SAS® Solution for CECL Features

Modeling & analytics 

  • Prepackaged templates for commonly used model types speed model development and make it easy to compute lifetime expected credit loss (ECL):
    • Hazard models.
    • Markov chains.
    • PD curves.
    • Loss given default (LGD).
    • Monte Carlo simulations.
    • State transition models.
  • Input data model supports ECL calculations and financial disclosure reporting.
  • Expected credit loss scenario attribution model computes the ECL impact of different scenarios.

Aggregation & reporting

  • Easy-to-use, self-service reporting templates and out-of-the-box visualizations for quick design and deployment of CECL reports.
  • Ability to aggregate or drill into results on the fly. 
  • Automated disclosures and posting into ledgers for enhanced accountability. 
  • Worksheets aggregate and merge ECL calculations with other financial data across business entities.
  • CECL report templates, which can be modified as needed, include:
    • Lifetime ECL.
    • Credit Quality.
    • Vintage Analysis.
    • Reconciliation to FAS 5 and FAS 14. 
  • Provides CECL financial disclosure reports (e.g., loss allowance, provisions and roll forward allowance).  
  • Enables financial aggregation according to FASB guidelines.
  • CECL project management dashboard monitors the status of monthly CECL reporting and related workflows. 
  • Data quality and reconciliation rules are provided in accordance with the principles of BCBS 239. Rules may be extended by business users as needed.


  • Unified workflow for greater control and improved collaboration and transparency.
  • Easily customizable accounting rules and data flows.
  • Ability to synchronize all risk and finance processes for improved auditability, transparency and repeatability.
  • Provides sample workflow templates for managing ECL computing process.
  • Additional SAS scripts for added flexibility and seamless integration of the disparate parts of the CECL process.
  • End-to-end data processing support, from inputting data models to reporting and generation of posting data.