How to increase credit without increasing default
Current Risk models are based on traditional econometric techniques like the Logistic Regression, which is a functional analysis that does not take advantage of all the customers´ information available in companies.
Consequently, the evolution of these models into Machine Learning non-linear methods and techniques, enable companies to have more effectively customer´s risk forecasting models, easily updated and more accurate than the static ones. Having more information of each customer can result in transactions risks reduction and increase of lines of credit.
Grupo AIA ´s models enable to fix credit limits per customer, based on customer´s credit history, and leave a margin for a potential increase. The objective is to minimize regulations costs, without affecting the business.
The risks models are recommended to:
