--- title: "Tech Giants' AI Cash Burn Intensity Surpasses Dotcom Bubble Peak; Depreciation to Be Future Focus" type: "News" locale: "en" url: "https://longbridge.com/en/news/289270765.md" description: "Morgan Stanley believes the AI capital expenditure-to-sales ratio of tech giants will reach 44% in 2027, comprehensively surpassing the 32% peak during the Dotcom Bubble. Cumulative depreciation over the next three years will exceed $520 billion, with Oracle's depreciation-to-revenue ratio soaring from 7% to 28%. Coupled with nearly $1 trillion in purchase commitments and over $800 billion in off-balance-sheet lease leverage, depreciation will become a core pressure variable on profit margins" datetime: "2026-06-10T04:13:54.000Z" locales: - [zh-CN](https://longbridge.com/zh-CN/news/289270765.md) - [en](https://longbridge.com/en/news/289270765.md) - [zh-HK](https://longbridge.com/zh-HK/news/289270765.md) --- # Tech Giants' AI Cash Burn Intensity Surpasses Dotcom Bubble Peak; Depreciation to Be Future Focus Wall Street is witnessing an unprecedented capital spending spree. On June 10, according to Zhuifeng Trading Desk, Morgan Stanley's latest research report revealed a stark reality: the "cash burn" intensity of hyperscale cloud computing giants represented by **Amazon, Google, Meta, Microsoft, and Oracle** in the AI sector has officially surpassed the peak levels seen during the millennium Dotcom Bubble. Morgan Stanley predicts that the capital expenditure-to-sales ratio (Capex-to-Sales) of these five giants will reach 36%, 44%, and 42% in 2026, 2027, and 2028, respectively, far exceeding the 32% peak during the Dotcom Bubble. If finance leases are included in the calculation, the actual capital intensity will be even higher. The core warning in the report is that **Microsoft, Oracle, Meta, and Google will face a massive depreciation tsunami exceeding $520 billion over the next three years.** If non-depreciation costs cannot be reduced or revenue expectations are not raised synchronously, this flood of depreciation will ruthlessly devour corporate profit margins. **Currently, the huge amount of "Construction in Progress" (CIP) temporarily masks this impact, but the moment of reckoning for the income statement will eventually arrive.** ## AI Cash Burn Intensity Comprehensively Surpasses Dotcom Bubble Peak Morgan Stanley's core judgment is that the intensity of the current AI capital expenditure cycle has no comparable precedent in history. The firm predicts that the capital expenditure-to-sales ratio of hyperscale cloud computing giants will reach 36% in 2026, climb to 44% in 2027, and remain at 42% in 2028. This figure comprehensively surpasses the historical peak of 32% for the communication services industry during the Dotcom Bubble. Of greater concern is that if finance leases are included in the calculation, capital intensity will rise significantly further. > Taking Microsoft as an example, Morgan Stanley expects its capital expenditure-to-sales ratio, including finance leases, to jump from 33% and 50% under the traditional metric for FY26 and FY27 to 44% and 64%, respectively. Oracle's situation is more extreme, with the ratio soaring from 76% and 115% to 101% and 189% during the same period. In terms of absolute scale, hyperscale cloud computing giants contributed over 150% of the growth in capital expenditure for US large-cap stocks in 2025 and are expected to account for about 40% of the total capital expenditure of the Russell 1000 Index in 2026, double the level of 2024. If AI-related supporting industries such as energy and industrials are included in the statistics, AI-related capital expenditure is expected to account for more than 50% of the total market capital expenditure. At the same time, **the speed of upward revisions to capital expenditure forecasts has significantly exceeded the adjustment magnitude for revenue and free cash flow forecasts, which is the structural contradiction most worthy of investor caution in the current AI investment cycle.** **Over the past nine months, market consensus forecasts for the capital expenditure of hyperscale cloud computing giants for 2026-2027 have been raised by a combined total of approximately $900 billion.** > From an individual stock perspective, the market consensus forecast for Google's capital expenditure in 2026 was raised by 139% compared to a year ago (June 2025); Meta and Amazon were raised by 85% and 81%, respectively; Oracle had the largest increase, reaching 175%. > > However, **at the same time, the adjustment magnitude for revenue forecasts is far less than that for capital expenditure.