An internship project on how to prevent and mitigate the risk of credit default on a retail bank

My name is Nuno Aires and I completed my master degree in Applied Mathematics to Engineering and Finance at the Engineering School, Polytechnic of Porto (Instituto Superior de Engenharia do Porto), Portugal.

As a student of this master degree I had the opportunity to make, as my final project, an internship at a Bank. This internship was one of several partnerships between ISEP and companies from different sectors (Industry, Finance, Insurance, Services). The guidance of the work was shared between two supervisors, one at the Bank and other at ISEP.

The internship took place in the models area and automated decision processes of the Bank and its scope was the prevention of credit risk: foreseeing the potential credit default is critical in the way that it allows for the Bank to enable preventive actions near their customers.

The internship had three main stages:

• The first one was to understand the business, the Bank´s informational system and the existing processes to prevent credit default.
• The second one was to make the analysis of a new database and to structure the information in order to integrate it in Bank’s informational system and to understand the information constraints.

• The final stage was to develop a new model from the combination of Bank’s information and the new database, with the aim of turning the prevention of credit default even more effective.

To produce the new model, we selected and used a scoring model, because it is a predictive and understandable modeling technique widely accepted in banking. It also allows the introduction of expert rules, to ensure business conditionings (legal or operative).

We have tested this new model and it proved to be predictive, statistically stable and responding to the proposed objective.

Nuno Aires (Ocidental Grupo)