The discipline of identifying, measuring, monitoring, and controlling the risk that a model produces adverse outcomes through incorrect or misused output, set out for US banks in supervisory guidance SR 11-7.
In practiceMRM treats every model as a potential source of loss: financial loss, reputational damage, or supervisory action. The framework requires a firm-wide model inventory, written policies, defined roles for owners, developers, and validators, materiality tiering, independent validation, ongoing performance monitoring, and board-level reporting. It is increasingly extended to AI and machine learning systems, with regulators expecting the same discipline applied to LLM-based applications where the firm relies on the output to make or support decisions.
A bank stands up an MRM function that holds the model inventory, mandates a standard model documentation template, and runs an annual validation plan in which each tier-1 model receives a full re-performance and each tier-2 model receives a targeted review.
This definition is maintained by Moweb partners and used in live client engagements. For how Model risk management applies to your estate, or to challenge a working definition, speak to a partner.