Argenta Bank has a comparatively small model validation team tasked with managing model risk and independently reviewing all models classified within the highest or second-highest tier categories. As validation backlogs accumulated, the team faced increasing challenges in meeting the growing demand for validation requests and ensuring compliance with regulatory requirements, particularly in monitoring its Internal Ratings-Based (IRB) models. 

Model validation activities were conducted manually, making the process lengthy and time-consuming. As models were required to undergo yearly internal validation, in compliance with ECB regulations, the team had to dedicate three months to repetitive tasks for each validation annually. This limitation prevented Argenta from conducting validations more frequently, expanding validation activities to other model types, and focusing on high-value and qualitative tasks.

To enhance their capacity and address the rising demand for model validation, as well as to accommodate new model types such as ALM and IFRS 9, the model risk management team at Argenta initiated a search for a technology solution. This solution needed to automate repetitive tasks, maintain version control to tackle reproducibility issues, and assist in monitoring efforts to increase the frequency and comprehensiveness of model testing.

Before using Yields’ solution, we were limited to conducting validations on a yearly basis. Now, we can perform validations monthly with just a few clicks, and the entire process is automated. Additionally, we have established a robust audit trail, and the platform enables us to effortlessly reproduce analyses from the previous year or even further back, providing a significant advantage when addressing regulatory requests.

Quinten Collier
Manager Validation & Model Risk Management
at Argenta

After testing and analyzing the added value of Chiron App, Argenta opted for a cloud-based implementation of the solution. The implementation enabled them to automate their model monitoring and validation tasks, progressively increase their testing frequency, and expand the number of models (traditional and AI/ML models) they can validate within a given timeframe.

The solution’s advanced automation capabilities have reduced the annual model validation time from 3-4 months to just 2 weeks, and it now conducts model monitoring automatically each month. As new model data becomes available monthly, it is promptly uploaded to the system in a structured format and standardized test scripts are run. Consequently, Argenta has gained the agility to detect potential model issues earlier and implement necessary changes in a timely manner. Throughout this process, the model validation team receives notifications, and an audit trail is maintained.

Furthermore, the solution reshaped the role of the model validation team, enabling members to concentrate on qualitative aspects, invest time in comprehensive analyses, and generate detailed reports with unprecedented speed and ease. Notably, the enhanced standardization and automation introduced by the solution have allowed Argenta to assign more model validation tasks to junior professionals and reduce dependency on senior personnel. 

Argenta has been a satisfied customer since 2019, and in that time, Yields’ customer success team has cultivated an intense and fruitful relationship with them. This relationship is simultaneously one of the main reasons why Argenta continues to use the solution. Yields’ customer success team understands the challenges, speaks the client’s language, and the support has been proven to be ongoing, detailed, and fast. Furthermore, Argenta’s custom feature requests are welcomed and incorporated into the product roadmap.