Under the latest risk-based reporting regulations, companies are required to perform multiple interacting projections to meet their regulatory requirements. These could be the various stresses under Solvency II, modelling of unexpected movements under IFRS 17, the stochastic and deterministic elements of US principles-based reporting, or hedging for models of guaranteed-benefit contracts.
Well-designed models will maximise the reuse of code and product definitions to minimise the operational risks both in reporting and in updating models. Nested Structures and Modules bring this modular development paradigm to allow users to extend and improve their complex models.
The Nested Structures feature in Prophet further allows the integration of your existing projection and regulatory models. This allows companies to project their balance sheets under those complex regulations, capturing the various stress scenarios resulting in the accurate modelling of the costs of capital, desirable for doing FLAOR/ORSA calculations. This includes modelling the evolving impact of Solvency II and IFRS 17 as well as all regional variations of those. This is particularly important for Interest Sensitive and Variable Annuity calculations where stochastic and decision-based calculations are frequently performed within other stochastic calculations.
The Modules feature in Prophet provides a framework whereby a set of calculations can be defined once and executed several times. This can result in a smaller and more manageable code base and provides greater control around the execution of calculations within Prophet models. Modules are particularly well suited to repeated calculations involving different bases, such as the shocks under Solvency II or similar.
Alternatively, companies can introduce simplified proxy models and aggregations into their business projections, giving faster access to a calibrated approximate view of their company’s future. This is particularly important for financial risks that can be compute intensive to simulate on a full projection model.