For decades, the Agency Mortgage-Backed Securities (MBS) market operated on an implicit, foundational constant: a loan’s representative credit score meant Classic FICO. Capital markets desks built trading algorithms around it, risk engines calibrated duration based on it and specified pool buyers paid billions in pay-ups resting on their predictable prepayment behavior.
That baseline is shifting.
Recent government-sponsored enterprise (GSE) pool data reveals an emerging trend that demands the immediate attention of secondary marketing desks, desk traders and portfolio managers: A growing percentage of loans in newly issued Fannie Mae and Freddie Mac pools are underwritten using VantageScore 4.0 (VS4).
While the Federal Housing Finance Agency (FHFA) framed its credit modernization initiative as a step toward expanding homeownership, allowing approved lenders to choose between Classic FICO and VS4 on a loan-by-loan basis introduces structural shifts into the secondary market. What is intended to be an operational update for primary originations is becoming a complex valuation problem for the secondary market.
The illusion of equivalence
To the uninitiated, replacing a 720 Classic FICO with a 720 VantageScore 4.0 sounds like a distinction without a difference. Both sit neatly in the mid-700 credit tier. But to an MBS quant or a specified pool trader, treating these scores as interchangeable ignores the fundamental math of credit scoring.
VantageScore 4.0 incorporates trended credit data—tracking debt paydown behaviors over a 24-month window—and factors in non-traditional consumer payments like rent and utilities. Classic FICO provides a static, point-in-time snapshot of credit utilization.
Consequently, the two scores evaluate risk differently: score mobility and the other borrower score behavior.
- Score mobility: Studies show VS4 tends to yield higher numerical scores for lower-to-mid-tier credit profiles compared to Classic FICO.
- Borrower score behavior: A borrower who lands in a 740 score bucket due to active 24-month debt management displays a very different financial profile than a legacy 740 Classic FICO borrower.
When originators are given “lender choice,” they naturally select the scoring model that optimizes borrower eligibility or loan pricing. If a borrower scores 665 under FICO but 695 under VantageScore, the lender delivers the VS4 loan. Over time, this creates adverse selection. “High-tier” pools become diluted with uplifted mid-tier borrowers, while “low-tier” pools lose their most credit-agile borrowers.
Why specified pool pay-ups face compression
This model variance presents a challenge for the specified pool (“spec pool”) market. Investors pay cash premiums—pay-ups—above standard to-be-announced (TBA) pricing for specific pool attributes (such as low loan balance, high LTV, or low FICO) because they want one thing: prepayment protection.
When interest rates drop, investors need to know how fast borrowers will refinance. For 30 years, Wall Street econometric models mapped Classic FICO scores against refinance propensity with high accuracy. The market has minimal empirical, live-market constant prepayment rate (CPR) performance data for pools containing high concentrations of VantageScore 4.0 loans.
If an investor buys a “Low FICO (< 700)” spec pool expecting a sluggish, refinance-resistant borrower base, but that pool is blended with VS4 loans containing trended-data borrowers who actively manage debt, the pool’s prepayment speed could accelerate significantly during a rate rally.
The result? Pay-up compression.
As institutional buyers realize that credit-based spec pools contain mixed credit scores, they will price in the uncertainty. The extra 32nds and 64ths that originators rely on for secondary execution will shrink across credit-sensitive spec buckets.
The emerging market bifurcation
The secondary market rarely leaves money on the table without adapting. As disclosure data makes VS4 loan concentrations visible down to the pool Committee on Uniform Securities Identification Procedures (CUSIP), market structure will inevitably bifurcate in the following ways:
- Tiered pay-up quotes: Dealers will stop quoting pay-ups purely on weighted average credit scores. Instead, trading screens will introduce VS4 concentration limits (e.g., “< 15% VS4” vs. “> 50% VS4”).
- The “Pure FICO” premium: Fully segregated, 100% Classic FICO pools will command a pricing premium over blended pools simply because they offer model continuity and historical performance data.
- Flight to non-credit spec stories: As credit-based spec pools take a haircut due to model noise, investor capital will rotate heavily into structural spec pools where prepayment protection is tied to physical loan traits rather than credit scores—most notably low loan balance (≤$85k / ≤$110k) and high-cost state pools.
What market participants must do now
Lenders cannot treat the credit score modernization initiative as purely an underwriting or LOS implementation project.
If capital markets desks do not monitor credit score model delivery in their pipelines, they risk dumping blended loans into spec pool deliveries unthinkingly, taking a direct hit on execution. Lenders must build algorithmic pooling strategies that monitor VS4 concentrations, walling off high-pay-up spec pools or routing pure VS4 originations into dedicated investor buckets.
For buy-side investors and desk analysts, the mandate is equally urgent: update risk engines to track loan-level credit model identifiers immediately.
While credit score modernization is often framed as a driver of borrower inclusion, its uncoordinated rollout threatens to create costly friction across the mortgage ecosystem. However, in the multi-trillion-dollar Agency MBS market, certainty is the ultimate currency. Until the secondary market has years of live prepayment data proving how VantageScore loans behave across rate cycles, blending credit engines without structural tiering will erode pay-ups, widen mortgage spreads, and force a total re-evaluation of credit-based prepayment protection.
Clifford Rossi, PhD, is the Executive-in-Residence at Johns Hopkins University, Carey Business School. Over a 25-year industry career spanning the S&L and 2008 financial crises, Dr. Rossi worked for both Fannie Mae, Freddie Mac as well as some of the largest banks and nonbank institutions in various C-level risk management positions.
This column does not necessarily reflect the opinion of HousingWire’s editorial department and its owners. To contact the editor responsible for this piece: [email protected].