What is your strategy for staying current with stress test validation and verification?

1 Validation vs verification

Validation and verification are often used interchangeably, but they have distinct meanings and objectives in stress testing. Verification is the process of checking that the stress test models are implemented correctly, according to the specifications and design. It involves testing the model code, logic, functionality, and performance, as well as the quality and integrity of the input data. Verification aims to identify and correct any errors, bugs, or inconsistencies in the model implementation. Validation, on the other hand, is the process of evaluating whether the stress test models are fit for purpose, meaning that they capture the relevant risk factors, reflect the economic and financial conditions, and produce reasonable and plausible results. It involves testing the model assumptions, parameters, calibration, sensitivity, stability, and robustness, as well as the appropriateness and relevance of the scenarios and the output data. Validation aims to assess and improve the model performance, accuracy, and reliability.

2 Why validation and verification matter

Validation and verification are essential for ensuring the quality, credibility, and usefulness of stress testing. Without proper validation and verification, stress test models may produce misleading, inaccurate, or inconsistent results that could compromise the risk management, capital planning, and decision making of financial institutions. Moreover, validation and verification are required by regulators, such as the Federal Reserve, the European Central Bank, and the Bank of England, who conduct regular supervisory stress tests and review the internal stress test frameworks of financial institutions. Failing to comply with the regulatory standards and expectations for validation and verification could expose financial institutions to reputational, operational, and legal risks.

3 How to stay current with validation and verification

Staying current with validation and verification means keeping up with the changes and developments in the stress testing environment, such as new regulations, methodologies, models, data sources, scenarios, and best practices. It also requires conducting periodic and comprehensive exercises that cover all aspects of the stress test models, from design to results. To do this, you should follow regulatory guidance and requirements for validation and verification, such as SR 15-18 and SR 15-19 letters from the Federal Reserve. Additionally, monitor industry trends and best practices for validation and verification provided by professional associations like GARP or PRMIA. Furthermore, you should implement a robust and independent validation and verification function within your organization with clear roles and responsibilities. Lastly, leverage available tools and technologies for validation and verification like software applications or frameworks that facilitate automation of tasks. For example, Matlab, R, or Python can be used to perform tests on models or data while Excel, Power BI, or Tableau can create visual reports.