Linkedin    Twitter   Facebook

Get Started
Log In

Linkedin

Category: Article
Home / Article /

AI’s Uneven Impact on Labor Demands a Local Housing View

By: Scott Anderson and Bernadette Kogler

AI is already disrupting parts of the U.S. labor market. The more contested question for residential mortgage investors is how and where. The dooms-day camp sees broad structural displacement where white-collar knowledge workers are replaced by AI agents. The optimists counter that AI makes workers more productive, that falling labor costs will expand demand leading to net employment gains. Both camps agree on one thing: the impact will not be uniform. Disruption will concentrate in pockets — specific occupations, metro centers, and income bands.

That specificity is precisely what makes this a mortgage credit problem. Local labor shocks have always transmitted to local housing markets through a familiar mechanism: income impairment reduces demand which pressures home prices, and price declines erode the collateral cushion that protects investors. What varies is the speed, the depth, and whether prices recover once the shock stabilizes.

Early indicators are consistent with the thesis. A Wall Street Journal analysis found the unemployment rate for recent college graduates ages 20–29 at 7.1% as of October 2024 — nearly double the 4.1% rate for the general population — with computer science graduates at 6.1% unemployment and computer engineering majors at 7.5%. The Federal Reserve of NY also published data showing labor market conditions worsened for recent college graduates at the end of 2025, with the unemployment rate climbing to about 5.7% in Q4 2025 — and underemployment rising to 42.5%, its highest level since 2020.

The occupations most exposed to AI displacement cluster in precisely the metros where high-balance prime mortgages are most concentrated, changing the traditional metrics that investors rely on for early warning. AI-driven labor disruption is not a macro thesis to be managed with national averages. While mortgage borrower occupation data is not complete or granular enough to drive AI job loss assumptions, property location data is. The risk needs to be identified, stress-tested, and monitored at a granular level — before it shows up in delinquency data.

The good news is that the data, models and historical reference points exist. With the growth of Non-QM lending, RTL and loan products that carry credit risk without an insurance wrapper, treating AI-driven labor disruption as a present-tense portfolio risk is not optional. To move from thesis to numbers, we ran a portfolio-level stress test on the universe of securitized Non-QM mortgages – starting with how AI exposure is measured at the local level.

Key Findings for investor consideration:

  1. AI labor disruption is expected to have a higher disproportional impact on white-collar professional jobs that are concentrated in certain “knowledge-work” centers.
  2. The largest job markets and the deepest credit pools are systematically the most AI-exposed, suggesting AI risk and portfolio risk overlap rather than diversify.
  3. Stress-testing the $196B universe of securitized Non-QM mortgage loans shows the magnitude is meaningful: under an AI-driven stress scenario calibrated to GFC-level shocks but re-allocated to AI-exposed geographies, RiskSpan’s credit model projects universe-wide cumulative defaults to nearly double and losses more than quadruple.
  4. Securitized deals benefit from built-in geographic diversification, which dampens the absolute impact across deals. Whole-loan investors, who construct their own geographic mix, face concentration risk that accumulates silently if exposure isn’t an explicit input into selection and surveillance.
  5. Act now — put the infrastructure in place to use granular data to monitor unemployment, HPI and loan performance and apply scenarios where relevant.

Measuring AI Exposure at the Local Level

Leveraging research published by the Brookings Institution, RiskSpan used their geographic AI exposure scores as the foundation for a credit stress scenario. The resulting county- and metro-level scores follow a similarly clear geographic pattern to that published by Brookings. Highly educated, high-paying knowledge-work centers — Santa Clara (42.8% exposure), King County WA, New York County, DC, Boston — show the highest exposure. Smaller industrial and rural counties score considerably lower. As a sanity check on the methodology, we aggregated and plotted state-level AI exposure against the share of each state’s adult population with a bachelor’s degree or higher (a well-understood proxy for knowledge-work concentration). The two track closely, which is the result one would expect — it confirms the Brookings measure is capturing what it intends to and isn’t producing arbitrary geographic rankings.

