Advanced Analytics

We take a data intensive and highly analytical approach to credit risk assessment and portfolio management

Our processes rely on:

  • Hundreds of input fields collected from the lending platforms or generated internally
  • Detailed loan level performance data that we continuously update and analyze on a daily basis
  • Advanced machine learning driven model development
  • Exhaustive individual level performance forecasting and tracking

The combination of all these tools helps us generate very accurate loan level performance forecasting which we further complement with robust portfolio management practices:

  • Rigorous diversification
  • Continuous monitoring
  • Timely weight adjustments
  • Full loan valuation framework

    We forecast all the components of credit performance at an individual loan level and across the full lifetime of the loan. This includes, default probability estimate, prepayments estimates, principal amortization, effective yield and return.

  • Continuous Validation

    We continuously collect actual loan performance and compare to our forecast. This allows us to detect performance deteriorations very early and make timely adjustments to our models or portfolio weights.

  • Machine Learning Algorithms

    Our models are build using machine learning techniques that have proven to be the most effective at predicting default risk. Our model training environment continuously incorporates new data which allows us to quickly recalibrate and deploy new models.

  • Diversification

    We manage volatility by measuring and controlling our concentration levels across multiple dimensions pertinent to credit performance and systemic risk correlation.

  • Real-time Monitoring

    We constantly look for opportunities to optimize our portfolios’ structure by monitoring and analyzing portfolio performance in real-time. This allows us to spot attractive segments to overweight and/or underperforming segments to exit in a timely way.

  • Automated Adjustments

    We have built-in limits and controls that prevent exposure breaches and automatically optimize or rebalance our portfolios by adjusting segment weights.