Model Identification Assessment
We establish what counts as a model across decision-making, management information and regulatory reporting, and build the process that maintains that judgement.
The framework around the whole model estate: what counts as a model, who owns it, how it is validated and what the board sees.
We establish what counts as a model across decision-making, management information and regulatory reporting, and build the process that maintains that judgement.
A single auditable record of every model you run, with role-specific workflow covering governance controls, documentation, dependencies, findings and monitoring results.
We independently validate any model against its performance, its intended use and supervisory expectations.
We write the standards for how models are built, validated, approved and owned across the firm.
We design and build the reporting that shows model risk across the lifecycle: development, the live estate, monitoring breaches and validations overdue.
We extend your model risk framework to cover generative AI, setting out how these systems are identified, tiered, tested, approved and monitored.
A European bank needed to know that the capital it holds against concentrated lending would stand up to scrutiny, so it asked us for an independent view of the model behind the number.
A UK lender needed every model in its IFRS 9 expected credit loss suite rebuilt, at the point when the team that had built them was no longer there.
Model risk frameworks are easy to write and hard to operate, so ask whether a bidder leaves you with a document or a working inventory: who owns each model, which validations are overdue, what broke in monitoring last month. At Gini we define what counts as a model across decisions, management information and regulatory reporting, write the standards for how models are built, validated, approved and owned, and build the inventory and reporting that keep the picture current.
Yes. Independent validation means the validator holds no stake in the model passing, so where Gini validates, it does not develop, and where it develops, somebody else validates. We test any model against its performance, its intended use and supervisory expectations, and write findings your model committee can act on, each with its severity and the work it needs named.
If those systems inform a decision, a management information figure or a regulatory report, then yes. The identification principle under SS1/23 reaches them whether or not your inventory does. At Gini we extend the framework to set out how such systems are identified, tiered, tested, approved and monitored, and we test live systems adversarially so the tier reflects what the system can be induced to do.
By materiality and use rather than by method, which usually shrinks the tier one population and reveals spreadsheets that belong in it. Gini runs the identification exercise across the business, assigns each model an owner and a tier against stated criteria, and builds the reporting that shows validation status and overdue work without anybody assembling it by hand.