Article content
Figure 1

Unlike earlier automation, generative AI concentrates in exactly the metros that house the largest mortgage markets: as the chart below shows, the most AI-exposed MSAs are also among the deepest non-agency RMBS markets.

Article content
Figure 2

One important caveat before moving on to how we used this data. As Brookings is careful to note, “‘exposure’ does not speak only to the displacement of workers; it also may involve their ‘augmentation’ through rapidly improving AI tools.” The same coder whose role is highly exposed could be made dramatically more productive rather than displaced; the same financial analyst could spend less time on rote modeling and more on judgment-intensive work. Whether exposure plays out as net displacement, net augmentation, or some mix will vary by occupation, employer, and how quickly the technology evolves. For purposes of this exercise, we focus on the displacement story: the scenario where AI exposure translates into labor market shocks that transmit to housing markets and mortgage credit. That’s not the only possible outcome, but it’s the one investors need to be prepared for.

Translating that exposure into a portfolio-level credit stress test on the Non-QM universe, the largest segment of the non-agency RMBS market, required two steps.

Step 1: Construct an AI-driven stress scenario calibrated to the GFC.

Rather than forecast AI displacement magnitude — a genuinely uncertain exercise — we sized the stress to a known reference point: the total HPI decline and unemployment increase observed during the GFC. The GFC offers a useful calibration episode because the empirical relationship between local unemployment shocks and HPI declines is strong and well-documented, regardless of which direction the causation primarily ran. The chart below shows that relationship across MSAs — a clear negative slope linking the spike in unemployment rates to the depth of the HPI decline.

Article content
Figure 3

The critical observation for our purposes is what happened in 2008: the metros most exposed to AI today (DC, Boston, San Francisco, Seattle) cluster toward the middle of that scatter plot. They experienced more moderate impacts than the hardest-hit metros, because the speculative and underwriting excesses that drove the GFC were concentrated elsewhere — Las Vegas, Phoenix, Miami, the Inland Empire. An AI-driven shock would invert that geography.

We took the GFC’s national-level shock magnitudes and redistributed them, concentrating the stress in MSAs with high AI exposure. Starting from Brookings’ occupation-level AI exposure scores rolled up to MSA and state-level, we scaled both the unemployment and HPI shocks to each MSA in proportion to its AI exposure share. The result is a per-MSA stress profile that adds up to GFC-magnitude pain at the national level but lands very differently on the map.

The chart below translates that scaling into implied HPI shock magnitudes across the metros where non-agency RMBS exposure actually sits. The metros at the bottom end of both axes (high portfolio balance and severe implied HPI shock) are where the AI thesis would hit hardest if it materializes.

Article content
Figure 4

Step 2: Run the loan universe through two scenarios.

We took the $196B universe of Non-QM mortgage loans (~500K loans across 810 active Non-QM deals as of April 2026) and ran these loans through RiskSpan’s proprietary credit model under two scenarios: a baseline and the combined AI-driven Unemployment + HPI stress scenario described above. Every loan received a projected cumulative default rate and cumulative loss rate under each scenario, isolating the AI-driven impact at the loan level.

Findings

Under the AI-driven combined stress scenario, universe-wide cumulative defaults nearly double (3.6% → 7.4%) and cumulative losses more than quadruple (0.5% → 2.2%). The steeper loss response reflects the familiar non-linearity between defaults and severities under HPI stress. The mechanism behind those aggregate numbers becomes clearer when you look at it geographically.

Finding 1 — AI exposure drives MSA-level credit performance.

