28 May 2026
By Chris Iggo, Chair of the Investment Institute and CIO for AXA IM Core, part of BNP Paribas Asset Management
The excitement engulfing artificial intelligence has been very much focused on equity markets — with the likes of the tech-heavy Nasdaq index hitting multiple fresh highs over the past 12 months.
However, fixed income markets have also been positively influenced by the AI boom.
Apart from the overall impact on investor sentiment and AI’s macroeconomic implications, there are numerous channels through which this new technology affects bond markets.
Here we discuss three: growing technology company issuance, disruption to corporate issuer business models, and the potential for enhanced investment processes resulting from AI developments.
The concentration of technology company shares in equity indices is well documented. The same dominance does not exist in fixed income markets, however, with debt issued by technology companies having a limited share in bond indices.
According to Bank of America/ICE, the Technology and Electronics sector makes up just 4.8% of its Global Credit index. And despite the focus on capital spending in the technology sphere, this share has remained constant in recent years. In the US investment grade market, the share is around 7%. It it is lower in Europe and, globally, represents just 5% of the high yield space.
However, issuance is on the rise. There have been several large bond deals in the last year from so-called ‘hyperscalers’ (the cloud computing behemoths), who are engaged in huge capital spending to support AI’s infrastructure expansion.
In the US, the face value of corporate bonds issued by technology/electronics companies rose by 11% in the year to March to some $685bn. Based on corporate guidance, there should be a lot more coming as the capital expenditure largesse shows no sign of slowing.
Most of this borrowing is occurring in the US. But issuers have tapped the euro and sterling investment grade markets, too. Google parent Alphabet rolled out its largest ever bond sale — a multi-tranche sterling issue — in February, including a 100-year maturity bond with a coupon of 6.125%.1 Elsewhere, Facebook owner Meta borrowed $30bn via the US market last year, while Oracle has been a prolific borrower.2
The tech giants’ share price performance underpins their quality. These are issuers with very strong earnings growth, robust balance sheets with limited debt, and strong credit ratings. The sector’s growth should be welcomed by investors as it means greater diversification in bond indices, which tend to be dominated by financials and more cyclical industrial companies.
While thematic investing is less common in bond markets, the growth in issuance from high quality, well-rated technology companies could find support from investors. Coupons are attractive and backed by very strong earnings growth. For investors who are already exposed to technology through their equity holdings, bonds issued from the key AI drivers who provide potentially more predictable returns should appeal.
Evidence highlighting how AI can disrupt business models by automating tasks, improving efficiency and speeding up processes is plentiful.
Earlier in 2026, the focus was on software companies and how certain application providers — client relationship management systems, for example — might see their products undermined by AI’s ability to code and build products more quickly. The technology ecosystem is complex, and smaller companies might find themselves more at risk of disruption. This has been a concern in the high yield, leveraged loan and direct lending markets, which have been a source of funding for smaller companies in this space.
Again, bond markets are less concentrated than equity markets. The technology sector accounts for only around 5% of the US high yield market with software around 3.5%. As concerns about the AI challenge to software services emerged at the start of the year, credit spreads in this sector widened.
However, investors are working hard to differentiate between those issuers that are more and less vulnerable to seeing their businesses threatened by AI — and not just those that reside in the technology sector itself. Sectors such as gaming and media could also see disruption. For active fixed income investors, the focus is on avoiding issuers where cashflows could be challenged and therefore where ratings could potentially suffer and default risk increase.
But again, this should not be overdone. It is still a relatively small sector, and not all technology companies are the same. Exposure to technology is more concentrated in the leveraged loan and direct lending markets than in the public high yield market.
Fixed income research, trading, and portfolio construction can all be enhanced by AI. In credit markets, the use of large language models (LLMs) and natural language processing can significantly improve fundamental analysis of issuers and detect information that can potentially be used to generate alpha and distinguish the value amongst issuers.
Extracting signals from bond issuers that can generate material investment decisions at the issuer and portfolio level can be made easier by AI. Trading has already been largely automated, something which has helped improve liquidity in corporate bond markets in recent years.
For portfolio managers, more powerful research through AI and the ability to use the technology to isolate material pricing information will be beneficial to how risks and the liquidity profile of a fixed income portfolio are managed.
Greater tech firm bond issuance because of AI’s rise will give fixed income investors access to the cashflows generated by this technological revolution. In the long term, the value rests on whether the assets being created by this investment (and which ultimately back the value of the bonds), can continue to generate sufficient return on capital to maintain strong balance sheets and healthy credit ratings, and high multiples and returns in the equity market.
Given where the world is today, we believe that betting against this revolution could appear to be a potentially risky approach to take.
Performance data/data sources: LSEG Workspace DataStream, ICE Data Services, Bloomberg, BNP Paribas AM, as of 20 April 2026, unless otherwise stated).
Past performance should not be seen as a guide to future returns.