AI Boom Faces $10 Trillion Revenue Hurdle, BCA Research Warns

The global artificial intelligence industry may need to generate $10 trillion in annual revenue to justify the immense capital expenditure currently flowing into data centers, according to a new report from BCA Research. The firm's analysis suggests that the market's enthusiasm for AI is overlooking significant long-term monetization challenges and that corporate profit margins are unsustainably high.

Inflated Margins Skew Valuations

BCA Chief Economist Peter Berezin noted that while the S&P 500's forward price-to-earnings ratio of 19x is in line with historical averages, this valuation is propped up by record-high forward profit margins of 16.7%. If these margins were to revert to 2019 levels, the index's forward P/E ratio would jump to a much more expensive 26.7x.

According to the report, these elevated margins are flattered by how major tech companies account for hardware purchases. The massive spending on servers and other equipment by tech "hyperscalers"—including Microsoft, Amazon, Alphabet, Meta, and Oracle—is treated as capital expenditure (capex) that is depreciated over several years, rather than as an immediate operating expense. This accounting treatment artificially boosts near-term profitability.

The Soaring Cost of AI Infrastructure

The scale of this spending is set to create a significant drag on future earnings as depreciation charges mount. BCA projects that annual depreciation for major hyperscalers will more than double, rising from $255 billion in 2026 to $581 billion by 2029.

This is driven by a forecast that total on-balance-sheet capex for these firms will hit $1.16 trillion by 2029. When including off-balance-sheet spending, BCA estimates total annual hyperscaler capex could approach $1.4 trillion by the end of the decade.

Ad

A $10 Trillion Monetization Challenge

To generate a profitable return on this sustained level of investment, the AI industry faces a monumental commercial hurdle. BCA's model assumes a 15% pre-tax return on invested capital and a 30% EBITDA margin, which it views as more realistic than Wall Street's more optimistic forecasts.

Based on these assumptions, the report calculates that:

• Hyperscalers alone would need to generate $7.4 trillion in annual revenue to justify a steady-state capex of $1.4 trillion.
• Factoring in AI spending from China, smaller cloud providers, and private ventures, the entire global AI industry would need to bring in approximately $10 trillion in annual sales.

To put this figure in perspective, Berezin noted that $10 trillion is equivalent to the total annual global expenditure on either healthcare or food, highlighting the immense scale of the monetization challenge facing the AI sector.