The rise of artificial intelligence (AI) and its impact on software is disrupting private debt markets, but it is also creating opportunities in infrastructure credit, according to AlbaCore Capital Group.
For credit investors, the implications of the idea that “software is dead” are “nuanced”, said David Allen, managing partner and chief investment officer at AlbaCore Capital Group. AI has the potential to enhance credit quality through productivity gains and efficiencies, while also presenting opportunities in infrastructure debt.
However, the other side of the argument is that AI could undermine long-term enterprise value by intensifying competition and disrupting business models.
For the positives, Allen explained that the expansion of data centre capacity is acting as a catalyst for a broader wave of infrastructure investment. The growth of data centres is driving demand for supporting assets such as power generation and grid connections, while also increasing the need for wider digital infrastructure.
“These adjacent assets often exhibit long asset lives, contracted or regulated cash flows and lower technology obsolescence risk, making them particularly well suited to a defensive infrastructure debt strategy,” he said.
However, concerns over the impact of software have also rattled the market, with volatility spilling into credit markets. The market reaction has spread to business development companies exposed to software, with the likes of Apollo, Ares, KKR and TPG seeing a sell-off in their listed vehicles in February.
Read more: Software sell-off sparks credit fears, but experts say debt is safe
Allen noted that spreads have widened since the software concerns emerged, yet software fundamentals remain among the healthiest in sub-investment grade credit markets. However, while fundamentals currently appear robust, loan-to-value ratios could become a bigger problem.
As the sector evolves, Allen said underwriting AI risk within software companies requires investors to consider a broad range of factors in combination, acknowledging that their interplay will shape a business’s resilience and adaptability in the face of ongoing AI-driven change.
Companies that own large, unique datasets have a clear advantage, as AI systems cannot easily recreate that data if it is not publicly available.
“Businesses that develop a substantial proprietary database, typically built over an extended period, often rely on an active, self-reinforcing ecosystem of data submitters and users. It is difficult for AI alternatives trained on only public data to replicate the value of these offerings.”
Allen also pointed out that the significance of proprietary data in mitigating AI risks extends beyond software companies to other industries.
Read more: Can AI address private credit fraud?
