AI in credit markets as investment races ahead of corporate profits

We went into our conversations with Robeco’s sector analysts expecting to hear about changing business models, widening competitive gaps and clear winners and losers. What we found was rather more surprising. Or rather, more mundane. And perhaps more interesting because of it.

AI adoption is widespread. Companies are spending money and experimenting with the technology and as a result processes are changing. But outside the sectors directly exposed to the infrastructure buildout or disruption, there is still surprisingly little evidence that AI is materially changing companies’ bottom lines.

That is striking given what is already happening in credit markets. A dominant theme has been the marked uptick in hyperscalers coming to market to fund development. Data center buildouts are expensive projects, and a significant share of the financing is expected to come from investment grade markets and, increasingly, high yield.
The impact is already visible in investment grade benchmarks, where technology and AI-linked issuers are becoming a much larger part of the universe. Hyperscalers alone now account for just over 5% of the US IG benchmark, while the share is considerably higher once AI-related infrastructure and utilities are included.

Perhaps none of this is particularly remarkable. The money being committed to AI is enormous, but so are the expectations. We have been told AI will transform business models and replace large parts of the workforce. Employees everywhere have been left wondering which jobs will still exist in a few years’ time, while companies are under increasing pressure to explain their AI strategy.

Behind all that investment and borrowing sits the assumption that eventually, the money will show up in company bottom lines. So far, that is proving much harder to see. What is obvious is that the AI funding cycle is already reshaping bond markets, but the transformation of the companies within them is proving much slower.

To understand why, we spoke with our financials and consumer analysts about what they are actually seeing at company level. Their experience suggests that, for now at least, the impact of AI is less about wholesale disruption and more about gradual changes in efficiency, customer interaction and the way existing businesses operate.

In financials, AI spending is already visible, but so far it is concentrated largely on improving existing processes rather than creating new sources of revenue. According to Robeco credit analyst Atif Ali, banks are applying AI across many areas from software development to know-your-customer processes, with the immediate opportunity primarily about efficiency and cost reduction.

Scale matters. Larger US banks have considerably greater capacity to invest than smaller European peers, potentially giving them an early advantage in modernizing systems and lowering costs. Atif expects some of that advantage to accrue to the largest players, particularly where competitors are constrained by legacy technology or smaller investment budgets. But for bondholders, the significance is much less dramatic: a bank generating somewhat higher profits as a result of AI does not necessarily become a materially different credit.

There are also natural brakes on disruption. Financials is highly regulated, meaning areas such as automated underwriting need to satisfy rules around fairness and model risk before adoption can accelerate. This is likely to make change more gradual than in less regulated industries.

That is reflected in Atif’s credit research today. AI spending and strategy are increasingly part of the discussion with companies, but the benefits have yet to come through materially in earnings or credit quality. Even where AI delivers meaningful cost savings, Atif expects much of this to be competed away rather than retained as higher profits, with customers ultimately benefiting through better pricing. For now, he does not view the absence of a leading AI strategy as a reason in itself to avoid a bank or insurer.

The picture is similar across consumer-facing companies. Machine learning is not new to the sector: hotels, cruise operators and other businesses have long used it for pricing, recommendations and demand forecasting. The more recent development is generative AI, which is starting to change how consumers research products and services and interact with companies.

Travel provides a good example. Booking and Expedia² are integrating with AI platforms, while hotel groups are introducing AI-powered trip-planning and customer-service tools. The customer journey is already fragmented across search engines, social media, online travel agencies (OTAs) and companies’ own channels. Generative AI and increasingly agents that can complete bookings themselves, could change the interface through which consumers research and arrange travel, but it is not yet clear whether this will materially alter existing competitive positions or simply provide another route into established platforms. While agents rely on existing platforms for supply, the likelier outcome is a reallocation of marketing and distribution spend rather than a material change in the sector’s underlying economics.

Credit analyst Yadeesh Moorthy sees no significant impact on the credit profiles of consumer-facing companies so far. AI adoption is beginning to generate measurable efficiencies at some of the larger platforms, mainly through the automation of customer service, but the overall financial contribution remains difficult to isolate and returns on wider adoption remain uncertain. There is not yet clear credit differentiation between winners and losers. Over time, companies with stronger balance sheets and greater financial flexibility may be better placed to invest and adapt, potentially widening the gap with smaller or more highly leveraged competitors.