Improving PD models predictivity and reactivity with transactional information and ML

PD models suffer from a relatively low reactivity in detecting sudden structural breaks in the economic cycle, as traditional data sources do not represent an ideal solution to capture the clients’ creditworthiness evolution in a timely manner. In light of the unprecedented crisis generated by the Covid-19 pandemic, it is required to evolve the modeling framework in order to be able to cope with such a gap.

Within the AI4Risk Expert Circle project, in this 60-minute webinar Prometeia proposes a solution - already being implemented in one of the major European FIs - that represents a powerful tool to improve both predictive capacity and reactivity, leveraging transactional information and Machine Learning techniques, both in case of structural breaks as well as normal circumstances.

In the following fireside chat, eminent speakers from the global risk management community will discuss the hottest topics related to the adoption of these methodologies within the banking processes.


Agenda

11.30 - Use case presentation
Emanuele Giovannini, Head of Credit Rating Modeling, Unicredit Italy
Giangiacomo Sanna, Senior Specialist, Prometeia

11.50 - Fireside chat with:
Chiara Capelli, Head of Credit Risk Modeling, Unicredit Group 
Sid Dash, Research Director, Chartis Research
Dmitri Kraynov, Head of Risk Modeling, Sberbank
Marco Stella, Partner, Prometeia

12.20 - Q&A session


Register for the webinar: even if you won't make it to attend it live, you will receive the video-recording shortly after.