This finding is by construction — we built the stress scenario by re-allocating shocks to AI-exposed MSAs, so the model output should track AI exposure. The MSA-level chart confirms it does. MSAs at the high end of AI exposure (DC, San Francisco, Boston, Seattle) see over an 8% percentage point (pp) increase in cumulative defaults compared to MSAs at the low end. Riverside, near the lower end of the AI exposure range at just under 30%, sees only a 2.6% pp increase in cumulative defaults. There’s MSA-specific noise from collateral mix, but the geographic risk concentration the AI thesis predicts shows up cleanly in the model output. This is the core mechanism: re-allocate the stress to AI-exposed geographies, and the credit impact tracks the re-allocation.

Article content
Figure 5

Finding 2 — The aggregate impact is shaped primarily by leverage.

Decomposing the universe results by current LTV bucket explains where the headline numbers come from. The first thing worth noting is what the LTV distribution looks like: less than 5% of total non-QM balance sits above 80% LTV, and roughly half is below 60%. That’s a meaningfully larger equity cushion than the borrower base of the mid-2000s, when the proliferation of low- and no-down-payment products left investors with thin protection against any HPI decline. Today’s Non-QM universe starts the AI stress scenario with a lot more skin in the game.

Within that distribution, stress impact scales monotonically with leverage, exactly as one would expect: low-LTV loans (≤40%) see a 1.3pp increase in defaults (from 0.96% to 2.24%), while high-LTV loans (>80%) see a 6.7pp increase (from 5.97% to 12.62%), though the relative impact on cumulative defaults is broadly stable across LTV buckets, with defaults about doubling in each bucket. The same pattern holds for losses, where the highest LTV buckets see losses rise by over 3pp under stress versus 0.16pp for the lowest LTV bucket. FICO declines slightly as LTV increases (from 756 for the lowest LTV bucket to 745 for the highest), so the projected cumulative default differences across LTV buckets are driven, at least partly, by underlying borrower credit as well.

Article content

Finding 3 — Deal-level concentration matters, but only in relative terms.

When we rolled up by deal rather than by MSA, the picture got more interesting. The absolute (pp) impact across deals shows essentially no relationship with the deal’s AI exposure share — high-AI-exposure deals don’t show systematically larger pp increases. But the relative impact (stress / base − 1) is clearly correlated with AI exposure.

One plausible explanation is that deal-level base default rates vary widely based on traditional credit characteristics — LTV mix, FICO mix, doc type (Full Doc vs. DSCR vs. Bank Statement), occupancy. That variation in base rates may be dominating the AI signal in absolute terms while leaving it visible on a relative basis. Either way, the relative measure is the cleaner lens for isolating AI exposure’s marginal contribution to deal-level risk.

There’s also a more reassuring story embedded in these charts, at least for securitized mortgage assets. Compare the deal-level scatter to the MSA-level scatter and you’ll notice the deal-level points cluster in a meaningfully narrower range — both on the AI exposure axis (X) and on the impact axis (Y). That tighter clustering is geographic diversification at work: even the most AI-concentrated deal in the universe pulls in loans from enough different MSAs that its blended AI exposure sits closer to the universe average than the most exposed individual MSA does. Diversification within deals is doing what it’s supposed to do — dampening the variance of outcomes — even though the unit of analysis is smaller than an MSA. The magnitude of the differential across deals is bounded by how concentrated any individual issuer is willing to get.

Article content
Article content

The implication for whole-loan portfolios is sharper. Whole-loan investors don’t get the benefit of pre-built deal diversification — they’re constructing their own geographic mix one loan or pool at a time, and concentration risk accumulates silently if it isn’t an explicit input into selection and surveillance. AI exposure is one such risk, and a relatively new one, but it sits alongside more familiar concentrations — natural disaster exposure, regional employment dependencies, single-industry metros — that warrant the same lens. The toolkit is the same in each case: granular geographic data, stress scenarios calibrated to defensible reference points, and loan-level credit modeling. Whether the goal is initial portfolio construction or ongoing surveillance of a seasoned book, the ability to identify these concentrations early — before they show up in delinquency data — is what separates active risk management from passive observation.

