SAS® Model Implementation Platform
Efficient, transparent, controlled risk assessment model implementation.
Significantly reduce the setup time for implementing extremely complex and computationally intensive bank stress test modeling systems.
Streamline model implementation.
A centralized platform and web-based interface simplify the development, deployment and maintenance of even the most complex bank stress test modeling systems (e.g., a Monte Carlo simulation, state transition framework). Once you estimate atomic models, you can quickly and easily group them into a system designed to produce desired results. You can reduce the total setup time for implementing such systems to less than one day – no specialized, non-SAS programming knowledge needed.
Centrally store models.
The SAS Model Implementation Platform stores models in a controlled, central environment, making them easy to call for execution. Model details are readily viewable, which greatly facilitates transparency. You can search atomic models based on key characteristics. Details on modeling systems are stored in an easily viewable format, making it easy to update and maintain them over time. Built-in unit testing and back-testing provide additional confidence in your results.
Run complex models fast, maintain them easily.
Quickly run very complex bank stress test modeling systems at the loan level on millions of loans – even when simultaneously running multiple economic scenarios. Distributed computing across a cluster – with our scalable, in‐memory risk engine – allows for massively parallel processing without the need to write distributed processing code. Fast execution means you can spend more time on analysis and exploration and less time waiting for runs to complete. High-level code means you can spend more time improving models and less time updating and maintaining them.
Visually explore results.
Once you run a portfolio, you can explore the results at an aggregate level, and then drill down to the loan level in seconds – even with millions of loan-level records. This in-memory drill-down capability enables very detailed, in-depth analysis of bank stress test results. You can explore any level of aggregation on the fly and then export results to an Excel file. Built-in, flexible, automatic attribution analysis makes it easy to compare with previous results.
Key Features
Quickly and efficiently execute a wide range of models used in bank stress tests and other enterprise-level risk assessments in a well-governed, scalable environment.
Fast, efficient model execution
Uses scalable, in-memory technology to quickly process complex loan-level modeling systems for all portfolios across the enterprise.
Simplified model & group setup
Provides prebuilt model templates and an easy-to-use graphical interface that simplifies the setup and maintenance of complex risk modeling systems, such as Monte Carlo state transition frameworks.
Centralized model execution library
Provides a sandbox environment for developing new data models and reports, and keeps all model information centrally stored for easy searching, versioning, tracking and unit testing.
Expected credit loss calculation environment
Supports calculation of loan loss models, such as those needed to support the IFRS 9 or CECL accounting standard. Makes all aspects of implementing and maintaining loan-level modeling easy.
Visual results exploration
Provides the ability to aggregate model results for millions of loans up to any desired level or drill down to the loan level in seconds. Lets you easily compare to prior results. Built-in attribution analysis enables you to understand the contributors to result changes.
Related Offerings
Check out these products and solutions related to SAS® Model Implementation Platform.
- SAS® Solution for CECLQuickly meet new US Financial Accounting Standards Board current expected credit loss (CECL) standards with best practices for modeling, workflow and reporting.
- SAS® Solution for Stress TestingMeet the challenges of enterprise stress testing and gain strategic advantage through advanced scenario-based planning.