Classifying a model vs tool for an effective model risk framework

Classifying a model vs tool for an effective model risk framework

By Elizae Dalvi, VP of Model Risk Management at BankUnited. 

Ahead of the Stress Testing USA: CCAR & DFAST Congress 2017, Elizae Dalvi, VP, Model Risk Management at BankUnited shared with us her insight. 

Elizae, can you please tell the Risk Insights readers a little bit about yourself, your experiences and what your current professional focus is?

I am VP Model Risk Management at BankUnited. I have been primarily focused on enhancing the bank’s Model Risk Management framework and managing its key aspects including model validation for stress testing, ongoing model risk monitoring, policies & procedures, and governance.

At the Stress Testing USA: CCAR & DFAST Congress last year, you presented on your insight regarding classifying a model vs tool for an effective model risk framework – Why is this a key talking point right now?

The question of what is a “model” is often debated and is of critical importance to application owners, model risk managers, and regulators. Getting the right answer to this fundamental question is important because it drives the formality, rigor, and frequency of model risk management activities.
The concept of a “model” as defined in the Supervisory Guidance OCC 2011-12/SR 11-7 leaves some latitude for interpretation:
“A quantitative method, system, or approach that applies statistical, economic, financial, or mathematical theories, techniques, and assumptions to process input data into quantitative estimates.”
Although every application should be tested to ensure it is producing accurate results and is appropriate for its intended use, it is not prudent that every application undergo the added costs of formal model risk management activities and the higher standards of regulatory expectations this entails. This is where it is valuable to have a methodology to classify a model vs tool across the enterprise.

How can financial institutions best manage tracking models across enterprise?

In my view tracking models across the enterprise should be addressed from multiple directions. First, there needs to be commitment and support for this from higher level management. Second, the 1st line of defense (model owner/business line) should understand their roles and responsibilities to ensure that they communicate to model risk managers about existing and new models. Third, there should be a partnership with the 3rd line of defense (Internal Audit) so they are also the eyes and ears of Model Risk Management.

Without giving too much away, can you outline the key challenges that arise when developing a concrete definition of a model based on SR 11-7?

The key challenge I see is how to apply a consistent, defensible, and prudent methodology to classify a model vs tool across the enterprise. This is critical so that applications that rise up the level of a model undergo the appropriate model risk management activities including a formal validation, ongoing monitoring, and documentation based on the standards of regulatory expectations. Consequently, tools can undergo a less formal validation process which results in lower costs and overall increased productivity for the institution.

Can you give a brief overview of the best practices in stress test model development?

Model development is an important stage in the model lifecycle in which to involve model risk managers because it provides the model owner/business line an opportunity to catch problems earlier in the process. Model owners can also solicit constructive suggestions from Model Risk Management to seek challenge for the model choices being made and evaluate its design, theory, and logic earlier in an effort to improve overall model quality.

What, in your opinion does the future hold for stress testing professionals, and how can they keep up with the increasing changes in the industry?

The role of a stress testing professional is always evolving because it requires working in multiple disciplines such as finance, statistics, risk management, and IT and keeping abreast of changing regulatory expectations. This offers a unique opportunity to learn and build a broad-based skill set in diverse areas of an institution’s operations.
In my view, to thrive in the stress testing profession it is important to provide meaningful insights into the portfolio’s key risk characteristics and the model results under stress to facilitate a practical and useful understanding to the model owner/business line of the risks involved and how to mitigate them.

This article represents the views of the author(s) only, and does not necessarily represent the views or professional advice of BankUnited.