Regulatory financial model development for major UK bank

Close up of safe


Justin Cooper was engaged by a major UK bank to assist in the development and automation of an approach for the calculation of PPI redress for mis-sold policies.  The services provided included an end-to-end process comprising of:

  • Developed, tested, and released financial models to calculate PPI redress.
  • Audited and reviewed existing PPI financial models.
  • Re-designed and developed financial models in line with FOS guidelines.
  • Developed highly automated tool and process to handle 80% of PPI claims with minimal manual intervention.
  • Managed PPI-redress calculation process.


  • Took ownership of existing financial models that had been developed in an ad-hoc manner, which were known to contain errors and had no documentation or version control.  These models were audited, refined and developed into fully FOS-compliant models for the purpose of PPI redress.
  • Developed several models from scratch for other financial products which required PPI redress, and integrated these into the automated process.
    put in place a software development lifecycle to manage the frequent necessary changes to the models, in order to remain complaint with the latest FOS guidelines. This included design, development, testing and release phases to a planned schedule.
  • Developed an automated system to populate the models and calculate redress for 80% of customer claims. This involved the bulk upload and processing of customer claims using transactional data, eliminating the need for human review and vastly speeding up a previously manual process.


  • The client was able to reduce the backlog of PPI claims and make significant progress in the speed of response to new claims. The client had suffered from a variety of past mistakes in their PPI redress process and calculations, which had led to significant fines.
  • The developed processes ensured that further mistakes in the calculation of redress were minimized and the process was fully auditable.