Goldman Sachs lifts 2026 bond forecast to $2.3T on AI demand

Goldman Sachs expects U.S. dollar investment-grade credit issuance to reach $2.3 trillion this year, up from its earlier estimate of $2.1 trillion, as companies continue raising debt to fund AI investments.

The bank also dialed up its net supply projection from $850 billion to $1.0 trillion. It further asserted that they anticipate a steady, high volume of activity through 2027, setting their USD IG gross supply estimate at $2.4 trillion. 

So far, according to Sachs metrics, 24% of year-to-date U.S. investment-grade bond volume originated from AI-related issuers, pushing the market to historic highs. In stark contrast, AI issuers accounted for a mere 6% of European volume, while total issuance grew by only 2%.

Goldman Sachs anticipates interest rate fluctuations

Goldman Sachs expects a strong pickup in market activity in September. In its Thursday report, it noted, “Summer slowdown [is] proving elusive—driven in large part by AI-related supply—and market participants are expecting a very busy September. ”

At the moment, hyperscalers are increasingly tapping EUR investment-grade credit, a move the bank says may narrow the euro’s relative technical advantage to the dollar. Nonetheless, the financial institution still expects interest rate fluctuations to remain the dominant factor for total returns, following a year-to-date period of negative performance for both currencies. 

Overall, corporate bond spreads have barely budged this year—instead, the jump in U.S. and German government yields has crushed returns. A continuation of current interest rate levels would materially impair returns. However, the firm forecasts an impending decline in yields that should alleviate some pressure, though total returns will likely remain below historical benchmarks. 

More recently, Germany’s 10-year yield touched a peak last seen in 2011, and the U.S. 10-year yield hit its highest level since November 2023. According to Goldman Sachs, the yield spike—amounting to 61 basis points in the U.S. and 52 basis points in Germany—has directly offset a large share of U.S. corporate bond payouts and erased the interest-rate safety net for European corporate debt. 

US companies have about $46 billion of outstanding bonds in the Eurozone

Overall, U.S. companies hold around $46 billion of outstanding bonds in the euro zone, representing a modest share of the overall market but close to 10% of gross new issuance. Amazon and Alphabet have topped the list of corporate issuers this year. 

According to a European Central Bank blog post, U.S. companies might push up interest rates for all kinds of businesses as they hoard debt. 

They are commanding such a large chunk of the bond market that the pain could easily spread to government bonds and international agency debt, the blog wrote.

Moreover, the blog’s authors argued that a massive influx of new bonds from tech giants risks overwhelming investor demand. As supply expectations continue to rise, this overcrowding could force companies across the market to pay higher interest rates to attract buyers, especially since investors have only a finite amount of capital to deploy.

AI investment is reshaping the corporate bond market

The surge in AI-related borrowing reflects the huge capital needed to build and expand AI infrastructure. Hyperscalers and other technology companies are investing heavily in data centers, advanced computing systems, networking equipment, and energy infrastructure to support the rapid growth of AI applications.

Much of this investment requires significant upfront capital, which is why the corporate bond market is a key source of financing for companies to rapidly develop AI capabilities without relying solely on cash on hand.

But investors, in light of the growing issuance among AI-related companies, face both opportunities and risks. Large technology companies generally have strong balance sheets and significant cash flows, making their debt attractive to investors looking for high-quality corporate bonds.

At the same time, the sheer amount of new supply could test how much additional debt the market can absorb without requiring companies to offer higher yields.

The trend could also influence the broader credit market. If major technology companies continue to dominate new issuance, other corporate borrowers may have to compete more aggressively for investor capital. This could push borrowing costs higher for companies with weaker credit profiles, particularly if government bond yields remain elevated.

Goldman Sachs’ higher issuance forecast, therefore, highlights not only the strength of corporate borrowing but also the growing financial footprint of the AI investment cycle.

As companies race to expand their AI capabilities, bond investors will increasingly need to assess whether the expected long-term returns on these investments justify the additional debt they are accumulating.

Goldman Sachs’ Churchman warned

Even as more companies continue to invest in AI, it does not mean the technology is completely risk-free. Earlier, Goldman Sachs’ Marquee head, Chris Churchman, expressed concern that over-reliance on artificial intelligence could erode the analytical capabilities of future finance experts.

He argued that an escalating dependence on automation could dull practitioners’ capacity for foundational logic and independent, raw reasoning.

“There’s a huge danger here that in the era of AI, we outsource our reasoning to these models, and we have cognitive atrophy that stops us from being able to reason from first principles ourselves,” Churchman asserted.

He also stressed that Wall Street needs to adopt AI without destroying its traditional mentorship system. He admitted that the bank is still trying to figure out how to use the technology while still protecting the real-world wisdom that employees only get from hands-on experience.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.