US investment-grade corporate bond issuance reaches $1.76 trillion this year, with AI infrastructure financing as the main driver

According to Bloomberg, as of October 26, U.S. investment-grade corporate bond issuance for the year had reached approximately $1.76 trillion, surpassing the $1.75 trillion record set during the COVID-19 pandemic in 2020 and setting a new all-time high for annual issuance.

The surge in investment demand for artificial intelligence (AI) infrastructure is an important reason driving this round of rising bond issuance. At the same time, companies are also locking in funding in advance while financing conditions remain acceptable. However, against the backdrop of rising U.S. Treasury yields and higher financing costs, why are companies still crowding into debt issuance? After issuance hit a record scale, what tests do the bond market's capacity to absorb supply and the return prospects of AI investment face?

Companies Rush to Lock In Funding

The primary driver of this bond issuance boom is the massive capital expenditure brought by tech companies' buildout of AI infrastructure. Leading tech companies such as Microsoft, Amazon and Alphabet continue to increase investment in data centers, computing power and chips, and their funding needs keep expanding, making the bond market an important financing channel.

Bloomberg data show that as of late October, AI-related financing in the global investment-grade bond market had exceeded $300 billion in 2026, significantly higher than the full-year level in 2025. Data center construction involves servers, chips, power and supporting infrastructure, with huge upfront investment, and companies need external financing to maintain their pace of expansion.

Lindsay Rosner, head of multi-asset fixed income investing at Goldman Sachs Asset Management, said the market is currently in one of the most capital-intensive corporate investment cycles in decades, with AI competition pushing companies to continuously expand financing scale.

In addition to meeting investment needs, companies issuing bonds early also aim to lock in financing costs and reduce refinancing uncertainty. Although U.S. Treasury yields are rising, as long as corporate credit spreads remain at relatively manageable levels, issuers may choose to raise funds as early as possible to avoid the risk of further rate increases or wider credit spreads in the future.

Moreover, corporate bond issuance is not entirely for new investment; it may also be used to repay maturing debt or replace existing financing. Therefore, record issuance scale does not mean corporate new debt is growing by the same magnitude, but continued expansion of AI capital expenditure is still putting pressure on corporate balance sheets.

The concentration of corporate bond issuance first tests the market's capacity to absorb funds. Pension funds, insurance institutions and asset management companies are important buyers of investment-grade corporate bonds, but as new supply continues to increase, investors may demand higher yields, and issuers also need to offer more attractive pricing terms. If supply growth exceeds allocation demand, corporate financing costs may rise further.

The rising concentration of AI-related debt in the credit bond market also deserves attention. PIMCO previously noted in an analysis that as of September 23, AI-related issuers accounted for nearly one-quarter of dollar-denominated non-financial corporate bond issuance, up from less than 13% in 2025 and about 4% in 2024. As tech companies continue to raise funds, investors will become more sensitive to changes in AI industry capital expenditure and commercialization progress.

The bigger test lies in whether investment returns can cover new debt. AI infrastructure investment is large and has a long payback period, and companies need to achieve revenue growth through businesses such as cloud services, model invocation and enterprise software. If computing power demand or commercialization progress falls short of expectations, problems such as insufficient data center utilization and increased depreciation expenses may weigh on cash flow, leaving companies to bear both high capital expenditure and interest expenses.

J.P. Morgan Asset Management previously estimated that by 2030, complete data center construction may require about $5 trillion in funding, of which about $2 trillion may be financed through the investment-grade credit bond market. This means AI-related financing demand may persist for a long time, but strong funding demand does not mean investment will necessarily generate corresponding returns.

Risk tolerance also differs among companies. Andy Li, an analyst at credit rating agency CreditSights, noted that leading tech companies maintain high ratings on the back of steady revenue and have ample debt safety buffers; but some mid-rated companies are following the trend of borrowing to expand production, and their debt risk resistance is relatively weak. In an environment of elevated interest rates, their debt repayment pressure will rise rapidly.

The trajectory of interest rates is also key. Analysis points out that the coupon rates of existing fixed-rate bonds will not change due to rising market rates, but companies that need to issue new bonds or roll over borrowings will face higher financing costs. If U.S. economic growth slows and corporate profits decline, combined with increased interest expenses, weaker issuers may face pressures such as wider credit spreads and greater financing difficulties.

Reporter Li Xizi, Intern Luo Zehui