Axtria Insights

Axtria Insights

CASE STUDY
Axtria Built a Loss Mitigation Engine (Prepayment, Delinquency, Default Curves, Optimization Engine) for a Large Mortgage Servicing Firm

Situation

The client wanted to build a world class automated loss mitigation platform incorporating optimization algorithms, prepayment curves, default curves, cash flow models


Approach

Axtria built a suite of models to construct the entire automated loss mitigation platform. Ket stages towards building the platform included:

Data Collection
  • Collected the following types of data for last 10 years
    − Borrower data
    − Mortgage data
    − Internal payment and performance data
    − Transaction data

Data Preparation
  • Validation and QC of data and performed logical checks
  • Clarified data Issues from client sources, documented key assumptions
  • Outlier and Missing value  treatment

Data Preparation
  • Validation and QC of data and performed logical checks
    − Cash out Incentive
    − Refinance Incentive
    − Borrower Risk  Profile
    − Home Equity (HPI)
    − Property Occupancy
    − Loan characteristics

Model Development
  • Modeled the pre-payment/delinquency/default probabilities using distribution fitting functions like sigmoidal
  • Performed loan age sub population validation and calendar validation

Implementation & Documentation
  • Documented key data assumptions, data analysis, and approach in detail
  • Thorough documentation of entire modeling exercise
  • Provided support in approvals and implementation

Result

Below are the results based on the models implemented to construct the automated loss mitigation platform :


Loss Mitigation Engine Prepayment Delinquency Default Curves Optimization Engine for Mortgage Servicing Firm