AI spending at Microsoft and Oracle is even higher than it looks, thanks to this accounting maneuver

Big Tech companies, including Microsoft, Oracle, Amazon, and Meta, are increasingly utilizing finance leases to fund massive AI infrastructure buildouts, effectively understating their true capital expenditures and managing free cash flow impact. This accounting strategy, exemplified by Microsoft's 70% increase in finance-lease liabilities to $46.2 billion and Oracle's $4 billion in new lease liabilities, reveals a significantly higher capital intensity for AI investments than headline figures suggest, with adjusted capex-to-sales ratios for Microsoft and Oracle rising to 38% and 58% respectively for FY2026 when leases are included.

Analysis

Major technology hyperscalers are increasingly employing finance leases to fund extensive AI infrastructure buildouts, a strategy that obscures the true capital intensity of these investments. This accounting maneuver allows companies like Microsoft (MSFT) and Oracle (ORCL) to avoid large, immediate impacts on free cash flow (FCF) that would result from traditional capital expenditures. For instance, Microsoft's finance-lease liabilities for fiscal-year 2025 surged 70% year-over-year to $46.2 billion, while Oracle recorded $4 billion in new finance-lease liabilities in its latest quarter, up from zero a year prior. This practice is particularly critical for Oracle, which is already experiencing negative free cash flow of $362 million while facing a 180% surge in consensus capex revisions. According to Morgan Stanley analysis, including these leases escalates the projected fiscal-year 2026 capex-to-sales ratio for Microsoft from 28% to 38% and for Oracle from 41% to 58%. The trend is sector-wide, with Amazon (AMZN) and Meta (META) also reversing previous strategies to embrace lease financing, evidenced by Meta's lease commitments ballooning to $52.6 billion. While this approach preserves FCF for near-term shareholder returns, it effectively converts upfront capex into long-term debt-like obligations, indicating that the full financial burden of the AI race is being deferred rather than reduced.

Key Decisions for Investors

• Investors must look beyond headline capital expenditure and free cash flow figures, scrutinizing balance sheets for finance lease liabilities to accurately assess a company's total investment and long-term cash commitments to AI.
• For Oracle, the combination of negative free cash flow and a heavy reliance on new lease financing presents a heightened risk profile, warranting close monitoring of its operating cash flow generation to ensure it can service future obligations.
• Valuation models based on traditional FCF should be adjusted to account for the deferral of cash outflows, as current metrics may overstate the capacity of companies like Microsoft, Meta, and Amazon to fund shareholder returns.
• Monitor the growth of total lease commitments as a leading indicator of future expenses and liabilities that will eventually impact earnings and constrain financial flexibility across the tech sector.