Morgan Stanley Dominates Wall Street With Explosive $2.3B AI Haul

Morgan Stanley collected $2.3 billion in debt and equity capital-markets fees tied to the artificial intelligence infrastructure boom in the first half of this year. That figure compares to $1.4 billion in the same period last year, a 64% jump that placed the bank ahead of every Wall Street rival in the AI fee race.

The result, reported by the Financial Times on Sunday, cements (MS) Morgan Stanley as the architect of a new class of financing deals funding the GPU clusters, power plants, and fiber networks that the AI buildout demands.

Morgan Stanley AI Fees Surge as Banks Race to Fund the Buildout

Morgan Stanley AI fees jumped because the bank moved early to design structures that other lenders were not yet offering.

It devised new debt and equity models suited to the unusual balance sheets of AI infrastructure companies, which often carry massive capital commitments before generating revenue.

The deals included investment-grade bonds for data center developers, convertible notes for AI chipmakers, and equity raises for hyperscale tenants signing long-term compute contracts. Each structure required bankers to translate the forward-looking economics of AI workloads into terms that fixed-income investors could underwrite.

Traditional project finance, used for decades on power plants and toll roads, works by securing debt against contracted cash flows.

AI infrastructure deals adapt that logic. A data center signs a 10-year lease with a hyperscaler, that contracted revenue becomes collateral, and a bank like Morgan Stanley packages the resulting cash flow into rated bonds that pension funds and insurers can buy.

The bank earns a fee at origination and often again if it underwrites subsequent equity.

Because AI tenants have stronger credit ratings than most real-estate lessees, the bonds price tightly, meaning low yields for buyers and high deal volumes for the banks arranging them.

How a $1 Trillion Capex Cycle Created a Fee Windfall

The numbers behind the fee boom are striking. The five largest US hyperscalers collectively disclosed more than $300 billion in combined capital expenditure for this year alone, a figure analysts at several banks estimate will exceed $400 billion once off-balance-sheet commitments are included.

A Nikkei analysis circulated on Sunday put total off-balance-sheet AI liabilities at US tech giants at $1.65 trillion.

Every dollar of that spending passes through a capital-markets structure of some kind. Construction debt, equipment leasing, tax-equity partnerships, and public equity rounds all require underwriters.

The firm’s edge is that it identified this pipeline earlier than rivals and staffed accordingly.

The bank’s technology-sector banking team, already strong from years of software IPOs, was repositioned toward infrastructure issuers starting in late 2024. That meant relationships with the independent power producers and data center REITs that hyperscalers rely on for physical capacity.

Those relationships paid out in the first half of this year as the pace of deal issuance accelerated.

From Software IPOs to Debt Architecture

Morgan Stanley AI fees did not emerge from a single product line. The bank combined capabilities from its project finance, leveraged finance, and equity capital markets desks to build composite deal structures that no single desk could execute alone.

A typical transaction might start with a private credit raise to fund construction, followed by a rated bond take-out once the facility is operational, then an equity raise for the sponsor once cash flows are established.

The firm earns fees at each stage, and its ability to stay in a transaction from groundbreaking to stabilization gives it an advantage over banks that specialize in only one layer.

Rivals have taken notice. Goldman Sachs, JPMorgan, and Barclays all expanded headcount in infrastructure and energy transition finance over the past 18 months.

But the bank’s first-mover position in AI-specific debt structures gave it a pipeline that will take competitors time to replicate.

When the AI Funding Wave Could Slow

The fee windfall carries concentration risk. If hyperscaler capex guidance softens in the second half of this year, deal volumes will fall faster than the revenue lines of diversified investment banks.

Morgan Stanley’s 64% fee growth implies that AI deals now represent a meaningful share of its total capital-markets revenue, which means a capex pullback would hit the bank asymmetrically.

The geopolitical backdrop adds uncertainty. US-Iran escalation pushed Brent crude up almost 4% on Sunday, and Asian chip stocks remained under pressure from the Moonshot AI model shock that rippled through markets on Friday.

A sustained risk-off episode could widen credit spreads and delay bond issuance across the AI infrastructure universe.

For now, the pipeline remains full. Power purchase agreements signed in 2025 are rolling into construction financing this year, and data center lease signings that closed in Q1 are generating bond mandates in Q3.

The bank’s bankers are positioned to collect fees on deals that were set in motion months ago, even if new mandates slow.

The deeper story is structural. Wall Street has spent two decades building expertise in software company IPOs.

The AI infrastructure boom is shifting the fee pool toward asset-heavy businesses that look more like utilities than startups. Morgan Stanley read that shift early, and the $2.3 billion figure is the first clean evidence that the bet paid off.

Read Next: Morgan Stanley Opens Spot Crypto To 8.6 Million E*TRADE Clients, Coinbase Just Got A New Rival