Design Phase of PillarOne.RiskAnalytics
The PillarOne project group started the design phase for the risk analytics parts. From our point of view the risk measurement process consists of the following steps:
- Structuring and Modelling
- Reporting and Presentation
With PillarOne.Reserving, a calibration tool for reserve risk purposes was introduced. PillarOne.Reserving is now open for further content related to premium and reserve risk estimation. It provides a link from general accounting data to risk model segments. Thus, the application provides parameters which are often used within risk models.
PillarOne.RiskAnalytics will cover the following tasks
- Define the risk model
- Import the model parameters
- Evaluate the risk model and persist results
- Analyse and present results
PillarOne.RiskAnalytics will start with the model definition, describing all the segment models, dependency structures, observable variables, pseudo balance sheet items, and the risk mitigation instruments. The defined model is able to be run with different risk mitigation instruments for comparison purposes. Risk measures and further statistical result variables can be predefined.
The second step of PillarOne.RiskAnalytics is the calibration of the defined model, which means the link between model and business statistics (respectively data). If necessary, balance sheet items might be loaded, as well as parameters of segment models, dependency structures and risk mitigation instruments.
As a third step, the model evalutation is introduced. A model simulation process is followed by post simulation calculations of statistical result variables. For selected input and output variables, a persistency structure with common database accessibility is generated. And therefore the parameters and results can be futher processed by Business Intelligence suites and the planned PillarOne-Reporting part.
Now, the discussion within the project focuses on PillarOne.RiskAnalytics. Join in, and share your requirements.
-- Jörg Dittrich