AI Stocks Echo Dot-Com Era: CAPE 38, Profit Reality Differs

The Shiller CAPE ratio for the S&P 500 currently sits between 38 and 40, a level exceeded only once in 155 years of data: the dot-com peak of 44.19 in March 2000, just before the Nasdaq collapsed 78% over two and a half years. Meanwhile, the top 10 companies in the S&P 500 now command 36% to 40% of the index’s market cap, nearly 50% higher than the dot-com era’s concentration of 27%. A Deutsche Bank survey shows 57% of institutional investors now flag an AI valuation crash as the biggest market risk. Jeremy Grantham, the GMO co-founder who predicted both the dot-com and housing busts, sees “slim to none” odds that the current AI rally avoids a similar fate. These statistics make the comparison to 2000 feel unavoidable. But they tell only part of the story.

The structural similarities between today’s AI rally and the dot-com bubble are not superficial; they are mechanical. Market concentration has blown past dot-com levels. The Nasdaq-100’s performance hinges on a few firms whose valuations depend on AI revenue growth that has not yet materialized at the scale priced in. Hyperscaler capital expenditure from Microsoft, Google, Amazon, and Meta is on track to reach $660 billion to $690 billion in 2026, the largest corporate investment program in history outside wartime. This spending is partly funded by shifting labor to infrastructure: Meta and Microsoft together cut up to 23,000 jobs while committing record capex, effectively moving money from payroll to data center construction.

Bank of America’s Savita Subramanian projects a year-end S&P 500 target of 7,100, with a bear case of 5,500, anticipating multiple compression as earnings growth slows in late 2026. The Motley Fool points to four classic bubble signals: retail investor euphoria, speculative capital concentration, valuations detached from fundamentals, and a narrative so compelling that skepticism feels disreputable. All four are present. OpenAI’s $852 billion valuation prices a profitless company at roughly double Coca-Cola’s market cap, even though Coke has generated profits continuously since the 1890s. Accel’s $5 billion AI fund, the largest venture capital vehicle ever, exemplifies the flood of private capital into AI. Public and private markets feed each other: venture-backed AI startups raise at extreme valuations, public AI companies spend aggressively to stay ahead, and the cycle pushes both valuations and capex higher.

The critical difference between 2000 and 2026, however, is profitability. At the dot-com peak, tech firms were collectively destroying capital. Cisco traded at 200 times earnings. Pets.com had no earnings. The thesis rested entirely on future revenue from an internet economy still years from generating the cash flows the market discounted. In 2026, the companies leading the AI rally are among the most profitable in history. Nvidia reported net income over $120 billion for fiscal 2026, with a forward P/E of about 41, elevated but nowhere near Cisco’s 200. The tech sector’s aggregate forward P/E is roughly 30, versus 50 at the 2000 peak. Apple, Microsoft, Alphabet, Amazon, and Meta generated a combined $350 billion in free cash flow in their latest fiscal years. These are cash-generating machines, not speculative ventures burning venture capital. They have chosen to reinvest at historically unusual rates.

Capital Economics analyst John Higgins offers a nuanced distinction between a “stock bubble” and a “fundamental bubble.” The stock bubble may already be deflating: the Nasdaq-100 corrected over 10% from its February 2026 highs before rebounding on trade optimism and strong earnings. But the fundamental bubble, built on actual earnings growth, is still expanding. Nasdaq-100 earnings rose 19% year-over-year in the most recent quarter. As long as AI-related revenue keeps growing at that pace, elevated multiples are justified. The bubble pops not when P/E ratios are high, but when the “E” stops growing. JPMorgan sees the S&P 500 reaching 8,000 if earnings momentum persists. Goldman Sachs forecasts a multi-year AI supercycle. The bull case is not that valuations are reasonable today, but that earnings growth will make them look reasonable in retrospect, an argument that was wrong about Cisco in 2000 but right about Amazon.

The variable that will determine which analogy holds is capital expenditure returns. Hyperscalers are spending $660 billion to $690 billion this year on AI infrastructure. Meta’s $27 billion deal with Nebius for AI cloud capacity is just one of many transactions, each larger than most companies’ entire capital budgets. The question is not whether this infrastructure will be used; it almost certainly will. The question is whether it will generate returns that justify the investment at the price paid. The fiber-optic cables laid in 1999 carry today’s internet. The companies that laid them went bankrupt. The technology was correct. The financial model was not.

There are structural reasons to believe the AI capex cycle is better supported than the fiber-optic buildout. Cloud computing operates on a consumption model where customers pay for usage, providing revenue visibility that speculative fiber networks lacked. The hyperscalers building the infrastructure are also its primary consumers, reducing the demand uncertainty that destroyed independent fiber companies. Oracle’s $553 billion in remaining performance obligations, Microsoft’s Azure backlog, and Amazon’s AWS contract pipeline all represent committed future revenue. But committed revenue is not collected revenue. The concentration of AI demand among a small number of large model developers and enterprise customers creates fragility. If OpenAI, the anchor tenant of Oracle’s Stargate project, faced financial trouble, the ripple effect through the infrastructure financing chain would be severe. If enterprise AI adoption plateaus at the “copilot” stage without progressing to autonomous agent workflows that justify the next order of magnitude in compute spending, the return on $660 billion in annual capex would fall below the cost of capital.

Both sides of the debate are correct, which makes the current moment so difficult to navigate. The bears are right that market concentration, CAPE ratios, and speculative euphoria have reached or exceeded dot-com levels. The bulls are right that the underlying companies are profitable in ways their dot-com predecessors were not. The resolution depends on a variable neither side can observe directly: the long-term return on the hundreds of billions invested in AI infrastructure this year. If those returns materialize, current valuations will be seen as fair prices paid early for a genuine technological transformation. If they do not, the CAPE chart will add a second peak to match March 2000, and the comparisons that feel alarmist today will feel prescient.

The Federal Reserve’s benchmark rate sits at 3.50% to 3.75%, providing less cushion than the near-zero rates that inflated asset prices between 2020 and 2022 but not the restrictive rates that typically trigger corrections. Section 122 tariffs of 10% to 15% on a range of imports expire on July 24, 2026, and their renewal or escalation will affect corporate earnings forecasts and consumer spending. The trajectory that brought technology markets to this point, a year of accelerating AI investment, record venture funding, and corporate restructuring around artificial intelligence, has created conditions that resemble a late-stage expansion more than an early-stage bubble. Late-stage expansions can last longer than skeptics expect. They also end more abruptly than optimists imagine. The honest answer to whether AI stocks are in a bubble is that the question cannot be answered until the capex cycle produces results, and the capex cycle has barely begun. Grantham is betting it ends badly. Goldman is betting it does not.

The market is pricing in both possibilities simultaneously, which is why it has been volatile in both directions, and will remain so until the revenue either arrives or does not.