Stocks tied to artificial intelligence have powered one of the most concentrated rallies in modern markets. Yet cracks appeared last week. The Nasdaq tumbled more than 4 percent in a single session. Hundreds of billions in market value vanished from the very names that led the surge.
Chipmakers stumbled. Cloud giants gave back gains. And suddenly the questions that had lingered in the background moved front and center. Is this the beginning of a painful reckoning? Or simply noise in a longer transformation?
Economist Owen Lamont once pushed back against easy bubble talk. His framework tracks four signals: stretched prices, widespread belief in endless growth, heavy equity issuance, and flood of investor money. In February he saw no full alignment. Fortune reported his updated view this week. The surge in earnings forecasts now stands out as the clearest warning. Companies and analysts project rapid profit jumps that justify current multiples. Lamont calls the optimism “pessimistic” in its detachment from reality.
But the spending keeps coming. Hyperscalers plan more than $700 billion in capital expenditures this year alone on data centers and related infrastructure. Total AI-related outlays could exceed $1 trillion between now and the end of the decade. Those numbers come from industry forecasts cited across multiple outlets.
Market Concentration Reaches Rare Levels
Forty-one stocks linked to AI now represent nearly half the S&P 500’s total value. Jim Bianco of Bianco Research delivered that assessment in The Guardian. The concentration exceeds anything seen in recent decades. Neil Wilson at Saxo UK went further. “The entire market has become one giant AI edifice,” he said. A repeat of the dot-com crash remains possible. Years of flat returns could follow.
The comparison feels familiar. In 1999 investors paid premium prices for future promise. Many internet companies never delivered profits. Today the leaders generate real cash flow. Nvidia sells chips as fast as factories can produce them. Microsoft and Google report strong adoption of AI tools in their clouds. Revenue grows. Yet the forward multiples still evoke late-stage dot-com excess.
Goldman Sachs’ James Covello has sounded cautious notes for years. On the bank’s podcast he cut to the core. “At some point, you’ve got to make money.” Investments exist to generate returns. The distance between spending and proven payoff has widened, not narrowed.
And the productivity data raises eyebrows. A National Bureau of Economic Research paper from February found that 90 percent of firms reported no measurable AI impact on workplace output. Executives nevertheless forecast 1.4 percent productivity gains and 0.8 percent output growth. The gap between expectation and experience echoes the famous productivity paradox of earlier computing waves.
Recent market action adds tension. Broadcom issued soft guidance. That news helped trigger the June sell-off. Tech names shed value quickly. Some traders wondered whether the AI story could sustain lofty valuations without rate cuts on the horizon. Inflation readings have stayed sticky. The Federal Reserve holds its powder dry.
Prediction markets reflect the uncertainty. Polymarket bettors assign roughly 24 percent odds that the bubble bursts by the end of this year. The leading scenario points to December 31, 2026. Those probabilities shift daily with new headlines.
Yet bulls see something different. AI infrastructure orders remain robust. hyperscalers continue to issue debt and equity to fund expansion. Smaller players stretch balance sheets more aggressively. One research effort tracks a “redline signal” based on when spending shifts from cash flow to borrowed money. So far the largest companies stay in safe territory. The warning light has not flashed red.
Michael Burry warned in May that markets had “jumped the shark.” His Substack note circulated widely. History shows bubbles can inflate further even after skeptics grow loud. Overvalued assets often become more overvalued before the turn.
Bank of America strategist Michael Hartnett has prepared a post-bubble roadmap. He sees the current setup as vulnerable to a sharp reversal. His team points to the same concentration risks and reliance on continued hype.
The American Prospect took a blunt line this week. Artificial intelligence may prove transformative, the publication argued, but current stock prices could still reflect wild overvaluation. Its analysis highlighted how a small group of platform companies now drives an outsized share of market gains. The parallel to 2000 feels uncomfortable.
Even so, differences exist. Many AI leaders boast real earnings and dominant market positions. The technology solves concrete problems in coding, drug discovery, and customer service. Adoption curves look steep. Enterprise pilots convert to paid contracts faster than expected in some cases.
The debate therefore centers on timing and magnitude. How much future growth already sits in today’s prices? What happens if capital spending peaks sooner than models assume? Hyperscalers cannot increase data-center budgets at current rates forever. Cash flow eventually must cover the outlays.
Recent X conversations show divided sentiment. Some users declare AI the largest bubble in history. Others insist real money flows today separate this cycle from past manias. One prediction market update from Polymarket on Friday showed the 2026 burst probability holding near 22 percent.
History rarely repeats cleanly. The dot-com bust destroyed companies with no revenue and weak business models. This time the core players possess scale, cash, and proven demand for their products. The risk lies in the periphery and in the assumptions baked into multiples.
If earnings growth disappoints, multiples will compress. That adjustment could prove orderly or chaotic. Much depends on whether the broader economy cooperates. Recession fears, geopolitical shocks, or energy price spikes could accelerate any unwind.
Investors therefore watch several markers. Quarterly guidance from chip and cloud companies. Actual AI-driven revenue disclosure. Evidence of measurable productivity gains at scale. Any slowdown in new data-center orders.
The whirlwind has arrived. Valuations face scrutiny they largely avoided during the melt-up. Smart money has not fled en masse. But the easy narrative no longer goes unchallenged. Markets will test whether the AI story rests on solid ground or merely on hope stretched thin.
Resolution may not come in one dramatic pop. Corrections can unfold over quarters. Leadership can rotate without a full bear market. Yet the concentration leaves little margin for error. When a handful of names account for so much index performance, any stumble reverberates widely.
So the question hangs. Have investors priced perfection? Or does the technology’s potential still outrun even today’s ambitious forecasts? The data will decide. And it will arrive company by company, quarter by quarter.
WebProNews is an iEntry Publication