SAS Products

SAS High-Performance Risk 3.8: Release Notes

June 2017

New features:
  • Risk cube descriptor folder structure. You can now organize risk cubes into folders and navigate through the folders in the user interface.
  • Risk work groups. You can use risk work groups to set permission on risk cubes and scenarios. Assignment to one or more risk work groups restricts access to cubes, which are stored in directories corresponding to those groups.
  • Historical simulation. Historical returns computed from time-series are now available.
  • Computed methods that roll up. In a stress cube, a computed method can be applied at the instrument level and summed to any aggregate level.
  • Cross classification filters can be stored with a risk cube and applied on every query. The filters are displayed in the user interface.
  • Risk factor sensitivity analysis in the UI. In a stress cube, select risk factors to view first and second order sensitivities.
  • Live parameter matrix data sets to provide and update parameters for stress testing, user-defined statistics, and computed methods. Previously, parameter matrices were stored in the cube and could not be updated.
  • Pre-query output variables. At query time, before aggregation, the PreQuery method runs on each instrument or base cell at each state to create new output variables. These variables are treated the same as regular output variables in queries, but are not stored in the cube. Pre-query methods can reference parameter matrices and have access to all simulated output variables across all horizons.
  • SAS Risk Dimensions option to simulate all risk factors. SAS Risk Dimensions optimizes to simulate only risk factors used by the current portfolio. This enables you to simulate all risk factors in SAS Risk Dimensions even when the current portfolio or portfolio template does not use them. This can be useful when simulating risk factors in SAS Risk Dimensions for input into SAS High-Performance Risk.
  • SAS Risk Dimensions option to simulate all time points in covariance simulation. SAS Risk Dimensions defaults to scale risk factor values with the square root of time between simulated horizons. This argument forces risk factor simulation along all of the time steps between simulated horizons.

Performance:

  • Performance for PostVar calculations has been improved.

Software Fixes:

  • Sub-cubes using a schema data set now retain their schema values when joined side-by-side (for comparison). Previously, the schema of the last risk sub-cube joined would override the schemas of the other sub-cubes when each sub-cube had a schema data set that had the cross classification variable as the primary key
  • Bug fixes for Loaddata for ScenarioCF
  • Bug fixes for Loaddata for scoring
  • Portfolio details in the user interface now shows quadratic instrument values
  • Computed variables are now present in the ScenTrans data set
  • Support for cash flows in multiple horizon scenarios is now available