Classic LGD models typically suffer from low predictive power, making life difficult for modelers in the attempt to reach acceptable levels of performance in order to pass validation tests as well as have an efficient and sustainable model. Such evidence depends on the methodology applied as well as the lack of eligible drivers to be used.
In this 60-minute live session, Prometeia will present a successful business case developed on the retail unsecured portfolio, which is deemed to be the most problematic segment to be modeled. The case shows an innovative approach for LGD development which is suitable, at different stages, both for regulatory as well as managerial purposes, leveraging both development sample construction and methodology through advanced Machine Learning techniques.
We will discuss the frontier of AI and Advanced Analytics in this field with experienced subject matter experts in Europe, both from the banking and consulting industry.
Introduction to LGD modelling via AI: the Prometeia approach
Marco Stella, Partner Prometeia
A retail unsecured portfolio business case
Ivan Prakhov, Manager, Prometeia
Fireside chat: AI and Advanced Analytics in Modelling
Dmitri Kraynov, Head of Risk Modelling and Research, Sberbank
David Eschwe, Head of Group Advanced Analytics, Raiffeisen Bank International