RiskSpan Launches Credit Risk Model for Non-QM Loans

RiskSpan released Credit Model 7.1 on Friday, a new credit risk model tailored for the expanding non-qualified mortgage (non-QM) market. This model, available through the RiskSpan Platform, allows users to analyze loan data and generate cash-flow projections within a unified environment, supplementing the company's existing prepayment modeling tools.

The launch coincides with significant growth in the non-QM securitization market. Morningstar DBRS reported that non-QM residential mortgage-backed securities (RMBS) issuance nearly doubled year-over-year in the third quarter of 2025, reaching a record $20.9 billion, a 97% increase from $10.6 billion in Q3 2024. Fitch Ratings noted that issuance across its rated non-QM and non-prime RMBS portfolio surged by over 800% between 2020 and 2023. KBRA projects that broader non-agency RMBS issuance, which encompasses non-QM loans, will see an additional 15% growth in 2026, reaching $160 billion.

RiskSpan stated that Credit Model 7.1 is engineered to better represent the unique characteristics of non-QM loans. It models borrower behavior across various documentation types, including bank statement, debt-service-coverage ratio (DSCR), and full documentation loans. The model incorporates 10 loan- and borrower-level variables, such as credit scores, mark-to-market loan-to-value ratios, debt-to-income ratios, and loan purposes, alongside three macroeconomic factors.

This new model was trained on approximately $87 billion in unpaid principal balance, derived from roughly 226,000 non-QM loans originated between January 2018 and August 2025. The release also features artificial intelligence-powered loan tape analysis tools and provides API access for clients who wish to integrate the model into their existing systems.