Ensuring completeness and accuracy of model inventory to monitor uses across all areas
The views and opinions expressed in this article are those of the thought leader and not those of CeFPro.
By Nikolai Kukharkin, MD, Head of Model Risk Management, TIAA
What are the strengths of using different models to monitor model accuracy across different areas?
Models are always approximations of reality, so they can’t be precise. Model Risk arises from incorrect or inappropriate application of models, including the reliance on a set of assumptions that may not be in line with the model’s objectives or may only be valid for a limited range of input data or market and economic conditions. If one uses different models, it helps to see a bigger picture, thus having a better chance to evaluate the accuracy of approximations that any model produces. In addition, different models can help to understand different types of model inconsistencies. For example, two different models can behave similarly when considered in a boundary case, but may reveal quite different dynamics in broader cases.