Unlocking the Data Center Supercycle with Trade Finance

中文标题

以贸易融资解锁数据中心超级周期

Citi09/28/2026updated 10/17/2026

Unknown economy

Citable research brief

Core conclusion · Tlines extraction

AI-driven data center demand is forecast to generate $3.8 trillion in global AI-related capital expenditure by 2030, but power access delays nearly 40% of U.S. projects, while complex contracts and semiconductor shortages raise costs. Trade finance tools such as standby letters of credit, inventory finance and export agency finance are presented as strategic enablers that help secure power capacity, hardware and infrastructure while preserving liquidity. Citi cites case studies including a $2.5 billion syndicated SBLC facility and a 20-year PPA.

Institution: Citi
Published: 09/28/2026
Research topic: Semiconductors
Time horizon: 2030

Semiconductors · SOX

Key arguments

AI-related capex is forecast to reach $3.8 trillion by 2030, with most directed to data centers.
Power availability is the primary constraint; interconnection approvals can take years and nearly 40% of U.S. projects face delays.
Trade finance structures such as SBLCs, inventory finance and export agency finance can reduce risk, preserve liquidity and accelerate timelines.
Utilities risk stranded assets: roughly 75% of requested interconnection capacity was withdrawn between 2000 and 2020.

Key numbers

Global AI-related capex by 2030: $3.8 trillion
U.S. data center projects delayed: Nearly 40%
Historic interconnection withdrawal rate (2000-2020): ~75%
LC support required per MW from utilities: $1.5 million to $3 million
Example syndicated SBLC facility size: $2.5 billion

Main risks

Power access constraints delaying nearly 40% of U.S. data center projects.
Semiconductor shortages leading to long lead times and prepayment requirements.
Utilities face stranded-asset risk, with about 75% of historic interconnection requests withdrawn.

Corresponding source evidence

•Power access is now the industry’s biggest constraint, delaying nearly 40% of U.S. data center projects.
•Citi’s experience delivering tailored trade finance solutions may offer clients an alternate source of funding that can help secure scarce power capacity, manage semiconductor supply constraints, and accelerate infrastructure buildout while preserving liquidity.
Citi Research estimates AI-related capital expenditure will reach $3.8 trillion by 2030 (see Figure #1) , with the majority directed toward data center development. · 2030

Context: the conclusion is Tlines' automated extraction, not the institution's wording. Quoted evidence is publisher wording. Scope and date must travel with the conclusion.

Verify at the original source ↗