AI Investment Boom Reshapes Global Bond Indices: What Investors Can Expect in 2027

• US hyperscalers have issued almost USD 400 billion in bonds over the past two years.
• Capital expenditure of around USD 5.5 trillion could significantly shift index weights by 2029.
• Debt-weighted indices increase concentration risks, making active security selection increasingly important.

US hyperscalers are rapidly becoming major issuers in the corporate bond market. A new study by asset manager Assenagon shows how the wave of investment in artificial intelligence is changing issuance volumes, maturity profiles and index weights across fixed-income markets. Alphabet, Amazon, Meta, Microsoft, Oracle and SpaceX have issued bonds totalling USD 392 billion over the past two years.

According to market estimates, their capital expenditure could reach approximately USD 5.5 trillion by 2029. Given the steadily growing volume of issuance, Assenagon’s fixed-income experts expect significant shifts in global bond indices. “In the bond market, index mechanics reward debt rather than quality: the more debt an issuer takes on, the greater its weight in the index,” says Robert Van Kleeck, Head of Credit Portfolio Management. He adds: “At Assenagon, we therefore take an active approach, selecting issuers on the basis of their credit quality, valuation and risk-return potential rather than the size of their debt. Our proprietary rating model forms the basis of this process.”
 

Rising Volumes, Longer Maturities and New Currencies

The wave of issuance by the technology giants is changing not only the size of the bond market but also its maturity profile. Forty-one percent of hyperscaler issuance has a maturity of more than ten years. By comparison, the corresponding share in the euro investment-grade market is just 7 percent.

At the same time, financing is expanding beyond the US dollar into nearly all major currency markets. Around EUR 35 billion has already been placed in the euro market. As a result, the new issuance is increasingly finding its way into euro-denominated corporate bond portfolios.
 

Funding Requirements Shift Index Weights

Based on the market estimates used in the analysis, the six companies’ cumulative capital expenditure could reach approximately USD 5.5 trillion by 2029. Even if only around 20 percent of this amount were financed through the bond market, their total debt could rise to approximately USD 1.1 trillion by 2029.

Because traditional corporate bond indices weight issuers according to the volume of debt outstanding, additional funding requirements automatically result in higher index weights. In Assenagon’s model calculation for 2029, SpaceX, with a weighting of 1.87 percent, Amazon, at 1.48 percent, and Oracle, at 1.45 percent, would be the three largest issuers in a global investment-grade ETF. As recently as August 2025, the leading positions were held exclusively by banks.
 

Index Weightings Do Not Reflect Credit Quality

The growing importance of technology issuers says little about their individual credit quality. Assenagon’s analysis reveals substantial differences in debt-servicing capacity, leverage, profitability and return on capital.

The divergence between external and internal ratings is particularly pronounced for Oracle and SpaceX. While agency ratings still place both issuers in the investment-grade segment, Assenagon’s internal fundamental analysis arrives at significantly lower credit assessments. The subsequent widening of both issuers’ credit spreads highlights the importance of identifying changes in leverage at an early stage.
 

Active Security Selection Works at Three Levels

From Assenagon’s perspective, simply replicating future index weights is therefore not sufficient. Three decisions are crucial when managing an active corporate bond portfolio:

• Selecting issuers on the basis of fundamentals, default risk and expected changes in credit quality.
• Identifying the most attractive bond from each issuer in terms of maturity, currency and credit risk premium.
• Sizing positions according to their attractiveness and risk rather than the volume of debt outstanding.

Alphabet provides an example of the differences this approach can uncover. After hedging the currency risk, sterling-denominated bonds offered around 30 basis points more spread than comparable euro-denominated bonds, according to the analysis. By contrast, an index-tracking portfolio generally buys all index-eligible issues, irrespective of whether their relative valuations are attractive.
 

Here you can download a high-resolution photo of Robert Van Kleeck.

München/Frankfurt, 23 September 2026