Automated Testing and Documentation at G-SIB
Regulatory expectations and requirements have intensified over recent years and model risk has taken centre stage on the regulatory agenda. As a result, banks across the industry are reconsidering their operating model and looking for ways to increase efficiency and effectiveness across the end-to-end model lifecycle.
In this case study, we look at how a Global Systemically Important Bank (G-SIB) has implemented Yields’ Model Risk Management platform to support its modelling lifecycle and model maintenance under evolving regulatory standards.
The bank is committed to practising effective MRM by enforcing model risk policies that ensure full compliance with the SR11-7 guidelines for MRM.
Model documentation provided by model developers is often poorly written and inconsistent, meaning it is inadequate to support a rigorous validation effort around the standards set by SR11-7.
Standardizing the model lifecycle is challenging, given the multiple data sources and analytics libraries, the lack of standardized testing templates, and the strict documentation standards.
The bank was looking into enhancing the efficiency and effectiveness of its model development process. Furthermore, the organization was performing a model uplift to align its model documentation with internal/external minimum documentation standards, and wanted a strategic solution in place to streamline future uplifts through automated testing and documentation generation.
Our MRM platform was deployed on-premise and integrated with the client’s enterprise risk systems and data stores to leverage the golden sources within the organisation. By enhancing the model lifecycle and introducing automated testing and documentation, the organization achieved the efficiency gains needed for model developers to re-focus on value-add activities.
REPRODUCIBILITY & CONSISTENCY
All tests run on the platform can be replayed in the future with the data and model as they were originally run.
AUTOMATED TESTING & MONITORING
Models are tested at scale through the execution of standardized reusable routines, which improves the efficiency, reliability and speed of repeatable tasks and tests.
AUTOMATED MODEL DOCUMENTATION
Automated generation of model documentation can be supported via predefined text blocks in 20% of the original time.
Integration allows information like model name, model owner, legal entity, etc. to be consistently consumed within the client’s infrastructure.
The Yields Advantage
Chiron Enterprise is a customizable model inventory and workflow management tool that streamlines the execution of end-to-end Model Risk Management processes.
Chiron App is a modular data science platform that automates all quantitative model testing and documentation.
Learn more about the Yields MRM Platform at www.yields.io
In 2022, Yields.io has won the ‘Model Validation Service of the Year’ award at the Risk.net technology awards for the fourth year in a row. Yields.io is widely recognised as a technology leader in MRM, and is supported by global partners such as KPMG and IBM.
Director of Sales