RiskSpan’s new Credit Model 7.1, announced on July 17, 2026, promises a purpose‑built credit analytics suite for non‑qualified mortgage (Non‑QM) assets, aiming to close a long‑standing data gap in the structured‑finance market.

RiskSpan, a long‑time provider of data and analytics for loan investors, introduced Credit Model 7.1 as the latest addition to its RiskSpan Platform. The solution pairs a Non‑QM‑specific credit engine with the company’s existing prepayment model, delivering a single, tape‑to‑cash‑flow workflow for investors, dealers, and risk managers. Built on a data set of roughly $87 billion in unpaid principal balances (UPB) from more than 226 000 Non‑QM loans, the model claims to capture borrower behavior that traditional agency‑focused tools miss.

What Credit Model 7.1 Brings

Credit Model 7.1 is a transition‑state credit framework that treats each loan‑type—Bank‑Statement, Debt‑Service‑Coverage‑Ratio (DSCR), Full‑Doc, and “Other”—as a distinct segment. Ten loan‑ and borrower‑level variables—including FICO score, mark‑to‑market loan‑to‑value (LTV), debt‑to‑income (DTI) ratio, and loan purpose—feed into three macroeconomic drivers. The model outputs probability‑of‑default (PD) and loss‑given‑default (LGD) estimates that can be rolled into cash‑flow projections in real time.

• AI‑driven tape cracking that automatically extracts loan attributes from dealer feeds, reducing manual data‑entry time by an estimated 40 % according to internal tests. AI‑driven tape cracking
• API access for seamless integration with downstream risk‑management platforms, enabling firms to embed the model into existing analytics stacks without a full platform migration.
• Back‑testing dashboard (beta) that lets users compare model outputs against historical performance, a capability rarely offered by third‑party vendors in the Non‑QM space.

Why Non‑QM Modeling Matters Now

The Non‑QM market has been on a rapid expansion trajectory. Morningstar DBRS reported that Q3 2025 Non‑QM RMBS issuance surged 97 % year‑over‑year to $20.9 billion, nearly doubling the prior year’s volume. Fitch Ratings notes an 800 % increase in rated Non‑QM/Non‑Prime RMBS issuance between 2020 and 2023, while KBRA projects total non‑agency RMBS—of which Non‑QM is a core component—to climb another 15 % in 2026, reaching $160 billion.

These growth rates have outpaced the analytical tools traditionally used by investors. Legacy credit models, originally designed for agency‑backed mortgages, often assume homogeneous borrower behavior and rely on outdated documentation categories. As dealer tapes arrive faster and allocation windows shrink, risk teams need a model that reflects the doc‑type driven nuances of Non‑QM borrowers.

A recent Gartner survey of 150 institutional investors found that 68 % consider “granular, asset‑class‑specific credit analytics” a top priority for 2025. Credit Model 7.1 directly addresses that priority by delivering a differentiated credit view for each documentation type.

Competitive Landscape

Few vendors offer a fully integrated Non‑QM solution. Existing alternatives fall into three buckets:

• Broad non‑agency models – Vendors such as Moody’s Analytics and S&P Global provide credit engines that span agency and non‑agency assets but lack doc‑type granularity.
• Standalone cash‑flow engines – Companies like BlackRock’s Aladdin focus on cash‑flow projection, leaving credit estimation to third‑party models.
• In‑house proprietary tools – Large banks often build bespoke models, but the effort and data requirements limit accessibility for smaller investors.

RiskSpan’s claim of being the only provider that couples a purpose‑built Non‑QM credit model with a prepayment engine and a unified workflow differentiates it from the competition. The integrated approach reduces vendor sprawl—a factor highlighted by a 2023 Forrester report that identified “vendor consolidation” as a key driver of operational efficiency in financial services.

Implications for Enterprise Finance Teams

For enterprise finance groups, the rollout of Credit Model 7.1 could reshape several workflows:

• Risk reporting – The model’s transparent, doc‑type‑specific PD/LGD outputs simplify audit trails, satisfying both internal governance and regulator expectations. risk reporting
• Portfolio optimization – By feeding more accurate credit risk into cash‑flow models, asset managers can fine‑tune tranche sizing and pricing, potentially improving yield spreads by a few basis points, according to RiskSpan’s internal simulations.
• Marketing and product positioning – Lenders that can demonstrate a data‑backed understanding of Non‑QM borrower risk may differentiate their loan products in a crowded marketplace, especially as fintech platforms increasingly bundle mortgage‑originated assets into embedded finance offerings. Fintech ecosystem

Enterprise teams that already leverage the RiskSpan Platform will likely see a low‑friction adoption curve, while newcomers may need to evaluate integration costs against the benefits of a single‑source analytics environment.

Market Landscape

The broader fintech ecosystem is witnessing a convergence of digital payments, open banking, and embedded finance. Non‑QM mortgages sit at the intersection, offering higher yields for investors willing to navigate higher credit risk. As open‑banking APIs make borrower data more accessible, models that can ingest real‑time financial statements—such as those based on bank‑statement documentation—will become increasingly valuable.

Simultaneously, blockchain‑based settlement solutions are gaining traction for RMBS issuance, promising faster clearing and reduced reconciliation overhead. A credit model that can output standardized risk metrics aligns well with these emerging infrastructures, facilitating smoother integration with distributed ledger platforms.

Top Insights

• Purpose‑built credit analytics: Credit Model 7.1 delivers doc‑type‑specific risk estimates, a capability missing from most legacy models.
• Data magnitude: Trained on $87 billion of UPB across 226 k loans, the model offers a depth of historical insight unmatched in the Non‑QM space. data set
• Integrated workflow: Combining credit, prepayment, and tape‑cracking reduces vendor sprawl and accelerates time‑to‑insight for risk teams.
• Market timing: With Non‑QM issuance projected to exceed $160 billion in 2026, the model arrives as demand for granular analytics peaks.
• Enterprise impact: Improved risk transparency supports tighter regulatory reporting, better portfolio optimization, and more compelling marketing narratives. marketing narratives marketing narratives

Get in touch with our fintech expert