Basel Model Implementation

The IQ Business Group assisted a large financial institution in the end-to-end retail Basel Model implementation.

The Business Challenge

A division within a large Financial organisation develops and is responsible for all the retail Basel capital models. When IQ was approached to assist, the department was manually running the capital models on a monthly basis for reporting purposes. The models required manual human intervention and modelers had access to change code as when required without following any formal procedures. The model calibration and application was combined and the models were very time consuming to run (between 2- 5 days). Not only were they time intensive, but they also often produced inconsistent results and were poorly documented.

The primary driver and need for our solution was the SARB and Basel requirements for models to be automated and to run in a controlled environment where there are change control processes and procedures in place to ensure proper governance techniques are followed.

The Solution

After analyzing the environment and engaging with the stakeholders (both directly and indirectly involved in the project), the IQ team compiled a fully comprehensive project plan to fully address all the issue.

The project was divided into 10 key workstreams with focus on the following:

  • The Integrated Model Framework– The integrated model framework is a framework that we developed within which all models are run. This piece of code reduces the need for human intervention and allows the models to run more efficiently and effectively.  IQ designed the IMF in such a way that it is fully automated with optimized code that takes a fraction of the time (3 hours) to run. If errors are encountered during the run, an email notification is sent to the relevant parties notifying them of the issues encountered and allowing them the opportunity to halt/continue with the process; thereby eliminating the need for a full run when issues are identified.
  • Technology requirements and platform–A stable and reliable technology environment was vital to the success of our project. Issues such as access, capacity, licensing and storage were encountered. IQ carefully crafted a technology BRD that addressed the challenges. A solution was identified to resolve the issues and ensure that the adequate technology was in place to enable the required modeling, testing and implementation to occur.
  • Model development and implementation–This workstream formed the largest part of our project and was seen by most to be the most critical portion. The process involved in implementing models was a lengthy and time consuming process. Once model methodology was approved by the relevant committees, IQ would develop the application code for each of the models and integrate the piece of code into the Integrated Model Framework. End to end testing was performed before implementation to ensure the overall capital impact could be quantified. This would allow our client to communicate the expected impacts to the SARB ahead of time. IQ prepared and obtained sign off on all the relevant BRDs, testing documents and governance measures that were required before model implementation.
  • RWAs and regulatory reporting– The purpose of this workstream was to provide regular communication and updates in terms of capital movements to all authoritative bodies including SARB, Barclays and Group Regulatory Finance. We identified 3 main intervention points where we met with the SARB to discuss model methodologies, project progress and capital movements.
  • Management understanding and training– The purpose of this workstream was to ensure that adequate communication, training and stakeholder management was evident throughout the project. Roles and responsibilities were defined, and workshops were held to ensure efficient transfer of knowledge to all the relevant areas
  • Policy review, alignment and embedding –The purpose of this workstream was to review the current policies and ensure alignment with the regulatory requirements. This was used to identify operational and methodological challenges to compliance, to adjust the policies, to develop new policies and to implement the policies.
  • Governance–The purpose of this workstream was to ensure governance processes and policies were adhered to. IQ's role was to ensure that models followed the correct internal governance procedures for approval and implementation, as well as ensuring external procedures were followed. IQ also ensured that the correct SOX controls were in place for model automation.
  • IT engagement–The purpose of this workstream was to define and agree the procedure to be followed for the production, testing and implementation of the models. Testing schedules were compiled monthly and all stakeholders were notified of the changes to ensure the process was managed carefully and did not allow for time slippage.
  • Data integrity and quality– The data integrity and quality workstream of the project required IQ to develop validation metrics for the models. This entailed creating data quality rules for each model that was approved by developing accuracy, consistency and completeness rules for each of the fields that were required in the models. These rules were then uploaded onto a central system and used for monthly monitoring by the MDA (Model governance and Data Assurance) team.
  • Internal media communication and project update– The purpose of this workstream was to ensure that all stakeholders were kept up to date with the progress of the project. A weekly status meeting was held with the CVM team where any issues/concerns were raised and where progress was discussed. Minutes of the meeting and a status update was sent to all stakeholders weekly. In addition to this, IQ hosted a bi-weekly steering committee to address the more pertinent issues with the project executive and sponsor.

The benefits our client received

The models no longer required manual human intervention and modelers have formal procedures which have been put in place. Benefits include:

  • Regulatory compliance
  • Model automation and optimisation
  • Stable capital numbers and reporting figures
  • Change control processes and procedures