ECL
Expected Credit Loss (ECL) Computation - CECL/IFRS 9
With high performance in-memory parallel ECL computations across portfolios, with a comprehensive suite of methodologies, you can perform loss computation at the most granular level and analyze change in each step by performing & comparing multi-period, multi-methodology runs.
Comprehensive Coverage
Functionalities to support all aspects of IFRS9/CECL including classification, EIR calculations, ECL computation and G/L reconciliation
Rapid Model configuration
Pre built templates for all modelling approaches namely PD, LGD, EAD, TTC to PIT Calibration, 12 month and lifetime PD and LGD supported. Transition matrix, roll rate and vintage loss methods. Supports champion challenger model compasirions.
Attribution
Explains the differences in the ECL numbers using multi dimensional attribution framework by quantifying the impact due to changes in models, scenario methodology, portfolio
Model/Scenario What If
Unique capability allowing business users to modify any model or scenario from front end to conduct what If analysis dynamically. Speed of execution allows rapid exploration of alternatives
Smart Narrative & Reporting
Automation of regulatory reporting; Management reporting and dashboards with charts and smart commentary
Transparency & Auditability
Establish data and result integrity with historical repeatability and version management. Describes the data used, methodology followed, computation sequence holistically. BCBS 239 compliant architecture.
Capital
Basel III/IV – Capital Computation
Compute required regulatory capital on demand or using predefined batch process. Comprehensively covers Credit Risk Standardized and Internal Rating Based approach, Market Risk - Standardized Method and Operational Risk – The Standardized Approach.
Multi Jurisdiction
Ability to configure and compute capital numbers using different regulatory treatment for home country as well as host country
Attribution
Explains the changes in RWA between any two given computations at atomic level across various dimensions such as – asset class, products, ratings, risk components.
Smart Narrative
Automation of regulatory reporting; Management reporting and dashboards with charts and smart commentary
Self Service
Allows configuration of every single formula and business rule from the front end. Any changes in the regulatory requirements can be configured by business user
What If
Enables banks to perform various levels of What-if Analysis to compute the impact on required capital and available capital due to change is ratings , model, scenario, collateral and default
Transparency & Auditability
Maintains complete audit trail of input data, computations performed, data extraction; thus allowing complete visibility for audit purposes
B/S Analytics
Advance B/S Analytics
Regulatory changes, profitability pressures and volatile business environments demand that Finance & Risk leadership look far beyond what existing ALM products deliver and leverage advance analytics (AI/ML) to dynamically model balance sheet impact for each scenario
Attribution
Understand movements in various risk and return metrics in terms of underlying products and individual transactions
Dynamic B/S Modelling
Dynamic and forward looking approach to forecast risk and return metrics incorporating business plans
Model Automation
Auto-generate model configuration parameters on a regular basis behavioural and new business profiles
What if Integrated Stress Testing
Bring all risk and return metrics of the balance sheet under a single analytical framework to model response to multiple strategies and scenarios
Smart Narrative & Reporting
Automation of regulatory reporting, Management reporting and dashboards with charts and smart commentary
Comprehensive Risk Monitoring
Automated daily tracking of various metrics against risk appetite, early warning indicators and set-up triggers with notifications
MRM
MONITRO – Automated Model Risk Management
Enables superior model risk management through intelligent automation of model monitoring, validation and reporting life cycle to improve model governance while achieving significant cost savings
Model Inventory
Comprehensive captures all critical information such as model type, tier, usage, versions over time, owner, validation and performance results over time with an ability to define schedule
EDA
Automatic missing data imputation using AL algorithms and data anomaly detection
Comprehensive Model Monitoring
Automated tracking of various metrics against pre defined performance thresholds, warning indicators and set-up triggers with notifications
Validation & Performance Analysis
Comprehensive coverage of model performance evaluation tests for rank-ordering, calibration, benchmarking and stability of classification and regression models
Regulatory Compliance & Reporting
Creates model validation documentation using custom formats for regulatory submissions and independent management reviews
AI/ML Challenger Benchmarks
Ease of creation of challenger model using AutoML feature that automatically builds challenger models to accelerate model recalibration to improve the model performance
CreditNext
Enhance, augment credit insights by leveraging an AutoML framework. Rapid build, rebuild, calibrate credit models.
Analyst Driven
Traditional Scorecard development in automated manner using Logistic Regression with comprehensive model performance analysis and model selection using overrides
Driverless
Auto ML model building using various algorithms on the sample data with automated HP tuning, covering GLM, RandomForest, Gradient Boosting and Neural Nets
Model Explainability
Explaining model output for specific observations and identifying important variables at model-level
Model Performance
Comprehensive coverage of model performance evaluation tests for rank-ordering, calibration, benchmarking and stability of classification and regression models
Model Documentation
Creates model documentation using custom formats for regulatory submissions and independent management reviews and maintains model version inventory
Model Deployment
Supports both RestAPI - Automated java object creation for deployment of non-linear models, model is available as REST API endpoint, making integration with web applications easier and ; Portable Model Object, suitable for Edge Deployment i.e. model is installed on device where predictions are needed