How Credit Markets Are Financing AI Infrastructure

Nov 17, 2026

Rising demand for capital to finance AI infrastructure and M&A is driving record bond issuance and new financing structures, even as borrowing costs rise.

Key Takeaways

  • Technology companies are increasingly tapping credit markets to fund AI infrastructure, supported by potential returns on invested capital of 25% to 50%.
  • Rising demand for financing led U.S. investment-grade corporate bond issuance to reach $2.2 trillion through September 2026, up nearly 25% year over year.
  • Credit markets are likely to finance more than half of approximately $3.8 trillion in projected hyperscaler spending from 2026 through 2028.
  • Growing connections between public and private credit, alongside new pools of institutional capital, are expanding financing options for companies investing in AI.
  • Despite higher bond yields, credit spreads remain near multidecade lows; capital availability is a more pressing issue for AI investment than borrowing costs.

Credit markets are undergoing a significant transformation as companies fund the buildout of artificial intelligence infrastructure and a pickup in mergers and acquisitions (M&A).

 

Financing for these investments is shifting from corporate cash flows and equity markets toward debt markets, creating demand for more complex financing structures and new sources of capital, noted investors at Morgan Stanley’s 2026 Credit Symposium.

 

Through September 2026, U.S. investment-grade corporate bond issuance reached $2.2 trillion, up nearly 25% from the same period in 2025. Meanwhile, capital flowing through insurance companies, annuity investments and private credit is expanding the pool of investors available to finance corporate borrowing.

 

This surge in activity comes amid broader economic and geopolitical challenges. Global growth has remained resilient despite persistent inflation, rising government deficits, geopolitical tensions and substantial financing needs across the public and private sectors.

 

At the same time, the AI capital expenditure cycle is creating new financing challenges. The interconnected relationships among hyperscalers, chipmakers, neocloud providers and data center operators are driving innovation in how companies structure and fund their investments.

 

“Financing solutions have expanded, and public and private credit have become more interconnected than ever before,” said Jay Hallik, Morgan Stanley’s Global Co-Head of Fixed Income. “Yet the importance of relationships, connectivity and great ideas has only grown.”

AI has moved from being a technology story to an equity story than a credit story, and increasingly, a macroeconomic story. The scale of investment is enormous, and there are many reasons for optimism. Usage is growing rapidly. More companies are reporting measurable benefits. However, the infrastructure required is enormous.

 

We're financing chips, data centers, power generation, transmission utilities and the broader supply chain. We're not simply financing a technology anymore. We're financing a whole ecosystem. That ecosystem involves public bonds, private credit, project finance, structured finance, securitization, commodities and hedging. It's an enormous opportunity for capital markets and for Morgan  Stanley. At the same time, some of the biggest constraints are no longer financial; they’re physical power availability and permitting, to name a few.

Insights

Jay Hallik

Co-Head of Fixed Income, Morgan Stanley

AI Investment Drives Demand for Credit

The scale and pace of AI infrastructure investment are contributing to record debt issuance by technology companies seeking to capture the potential economic returns from the technology.

 

According to Morgan Stanley Research's analysis of existing AI computing infrastructure, returns on invested capital range from 25% to 50%, depending on the business model.

 

For companies seeking to finance large-scale investments, credit markets offer a deep, liquid source of capital that can accommodate substantial borrowing needs.

 

“AI data center financing is an area of both incredible growth and incredible innovation within credit markets,” said Katy Huberty, Morgan Stanley’s Global Director of Research. “The returns have to justify the spend.”

We were looking at our AI CapEx forecast for 2026 a year ago. It's 74% percent higher. And even through the early part of this third quarter earnings season, the hyperscalers that we've heard from are underwriting, both their spending commitments, but also the fact that they have to tap into the full spectrum of capital markets to fund those investments.

 

And if there is one difference this cycle, it's the sheer scale that is creating both capital and physical constraints, like power. And that adds a whole new layer of complexity.

Insights

Katy Huberty

Global Director of Research, Morgan Stanley

New Financing Structures Support the AI Buildout

The scale of anticipated AI investment is also changing how companies access capital.

 

At the symposium, Morgan Stanley credit specialists discussed how public and private credit markets are evolving to support the development of the AI ecosystem.

 

Vishwanath Tirupattur, Chief Fixed Income Strategist and Director of Quantitative Research at Morgan Stanley, estimates that credit markets will finance more than 50% of hyperscalers' spending from 2026 through 2028, with total spending now projected at approximately $3.8 trillion.

 

Meeting that demand will require a range of financing channels and structures, reflecting the different capital needs and business models across the AI ecosystem.

 

The growing connections between public and private credit markets could play an important role in meeting those needs as issuers seek financing solutions tailored to increasingly complex investments.

Vishwanath: If you had this conversation two years ago or a year and even a year and a half ago and looked at would the depth, the breadth and the complexity of various credit market channels to handle this. And we would have probably not even conceived this.

 

But the key underlying premise for all of this is the demand for compute vastly exceeds the supply of compute. This demand-supply imbalance is there today, and we expect it to persist for quite a number of years. This demand-supply imbalance for compute creates the opportunity set for, for example, the hyperscalers to look at a return on invested capital that is well north of 25%, 25%, 40% type of return on invested capital.

 

I mean that that is in line of sight for them in line to monetize and realize that you need this massive amount of infrastructure build out to enable that and, and therefore that strong continued commitment to doing that.

 

Bill: You're starting to see real scale, real depth of markets. And whether it's the investment grade market, the anonymous grade market, the project finance, the bank lending market, but nevertheless, the scale of the opportunity is so large that there's no boat big enough for all the capital needs here.

 

And then from a complexity perspective, I would say we are Morgan Stanley, we've been trying to innovate new solutions and try to keep them as simple as possible and as transparent as possible. We have had to try to structure the transactions to narrow the risks or to isolate the risks for people.

 

Rich Myers: What's changed, obviously, is the scale. But also the kind of singularity of the offtakes and the contracts such that you have now single projects with single tenants that are in the tens of billions of dollars.

 

And so rather than kind of real estate credit or securitized products credit in the sense of diversification, these have now moved fundamentally into more corporate credit type underwrites because of that, you know, non-diversified single credit nature of where the cash flow is coming from.

Research

The Capital Behind the AI Infrastructure

Vishwanath Tirupattur, Bill Graham and Rich Myers

Capital Availability Takes Priority Over Borrowing Costs

Credit markets have remained resilient despite a substantial increase in bond yields in 2026. Credit spreads—the additional yield investors demand to hold corporate debt rather than comparable government bonds—remain near multidecade lows.

 

Speakers at the symposium suggested that higher borrowing costs are unlikely, on their own, to derail the AI investment cycle. Instead, the availability of capital may be a more important constraint for companies seeking to fund large-scale infrastructure projects.

 

“If you are a hyperscaler looking at a 25% to 50% return on invested capital, it doesn’t look like a major impediment if the cost of financing goes from 5% to 6% or even 7%,” Tirupattur said.

Research

Cost of Capital and Returns

Vishwanath Tirupattur, Rich Myers and Bill Graham