Mundo

27 feb. 2026

Daily: What do AI disintermediation risks mean for credit markets?

Video: The AI Show | NVIDIA earnings takeaways and the Anthropic vs software debate (4 mins) Video: Investors Club | Tech turbulence and trade crosswinds (10 mins) Video: CIO's Jeff Harwood on opportunities in US financials (6 mins) Podcast: CIO's Ulrike Hoffmann-Burchardi on investing in bits vs atoms (5 mins) AI disintermediation risk has emerged as an important market narrative in recent weeks. US software and IT services have been hard hit, with the S&P 500 software index down close to 30% from its fall 2025 high, as rapid advances in agentic AI inject new uncertainty over the terminal value of traditional software models.

Puntos clave

  • 01That volatility has also been felt in credit markets.
  • 02Investment grade (IG) spreads remain near historical tights, and the overall moves at the index level have been orderly.
  • 03But beneath the surface, pressure is starting to show in high yield (HY) and sector dispersion has risen sharply.
  • 04Sectors perceived as most exposed to AI disruption, such as software, services, and insurance brokers, have seen significant spread widening.

Credit from relatively insulated sectors, especially those with more tangible physical assets or moats, like energy, utilities, and capital goods, have outperformed. We think AI disintermediation risks in credit are real. But the more pressing question is how and where this translates into pressure on cash flows, leverage, and refinancing risk: Disruption risks are credible, though credit should be buffered. Certain business models, especially those reliant on easily automated workflows or commoditized services, face credible long-term threats from AI.

However, actual credit impairment typically requires sustained deterioration in earnings and cash flow, which triggers downgrades or refinancing challenges. Most IG and large HY issuers retain strong balance sheets and financial flexibility, making them resilient to near-term disruption. Risks are more pronounced for smaller, highly leveraged HY issuers, but these represent only a small fraction of the broader market. Agentic AI tools are accelerating automation risk, particularly for software and services.

Publicado por UBS

Continuar lectura