Master Thesis: Fraud Modeling

Background:

We are a fintech bank specializing in SME business loans, with a strong focus on developing highly accurate credit risk assessment models. First-party fraud is a growing challenge in our sector. Unlike identity theft, this type of fraud occurs when a legitimate customer applies for credit with the deliberate intention of never paying it back.

The most blatant indicator of this behavior is a customer who defaults on their very first payment and rolls directly into delinquency. Distinguishing these intentional fraudsters from honest customers experiencing sudden financial distress is incredibly difficult, yet crucial for robust risk management.

Project Objective:

Leverage predictive models to identify first-party fraud in SME business loans through in-depth historical data analysis and feature engineering.

Responsibilities:

• Data Analysis: Analyze historical loan and repayment data to uncover behavioral patterns of intentional first-payment defaults.

• Feature Engineering: Identify and create variables that distinguish intentional fraudsters from honest customers in financial distress.

• Model Development: Develop a predictive model that integrates the newly engineered features.

• Validation and Testing: Validate the model to ensure its accuracy and robustness.

• Documentation and Reporting: Document the analysis, findings, and model results.

How To Apply:

This is what we need from you:

• A brief description of yourself.

• Attach a CV, make sure it contains relevant courses and projects that you have completed.

• Note: This Master Thesis Project is based on-site in Stockholm.

This master thesis student will join our team from January 2027 - June 2027. The student gets a computer, access to the office and a 20 000 SEK compensation for a successful master thesis project. If you have any questions about your application, please reach out to alexander.eriksson@qred.com.