From the studio

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)

Thought of the day

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.

That volatility has also been felt in credit markets. Investment grade (IG) spreads remain near historical tights, and the overall moves at the index level have been orderly. But beneath the surface, pressure is starting to show in high yield (HY) and sector dispersion has risen sharply.

Sectors 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.

Software and services are more exposed. Agentic AI tools are accelerating automation risk, particularly for software and services. These concerns could linger for some time, making it difficult for investors to have conviction on current firms’ growth and profitability. Still, companies with strong competitive moats, regulatory barriers, or entrenched business-to-business relationships are better positioned to adapt. Many large tech firms are actively investing in AI to defend and extend their franchises, while consumer-facing firms with low switching costs remain more vulnerable. The ability to integrate AI, maintain proprietary data, and respond strategically will be key differentiators.

Credit fundamentals, technicals remain healthy. Across most sectors, corporate fundamentals are robust: Balance sheets are strong, cash generation is healthy, and refinancing needs are manageable. The macro backdrop remains supportive, with solid economic growth, stable labor markets, and accommodative policy. Recent large-scale tech bond issuance has been a worry in equity markets—but by contrast has been well absorbed by the credit market.

So, we think credit markets are likely to remain resilient, with these AI-related risks unlikely to spill over into sustained stress. Any wider bouts of AI-linked volatility in credit could be an opportunity for selective buying opportunities, in our view, particularly in high-quality issuers with robust fundamentals. Selective participation in new tech issuance and credits from sectors poised to benefit from AI investment can also enhance portfolio yield and diversification. At the same time, we favor credit exposure to top-rated tech companies and asset-heavy sectors such as utilities, energy, and industrials.

From an equity perspective, we have moved our ratings on US information technology and communication services down to Neutral recently, reflecting a more balanced risk-reward outlook. Read more on our credit views in our recently published report, “AI disintermediation risk: Credit perspective.”