Publications

Credit analysis should use AI and abductive reasoning
- June 1, 2026: Vol. 20, Number 6

To read this full article you need to be subscribed to Institutional Real Estate Europe

Credit analysis should use AI and abductive reasoning

by Marek Handzel

Real estate credit analysis needs to include abductive reasoning and make more use of large language models to create a more honest, forward looking view of credit risk.

Speaking on a podcast recorded by Institutional Real Estate, Inc, Vincent Noble, head of asset-based lending at Federated Hermes, said credit providers have so much analytical power at their disposal now that they should use it extensively to dig as deeply as possible, and as early as possible, to understand deal risk.

Federated Hermes has begun using what Noble calls a structured pre-mortem alongside traditional analysis in order to deepen its analysis by not relying too heavily on a small number of criteria.

“We’re trying to dig beyond the metrics and say, what is really driving the risk?” said Noble.

“The best analysis starts with asking the right questions, and for real estate in particular, where even two buildings — that on the face of it seem similar and are maybe located

For reprint and licensing requests for this article, Click Here.

Forgot your username or password?