Model Management That Fits Diverse Needs

As those of you who follow this blog know, I've been discussing the importance of centralized model management. But centralization should not translate into one-size-fits-all.

Indeed, banks vary greatly in how they use analytics across their enterprise and the scope of their efforts to standardize lifecycle model management processes. Among our clients, we see such diversity:

• A top-five US bank is centralizing management of every model across its vast enterprise.
• A leading Asian bank is initially focusing on ensuring their models in development achieve Advanced Internal Rating Based status under the Basel III global standard.
• A top-five Australian bank seeks to bridge current process inconsistencies around model tracking and validation of Basel rating models and decision models across different countries.

Because of these varied needs and goals, configurable workflow tools are a core component of any model management infrastructure. It’s essential to have the flexibility to align workflows with the needs of analytic teams. Modeling methods, documentation and formal approval requirements vary across geographies and markets.In addition, requirements—as well as methodological pitfalls—depend partly on the types of analytics being built. For instance, expert (judgment-based) models are developed with a very different process and have different documentation needs than empirically derived models.

To enable this vital flexibility, your model management solution must allow for different workflow processes, each tuned to the needs of their users while allowing reuse of common processes.

Also essential are centralized repositories, which provide executives, managers and compliance officers with a centralized view of the status of models across the organization, as well as a single location to investigate and address questions raised by regulators. These repositories also support collaboration, where appropriate, across teams, enabling cross-training and dissemination of best practices throughout the organization.

For more information on this approach, I invite you to read our latest Insights white paper, “Reducing Regulatory Drag on Analytics Teams.”