Model Risk and Basic Approaches to its Estimation on Example of Market Risk Models

Abstract

Model risk is currently a topic of great interest both to the academic community and to the financial industry; however, there is not yet any generally accepted approach to measuring it as of now. We give a review of basic model risk definitions and different indicators and approaches to model risk estimation and calculation within a common setting. The subject of this work is model risk itself that arises while using models to estimate financial risk and its quantitative indicators. The study is of particular relevance to financial institutions which for a long time have been actively applying different risk models. Since there is no generally accepted way to do so, choosing the exact approach becomes an important step in modelling. The aim of the paper is to demonstrate different approaches to model risk estimation on the same example of estimating the model risk of Value-at-Risk models and compare them within one setting. Methods used: analysis of time series, theoretical distribution parameters estimation. The result of the work is a list of methods and approaches for estimating and calculating model risk with the examples of applying these methods to a real risk-management task with appropriate interpretation. This work can be useful for researchers and risk managers of financial institutions.

Keywords

• model risk
• method of model risk estimation
• conservativeness
• accuracy and efficiency of estimations
• value-at-risk
• market risk