Introduction

The use of models within the financial industry has become prolific to the degree that their deployment has become responsible for a risk factor in and of itself: that of model risk. Although the benefits financial models can offer are considerable, a model that is ill-fitted to its task, applied improperly, poorly understood by its users or not subject to rigorous checks can potentially cause massive damage to a business.

Key findings

• Why validate ALM models? – Validation of ALM models is vital for a number of reasons. The most obvious of these is regulatory requirement. In the US, the law is laid down in the OCC Bulletin 2000-16: Risk Modelling, Description: Model Validation document, while European countries are subject to the regulations of Basel II and Solvency II.
• Current industry practice – In addition to the guidance provided by the regulators, ALM departments must decide what model validation procedures to follow, both during and after the implementation of an ALM model, and how often these procedures should be followed in the future. In addition, they must decide whether internal or external validation is more appropriate at any given time.
• What should be included in the validation process? – The elements included in a typical validation vary between institutions depending on their size, business lines and other factors: there is no ‘one size fits all’ process that can be used across the board. However, common processes include looking at data integrity, cash flow definitions and back-testing of model predictions versus actual results.
• Quantitative versus qualitative approach – It is important that validation should not be viewed as a purely quantitative exercise, although this approach by necessity provides much of the data necessary for a successful model validation. The Basel II validation subgroup has been at pains to stress that qualitative elements must stand beside the necessary mathematical exercises.

Conclusion

Most participants in this survey feel – and this is backed up by the regulators themselves – that the onus for a thorough and successful validation approach falls directly on the banks and insurance firms, which results in the need for a clear set of rules for best practice approach. While this varies between firms, often relating directly to their size, the need for both internal and external validation seems to be well-recognised.

To obtain a free copy of this section in full, please contact us at marketing@lepus.com with your name, job title, firm, phone number and email.

Tagged with: 2009, model risk, risk, Risk Research Report