AI Infrastructure Capex to Hit $7.6 Trillion, Goldman Sachs Says
1h ago|4 min readStandard
Fazen Markets Editorial Desk
Collective editorial team · methodology
artificial-intelligenceinfrastructure
Fazen Markets Editorial Desk
Collective editorial team · methodology
Trades XAUUSD 24/5 on autopilot. Verified Myfxbook performance. Free forever.
Risk warning: CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. The majority of retail investor accounts lose money when trading CFDs. Vortex HFT is informational software — not investment advice. Past performance does not guarantee future results.
Goldman Sachs announced on 4 June 2026 that capital expenditure from four major technology firms is projected to reach $5.3 trillion by the end of 2030, exceeding the GDP of Japan. Total industry spending on artificial intelligence infrastructure, encompassing data centers, power grids, and computing hardware, is estimated to hit $7.6 trillion over the coming five years. This projection represents a substantial upward revision from the firm’s prior models, driven by an accelerated buildout schedule. The four hyperscalers—Meta, Microsoft, Amazon, and Alphabet—are on track to spend up to $725 billion on capex in 2025 alone, more than double the $360 billion allocated for 2024.
This capital expenditure forecast arrives during a period of heightened market volatility for technology stocks, as seen in the day's selloff. The scale of projected investment is unprecedented in the technology sector, rivaling the infrastructure booms of the early industrial era. The last comparable surge in corporate capex was the telecommunications fiber-optic buildout of the late 1990s, which saw over $1 trillion invested before the sector contracted. A key catalyst for the revised forecast is the rapid commercialization of generative AI models, which require orders of magnitude more computing power than previous technologies. This has forced a fundamental recalculation of the physical plant needed to support widespread AI adoption.
The projected $5.3 trillion in cumulative capex from Meta, Microsoft, Amazon, and Alphabet would surpass the current nominal GDP of Japan ($4.2 trillion), the United Kingdom ($3.1 trillion), India ($3.9 trillion), and France ($2.9 trillion). The 2025 target of $725 billion represents a 101% year-over-year increase from estimated 2024 spending. This growth rate far outpaces the average annual capex growth of 15% these firms exhibited between 2020 and 2023. For perspective, the combined 2025 capex of these four companies alone would be greater than the total market capitalization of many national stock indices. As of 04:44 UTC today, these stocks exhibited significant volatility: Microsoft traded at $427.34, down 7.20%, Alphabet traded at $358.99, down 4.62%, and Amazon traded at $250.02, down 4.30%.
| Metric | 2024E | 2025E | Growth |
|---|---|---|---|
| Hyperscaler Capex | $360B | $725B | +101% |
| Total AI Infrastructure Spend | N/A | Annual Run Rate ~$1.5T | N/A |
The capital allocation shift has profound second-order effects across global markets. Semiconductor manufacturers like NVIDIA and AMD are direct beneficiaries of the computing hardware demand. Power utility and electrical grid companies stand to gain from the massive energy requirements of new data centers, which can consume as much power as a mid-sized city. Conversely, high capex depresses near-term free cash flow for the hyperscalers, raising investor concerns about returns on invested capital. This has contributed to the sharp price corrections in their stocks. Goldman notes that private markets are increasingly funding the buildout, with private data center construction accelerating materially. Institutional flow data indicates hedge funds are building long positions in semiconductor capital equipment and renewable energy ETFs while shorting the hyperscalers.
Key catalysts will determine the pace of this investment cycle. The Federal Reserve's policy decision on 17 June will influence the cost of capital for these extensive projects. Second-quarter earnings reports from Microsoft on 22 July and Amazon on 24 July will provide updated capex guidance for 2025 and beyond. Markets will monitor the 50-day moving averages for hyperscaler stocks as a technical signal for institutional sentiment; a breach below could signal further de-rationing. The level of long-term Treasury yields, particularly the 10-year note, remains critical as it dictates the discount rate for these multi-year investment payoffs.
The AI infrastructure buildout is a primary driver of soaring electricity demand in the United States. Data centers require constant, massive power loads for computing and cooling. Goldman's analysis implies a doubling of data center power consumption by 2030, which will strain existing grid infrastructure and accelerate investment in nuclear, natural gas, and renewable generation. This trend directly benefits utility stocks and power equipment manufacturers.
Private equity and infrastructure funds have become major players in financing data center construction. These entities provide capital for the physical assets—land, buildings, power substations—which are then leased to the technology hyperscalers. This model allows tech firms to keep some capex off their balance sheets while enabling private investors to capture stable, long-term yields from essential digital infrastructure.
Historically, massive investment cycles often lead to periods of overcapacity and diminished returns. The risk is that the rush to build AI infrastructure outpaces actual demand for AI services, resulting in underutilized data centers. Goldman's report acknowledges this concern, noting that investors are questioning the long-term profitability of such immense expenditures, which is reflected in the recent selloff of cloud and AI-related stocks.
AI infrastructure investment is scaling to levels that reshape global capital and energy markets.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. CFD trading carries high risk of capital loss.
Vortex HFT is our free MT4/MT5 Expert Advisor. Verified Myfxbook performance. No subscription. No fees. Trades 24/5.
Position yourself for the macro moves discussed above
Start TradingSponsored
Open a demo account in 30 seconds. No deposit required.
CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.