** Taking changes within one year as an example, the capital expenditure for Microsoft, Oracle, Meta, Google, and Amazon was raised by **approximately $44 billion, $39 billion, $61 billion, $109 billion, and $89 billion**, respectively, while the corresponding upward adjustments for revenue were significantly smaller. This "scissors difference" implies a continuous rise in capital intensity, which will ultimately put pressure on profit margins through rising depreciation expenses. ## Nearly $1 Trillion in Purchase Commitments and Over $800 Billion in Lease Commitments Constitute Massive Off-Balance-Sheet Leverage In addition to capital expenditures already reflected on the balance sheet, hyperscale cloud computing giants have accumulated massive off-balance-sheet obligations through purchase commitments and lease commitments, which are risk exposures easily overlooked by the market in the current AI investment cycle. **Regarding purchase commitments**, as of the latest disclosure, the combined purchase commitments of Google, Microsoft, Meta, Amazon, Nvidia, and Oracle have approached $982 billion, close to $1 trillion. > Among them, Google is the highest at $332 billion; Meta is $238 billion; Amazon is $155 billion; Microsoft is $142 billion; Nvidia is $104 billion; and Oracle is $11 billion. > > Morgan Stanley believes that unless companies expect these contracts to result in losses, these obligations will not appear on the balance sheet until the relevant goods or services are delivered and accounts payable are recognized. **Regarding lease commitments**, the total value of leases not yet commenced by hyperscale cloud computing giants has exceeded $822 billion, with Google at $261 billion, Amazon at $197 billion, Meta at $183 billion, Microsoft at $106 billion, and Oracle at $76 billion. In addition, operating leases and finance leases already recognized on the balance sheet amount to **$179 billion and $86 billion, respectively.** The firm believes that the existence of these off-balance-sheet commitments means that the actual operating leverage borne by hyperscale cloud computing giants is far higher than what is presented in financial statements. Meanwhile, the mismatch between the timing of AI monetization and supplier payments is driving up days payable outstanding (DPO) for various companies. Currently, the unpaid capital expenditure embedded in accounts payable and accrued expenses for major AI players totals approximately $110 billion. ## Depreciation: The Core Variable for Next Phase of Margin Pressure Morgan Stanley explicitly lists depreciation expenses as the "next key item to watch for margin pressure." Morgan Stanley estimates that **the cumulative depreciation expenses for Microsoft, Oracle, Meta, and Google over the next three years (FY26-FY28) will exceed $520 billion.** As the proportion of depreciation expenses to revenue continues to rise, companies must rely on synchronous decreases in other expense items or significant revenue growth to hedge against this in order to maintain profit margin expectations. From an individual stock perspective, the pressure is most prominent for Oracle and Meta: **Oracle's depreciation expense-to-revenue ratio is expected to rise from 7% in FY25 to 28% in FY28; Meta's ratio is expected to jump from 9% to 19%.** Currently, a large amount of capital expenditure remains in the "Construction in Progress" (CIP) account, **not yet transferred to fixed assets and beginning depreciation.** This accounting mechanism objectively delays the impact of capital expenditure on net profit and profit margins, but this delay does not mean the disappearance of pressure, but rather its accumulation. Data shows that the CIP balances for Oracle, Meta, and Google grew by **approximately 200%, 90%, and 55%**, respectively, over the past year. As these projects are completed and transferred to fixed assets, **depreciation expenses will accelerate in the coming years, and the impact on profit margins will be more concentrated and significant at that time.** ### Related Stocks - [ORCL.US](https://longbridge.com/en/quote/ORCL.US.md) - [ORCX.US](https://longbridge.com/en/quote/ORCX.US.md) - [MS.US](https://longbridge.com/en/quote/MS.US.md) - [AMZN.US](https://longbridge.com/en/quote/AMZN.US.md) - [GOOGL.US](https://longbridge.com/en/quote/GOOGL.US.md) - [GOOG.US](https://longbridge.com/en/quote/GOOG.US.md) - [META.US](https://longbridge.com/en/quote/META.US.md) - [MSFT.US](https://longbridge.com/en/quote/MSFT.US.md) - [NVDA.US](https://longbridge.com/en/quote/NVDA.US.md) - [MS-O.US](https://longbridge.com/en/quote/MS-O.US.md) - [MS-Q.US](https://longbridge.com/en/quote/MS-Q.US.md) - [MS-E.US](https://longbridge.com/en/quote/MS-E.US.md) - [MS-I.US](https://longbridge.com/en/quote/MS-I.US.md) - [MS-L.US](https://longbridge.com/en/quote/MS-L.US.md) - [MS-P.US](https://longbridge.com/en/quote/MS-P.US.md) - [MS-A.US](https://longbridge.com/en/quote/MS-A.US.md) - [MS-F.US](https://longbridge.com/en/quote/MS-F.US.md) - [MS-K.US](https://longbridge.com/en/quote/MS-K.US.md) - [ORCL-D.US](https://longbridge.com/en/quote/ORCL-D.US.md) - [NVD.DE](https://longbridge.com/en/quote/NVD.DE.md) ## Related News & Research - [Oracle faces pivotal Q4 earnings with AI growth in focus](https://longbridge.com/en/news/289265936.md) - [Oracle Q4 Earnings Preview: Expect AI Demand to Push ORCL Stock Even Higher](https://longbridge.com/en/news/288884508.md) - [Apple stock: Buy, sell, or hold?](https://longbridge.com/en/news/288947515.md) - [Forget AI Models. This Company Sells What Every AI Giant Needs.](https://longbridge.com/en/news/289086501.md) - [What Bezos Knew In 1999 That AI Investors Are Missing Today](https://longbridge.com/en/news/288932329.md)