI divide means for markets
In a Reuters Open Interest column, journalist Jamie McGeever argues that the bigger the promises around AI become, the more entrenched both bulls and bears grow. Optimists point to transformative potential and surging demand for data centers and chips. Skeptics focus on the enormous upfront costs and the lack of clear, near-term profits for many companies pouring money into the technology.
Bank of America strategists describe a “wealth transfer” playing out in real time. The biggest cloud and consumer tech firms are spending heavily on capital expenditures—large upfront investments in equipment like servers and networking gear—while chipmakers and other suppliers book the revenue today. If that spending continues to rise faster than the returns it generates, even results that look good on paper can disappoint because expectations have been set so high.
Because AI exposure is concentrated in a relatively small number of stocks and indexes, crowded positioning can turn modest surprises into outsized moves. When many portfolios own the same names, a slightly weaker-than-expected earnings report or a cautious outlook can trigger synchronized selling. That dynamic can push volatility higher even if the long-term case for AI remains unresolved.
Earnings season as a reality check
The upcoming second-quarter earnings reports will serve as a critical test. Investors want to see whether the heavy spending on data centers and chips is starting to generate real profits. For the bulls, the bar is high: they need evidence that the investment is paying off. For the bears, the focus is on the mounting costs and whether they will ever be recouped.
Bank of America's 82% crowded-trade reading raises the odds of what the bank calls “air pockets”—sudden, sharp drops in prices when the market rebalances. In this environment, the key question shifts from “who is right about AI?” to “who still has room to add risk?”. Earnings do not have to disprove the AI story to move prices; they just have to come in a bit below the most optimistic forecasts.
That is how selloffs can feed on themselves in AI-heavy pockets like semiconductors. The first decline pushes volatility up, which can force more trimming by funds with strict risk controls. The selling concentrates in the same liquid AI-linked names, amplifying the move. South Korea's KOSPI index offers a recent example: its three biggest single-day declines since the 2008 Lehman Brothers bankruptcy have all occurred this year, including a 9% fall on Monday after a sharp drop in SK Hynix, a major memory chipmaker.
What it means for everyday investors
For ordinary investors, the key takeaway is that the AI trade has become a crowded consensus. That does not mean the technology is a bubble or that it will fail, but it does mean that the potential for sharp, sudden reversals is higher than usual. When everyone is betting the same way, the market becomes more vulnerable to surprises.
Investors should pay close attention to earnings reports from companies with significant AI exposure, especially those in the semiconductor and cloud computing space. The results will not just be about profits and losses; they will be about whether the narrative of AI-driven growth can withstand the scrutiny of actual numbers.
The broader market is also watching. As Latin American markets rise as steady oil and soft inflation offset AI jitters, the divergence highlights how AI's influence is not uniform across regions or sectors. Meanwhile, Bank of America sees Computacenter as an AI infrastructure powerhouse, underscoring that the investment theme extends beyond the biggest tech names.
Ultimately, the next few weeks will test whether the AI rally has real legs or whether it has run ahead of fundamentals. For now, the divide on Wall Street is getting harder to ignore, and the earnings season will provide the first major clue about which side is right.
