Goldman Sachs Forecasts $1 Trillion AI Capex in 2027
1h ago|4 min readStandard
Fazen Markets Editorial Desk
Collective editorial team · methodology
ai-capexgoldman-sachssemiconductors
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.
Research from Goldman Sachs projects AI-related global capital expenditure will surge past $1 trillion by 2027. The investment bank's analysis, citing historical technology investment cycles, anticipates a massive build-out of data center, semiconductor, and power infrastructure. The forecast arrives as technology leaders race to secure supply chains and expand computing capacity. As of 13:55 UTC today, Goldman Sachs stock traded at $1,016.46, down 1.51% on the day within a range of $1,000.45 to $1,017.61.
Large-scale technology investment cycles have historically been driven by fundamental shifts in computing architecture. The internet boom of the late 1990s catalyzed billions in telecom and enterprise software spending. The smartphone era, beginning in earnest after Apple’s iPhone launch in 2007, triggered a decade of mobile infrastructure and app ecosystem investment.
The current macro backdrop features elevated interest rates, with the Federal Funds rate above 5%, which typically dampens speculative capital spending. However, the tangible productivity gains demonstrated by generative AI models have overridden traditional cost-of-capital concerns for leading firms.
The immediate catalyst is the unprecedented demand for AI training and inference compute, which exceeds available data center capacity. This has created a multi-year backlog for critical components like high-bandwidth memory and advanced networking chips. Corporate earnings calls in 2025 and 2026 have consistently highlighted AI infrastructure as the top capital allocation priority, signaling sustained commitment.
The $1 trillion annual capex figure represents a compound annual growth rate exceeding 70% from estimated 2024 levels of approximately $150 billion. This projected spend would equal roughly 1% of forecasted 2027 global GDP. For comparison, total global semiconductor industry revenue reached $580 billion in 2025.
A significant portion of this expenditure targets data center construction. Industry analysts project a need for over 100 gigawatts of new global data center power capacity by 2027, a near-doubling from 2024 levels. This power demand alone could require over $200 billion in electrical grid and generation investment.
The capital intensity mirrors prior super-cycles but at a larger scale. Telecom capital expenditure peaked at roughly $350 billion annually during the dot-com bubble. Cloud infrastructure spending by hyperscalers reached a collective peak of about $180 billion in 2025 before the AI acceleration.
Goldman Sachs equity is down 1.51% in today's session, underperforming the broader S&P 500 index. The bank’s analysis suggests AI capex could contribute 0.5 to 1.0 percentage points to US GDP growth annually through the end of the decade.
Direct beneficiaries are semiconductor equipment makers like Applied Materials and ASML, chip designers NVIDIA and AMD, and memory producers SK Hynix and Micron. These companies are critical to building the physical AI hardware stack. Secondary beneficiaries include utilities and industrial firms involved in power infrastructure build-out, such as Eaton and Quanta Services.
A counter-argument is that such aggressive forecasts assume continuous, unimpeded technological scaling. Physical constraints, including silicon fab construction timelines, chip packaging capacity, and energy availability, could delay this spending or cap its ultimate magnitude. Geopolitical tensions affecting Taiwan Semiconductor Manufacturing Company’s supply chain present a material risk.
Positioning data shows institutional investors have been accumulating shares in the semiconductor capital equipment sector since late 2025, anticipating a multi-year order cycle. Flow analysis indicates capital rotating out of consumer discretionary and traditional software names into hardware and infrastructure-focused equities.
Key catalysts include quarterly earnings reports from major cloud providers beginning in mid-July 2026. Guidance on 2027 capital expenditure plans from Microsoft Azure, Amazon AWS, and Google Cloud will validate or challenge the trillion-dollar forecast. TSMC’s monthly sales data, released around the 10th of each month, provides a leading indicator for advanced chip demand.
Levels to watch include the Philadelphia Semiconductor Index (SOX) holding above its 200-day moving average, currently near 4,200 points. A sustained break below this level could signal weakening conviction in the capex cycle. Another critical metric is the book-to-bill ratio for semiconductor equipment, which has remained above 1.2 for six consecutive quarters.
Retail investors gain exposure primarily through exchange-traded funds or individual stocks. Broad semiconductor ETFs like the VanEck Semiconductor ETF offer diversified access. The forecast implies sustained revenue and earnings growth for underlying holdings, but valuations are already elevated. Investors should focus on companies with pricing power and protected intellectual property, not just thematic exposure.
The dot-com bubble was characterized by speculative investment in business-to-consumer internet companies with unproven revenue models. Current AI investment is predominantly driven by established, cash-rich corporations like Microsoft and Meta building tangible infrastructure. The capital is directed toward physical assets—data centers and chips—rather than marketing and customer acquisition.
The primary spenders are the hyperscale cloud providers: Microsoft, Amazon, and Google. Meta Platforms is also a major investor in AI infrastructure for its advertising and social products. These four companies are estimated to account for over 60% of total AI capex in 2026. Their collective spending is driving the entire supply chain, from chip designers to cooling system manufacturers.
Historical tech adoption patterns support an unprecedented wave of AI infrastructure investment, with capital allocation decisions happening now.
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.