Goldman Sachs: AI Hyperscaler Capex Could Hit $1.4 Trillion by 2027, Far Above $920B Consensus

Markets3 sources·Jun 12

compute-infrastructure earnings-and-markets goldman-sachs

Summary

• Goldman Sachs forecasts hyperscaler AI capital expenditure reaching up to $1.4T in 2027 — 52% above analyst consensus of ~$920B
• Goldman's team led by Ryan Hammond expects spending to be funded by strong operating cash flows plus new investment-grade debt
• Goldman anticipates further equity gains for stocks exposed to the AI infrastructure theme as the spending boom outpaces current market expectations

Adjust signal

Details

Type	Key Point	Context
Financials	$1.4T Goldman 2027 Forecast	Goldman Sachs upper-bound estimate for total hyperscaler AI infrastructure capital expenditure in calendar year 2027
Stat	$920B Analyst Consensus	Current Wall Street consensus estimate for 2027 hyperscaler capital expenditure — approximately $480B below Goldman's projection
Market Impact	52% Above Consensus	Goldman's forecast implies analysts are dramatically underestimating the pace and scale of the AI infrastructure buildout
Strategy	Debt + Cash Flow Funding	Goldman expects investment-grade debt issuances to supplement already-strong hyperscaler operating cash flows to fund the spending surge
Insight	AI Infrastructure Stocks to Gain	Goldman anticipates further equity gains for companies with significant exposure to AI infrastructure as capex exceeds market expectations
Industry Update	3 Spending Bottlenecks Identified	Analysis identifies power, memory chips, and optical networking bandwidth as the three primary constraints limiting AI capex deployment — GE Vernova, Micron, and optical bandwidth plays named as beneficiaries
Financials	$765B Estimated 2026 Capex	Goldman Sachs now estimates $765B in AI infrastructure investment will be made in 2026 alone, with Microsoft, Amazon, and Alphabet as the largest contributors

Financials

$1.4T Goldman 2027 Forecast

Goldman Sachs upper-bound estimate for total hyperscaler AI infrastructure capital expenditure in calendar year 2027

Stat

$920B Analyst Consensus

Current Wall Street consensus estimate for 2027 hyperscaler capital expenditure — approximately $480B below Goldman's projection

Market Impact

52% Above Consensus

Goldman's forecast implies analysts are dramatically underestimating the pace and scale of the AI infrastructure buildout

Strategy

Debt + Cash Flow Funding

Goldman expects investment-grade debt issuances to supplement already-strong hyperscaler operating cash flows to fund the spending surge

Insight

AI Infrastructure Stocks to Gain

Goldman anticipates further equity gains for companies with significant exposure to AI infrastructure as capex exceeds market expectations

Industry Update

3 Spending Bottlenecks Identified

Analysis identifies power, memory chips, and optical networking bandwidth as the three primary constraints limiting AI capex deployment — GE Vernova, Micron, and optical bandwidth plays named as beneficiaries

Financials

$765B Estimated 2026 Capex

Goldman Sachs now estimates $765B in AI infrastructure investment will be made in 2026 alone, with Microsoft, Amazon, and Alphabet as the largest contributors

Goldman Sachs AI infrastructure spending forecast vs. analyst consensus, with 2026 bottleneck and stock analysis

What This Means

Goldman Sachs is making a bold contrarian call that Wall Street is substantially underestimating the scale of AI infrastructure investment. A potential $480 billion gap between Goldman's $1.4 trillion forecast and the $920 billion consensus for 2027 hyperscaler spending would, if realized, rank among the most significant capex underestimates in technology history. This projection validates the thesis that the AI buildout is a sustained multi-year supercycle rather than a near-term peak, with prolonged demand for chips, data centers, and power infrastructure. Analysis identifies power supply (GE Vernova), memory (Micron), and optical bandwidth as the three specific bottlenecks where spending will concentrate — and where the most investable opportunities lie.

Sources

Update history (2)

Jun 22Added Motley Fool analysis linking to event. New article identifies $765B as Goldman's 2026 AI capex estimate and names three specific spending bottlenecks: power (GE Vernova), memory (Micron Technology), and optical networking bandwidth. This added 2026 datapoint and bottleneck/stock-pick context was integrated into the tier3 detail table and key_facts; the core 2027 forecast narrative ($1.4T vs $920B consensus) remains unchanged.

Jun 14Added corroborating source: Motley Fool analysis citing Goldman Sachs $1.4T capex forecast as basis for stock picks in Nvidia, AMD, and Micron. No new information about Goldman's forecast itself; source linked to increase coverage count.

Similar Events

Goldman Sachs Forecasts AI Capex Surge to Continue Through First Half of 2028

Aug 31

Big Tech Has Already Spent $1.1T on AI Infrastructure Since 2023; Plans $745B More in 2026 as Analyst Projects $1T+ for 2027

Jul 23