AI Infrastructure Accounting Inflates S&P 500 Forward Earnings
Major technology companies are using extended depreciation schedules for AI hardware to delay recording costs, creating an $854 billion gap in reported expenses.
A timing mismatch between AI revenue recognition and expense recording is artificially inflating S&P 500 forward earnings. While hardware providers like Nvidia Inc. report immediate gains from AI sales, the purchasing companies are lengthening depreciation schedules to delay recording the associated costs.
Alphabet Inc., Meta Platforms Inc., and Microsoft have extended the estimated useful lives of their servers and network equipment to between 5.5 and 6 years. This accounting shift ensures only a fraction of the initial cash outlay is expensed annually. Alphabet now depreciates such equipment over six years, while Meta increased its estimated useful lives to 5.5 years in January 2025.
This practice creates an estimated $854 billion gap between actual cash spending and reported expenses for 2027. If these outlays were treated as immediate expenses, S&P 500 earnings would drop by approximately 24%. Such a correction would raise the forward valuation multiple from 18.8X to 24.8X, suggesting the index may be in bubble territory.
