dv01’s proprietary risk models use macro and micro insights, as well as loan level data from five major US online lenders, to predict default and prepayment risks for online loans. Our cashflow engine, built from the ground up, integrates with our risk models to power additional Intelligence offerings, such as anomaly detection and stress testing.
Gain in depth understanding into origination and performance trends
Review how your cut of the pool compares to the full universe of loans
Identify under and over performing cohorts with our AI algorithm
Predict future returns based on prepay, default, and severity assumptions
Analyze how adverse economic scenarios impact marketplace asset performance
Years of credit research show that one-period transition matrix models generate compounded default curves that are not consistent with historical/empirical curves (Bluhm C and L Overbeck 2007).
This approach utilizes historical data from more mature lenders to infer future behavior for lenders without as long a history, while still accounting for differences between the lenders.
Some models are built with a one-size fits all approach. dv01 models are built by domain specific experts and take into consideration unique data characteristics, such as the structure and dynamics of default and prepayment curves, to intelligently select predictive variables per model.
dv01 has access to more loan level detail than any other online lending reporting platform. Direct integration with lenders means we receive information direct from the source, and normalization and data validation processes ensure all data used for modeling is standardized and updated.
Leader in online lending data integration and analytics, for both loans and bonds
Independent third party, focused on objective assessment of portfolio risk
Access to extensive historical data for all major lenders, covering >90% of the market
Expertise in traditional fixed-income markets, machine learning, and scalable IT infrastructure