RiskSpan Debuts NonQM Credit Model to Capture Surging Market

The new model addresses a critical friction point in structured finance: the disconnect between rapid market growth and outdated evaluation tools. While Fitch Ratings tracked an 800% growth in Non-QM portfolio issuance between 2020 and 2023, the industry has struggled with models that fail to account for the specific borrower behaviors tied to bank statements, debt-service coverage ratios, and full-documentation loans. Credit Model 7.1 shifts the paradigm by utilizing a transition-state framework trained on $87 billion in unpaid principal balance across 226,000 loans.

RiskSpan’s platform integrates this credit model with its existing prepayment engine, offering a unified tape-to-cashflow workflow. By incorporating ten loan-level factors, such as mark-to-market LTV and DTI, alongside three macroeconomic drivers, the tool provides the granularity institutional investors currently lack. Divas Sanwal, Head of Modeling at RiskSpan, noted that generic frameworks often overlook the nuances of documentation types, necessitating a rebuild from the ground up to ensure transparency for risk teams and auditors. The release also features AI-powered tape cracking and API access for seamless workflow integration.