Financial services companies face many challenges when scaling up analytics across multiple business areas, resulting in more than a year of new system deployment time. Increased cost due to inefficiencies, high loss rates, and more strategic projects abandoned in key use cases (such as Artificial Intelligence and Machine Learning initiatives, to name a few) are the next problem, and that only hurts the business.
In a time of rising interest rates and heightened economic uncertainty, financial institutions must consider operating costs and increased workloads, as analyst data show that model sizes of U.S. bank inventories have increased by about 31 percent since 2021.
A 2022 McKinsey survey of risk management model managers (MRMs) at 27 North American banks revealed reasons for this increase in model inventory size, including a surge in bank mergers and acquisitions that has led to the integration and expansion of model landscapes, expansion of MRMs, and major reconstruction efforts.
Model development and validation teams are facing peak workloads because of these challenges, resulting in stagnant business as usual. Studies show that the current modeling lifecycle can be 15 to 18 months, and pain points in the company can be detected throughout the process.
The modeling team's current lifecycle can be a pain point. Lack of formalized model acceptance processes, lack of early engagement by data engineers, poor clustering of use cases (especially with a bloated model inventory and early engagement) are all inherent problems for any company.
These are all inherent problems. To address these model lifecycle challenges, modeling and analytics leaders at financial institutions can accelerate the delivery of value for strategic technology use cases and free up resources throughout the model lifecycle by deploying four types of efficiency levers, potentially reducing time to market by 50 percent.
Model lifecycle transformation consists of four key phases, and each phase needs to be managed strategically from concept to deployment. This process begins with roadmap and communication, including understanding pain points and assessing baseline efforts such as metrics, cycle time, supporting infrastructure and workflow management tools.
Support tools are selected during the design phase to prioritize quick wins, and materials are designed to educate affected teams. A change story is developed based on the target state and current situation in collaboration with executive sponsors.