News 01/12/25

Controlling model risk is crucial for asset managers

Complex models have become an integral part of asset management. These models are used to assist in portfolio choices and risk management, but they may also introduce new vulnerabilities. Errors in development or use can lead to incorrect decisions and damage. The report ‘Model risk management by asset managers’ shows how the sector is currently dealing with these challenges and what steps asset managers can take to better manage model risk.

In short

More models, more risk – Complex models and AI increase the likelihood of errors and financial damage.
Practical tools – Eight good practices help asset managers get a grip on model risk.
Clear agreements and dialogue are essential – There is no uniform approach; knowledge sharing and dialogue are necessary.

The study is based on a broad survey of 250 market participants and an in-depth questionnaire completed by 14 asset managers. The aim is to provide insight into current practice and encourage efforts to strengthen model risk management.

More models, more risk

Due to AI and smart data analysis, asset managers are using increasingly complex models. This helps them make decisions, but it also entails risks. If there are errors in the development or use of a model, this can lead to incorrect choices and damage. That is why the Financial Supervision Act (Wft) stipulates that asset managers must maintain control and integrity in their business operations – including the proper management of model risk.

Practical tools

How do you tackle model risk? This report contains eight concrete tips. Start with a clear structure for who is responsible for models. Ensure that everyone shares a common understanding of the term ‘model’ and determine how much risk you are willing to take. Maintain an overview with a central list of all models and always validate them before use and regularly afterwards. Work with a life cycle: from development to termination. Also draw up rules for models from external parties and ensure that you have sufficient internal knowledge to recognise and manage risks. These guidelines are not mandatory, but they help to keep control of a complex and rapidly changing area.

Clear agreements and dialogue are essential

Our research shows that asset managers recognise the importance of model risk management, but that definitions and approaches vary greatly. Governance is often not explicitly defined and risk appetite is rarely formally stated. Yet this would strengthen controlled and ethical business operations. That is precisely why it is important to continue the dialogue on this subject – with regulators and with parties that supply models, such as external models and AI applications. By sharing knowledge and experiences, the sector can work towards a common understanding and a more consistent approach. The AFM will continue to monitor this topic and invites the sector to work together to strengthen model risk management.