What This Means for Investors

History offers instructive analogies including the energy-patch metros of the mid-2010s when oil prices collapsed — Houston, Oklahoma City, and Calgary experienced home price corrections in the range of 5 to 15 percent. Regional manufacturing declines showed a similar story over a longer time period — the long contraction of auto-sector employment in Detroit, for example, played out in local housing markets over years while national HPA remained positive. In neither case did the national trend warn you that specific MSAs were in distress.

Acting now means two things in practice. The first is tracking — building the early-warning infrastructure before you need it. That means monitoring MSA-level unemployment, not just the national rate; watching labor force participation shifts; and tracking occupational mix changes in the markets where your collateral is heaviest. These indicators move before delinquency does. The second is scenario preparation — running stylized AI displacement shocks against the portfolio today using MSA-level or county-level HPA overrides and occupational exposure overlays. The goal is not a forecast but rather a map of your exposure — which corners of the portfolio are most sensitive to particular labor market stress, and what the loss distribution looks like if the more adverse scenarios begin to materialize. That map is worth building now, while the cost of being wrong about the timing is low.

RiskSpan is releasing a new Non-QM-specific credit model in the coming weeks, estimated directly on Non-QM performance data and designed around the underwriting features (DSCR, bank statement, expanded LTV) that distinguish these segments. Re-running this scenario and other stress scenarios on the new model is on the near-term roadmap and may sharpen some of the findings reported here. For investors who want to evaluate their own exposure, the analytical infrastructure described here can be applied to specific portfolios on request.

Note: The Brookings Institution has been tracking AI’s labor market impact since 2019. A 2024 Brookings piece analyzed data from OpenAI to measure occupational AI exposure as the share of tasks where AI could reduce human completion time by at least 50%. AS one would expect, the cognitive/manual divide is evident in the rankings, with computer and mathematical work atop the list at 75% while construction sits at the bottom at 5.6%.

Their 2025 piece extends that work to geography by multiplying each occupation’s exposure rating by its share of local employment in each county or metro, then aggregating. Counties with large workforces in highly exposed occupations (software developers, financial analysts, lawyers, marketing professionals) score higher than those whose employment skews toward construction, agriculture, or in-person services.

Key References:

Generative AI, the American worker, and the future of work | Brookings

The geography of generative AI’s workforce impacts will likely differ from those of previous technologies | Brookings

There Is Now Clearer Evidence AI Is Wrecking Young Americans’ Job Prospects – WSJ

www.newyorkfed.org/research/college-labor-market

Models & Markets Update: April 2026 

Register here for next month’s call: Thursday, May 21st, 2026, 1 p.m. ET. 

Key Takeaways 

  • Prepayment models continue to perform well, with March speeds driven by a February rate rally and day count effects 
  • A new Non-QM Credit Model (CM 7.1) is on track for release near end of Q2 2026, with a dedicated webinar planned for end of May or early June 
  • Housing turnover analysis reveals rate sensitivity at positive refinancing incentive levels — a finding that will inform the next prepayment model 
  • Mortgage rates hit a six-month high in March before pulling back; rates are expected to remain above 6% through 2026 and 2027 
  • No Federal Reserve rate cuts are expected in 2026; the consumer remains under pressure from elevated rates and rising credit card balances 

You can read the recap below or click here for the entire recording. 

Prepayment Model Back-Testing: April Factor Data Update 

The prepayment model continues to track realized speeds closely across Agency collateral. Results are available on the Edge platform under the Vertex module. 

Fannie/Freddie — Discount Coupons (WAC 5.5 and Below) 

Discount coupons showed a modest uptick in March speeds, driven primarily by two factors: 

  • A day count effect: March had three more collection days than February 
  • Seasonal turnover patterns typical of the spring housing market 

Figure 1: FN/FH Discount Coupon Back-Testing — Model CPR vs. Observed CPR 

Fannie/Freddie — Premium Coupons (WAC 6.0 and Higher) 

Premium coupons saw a sharper increase in prepayment speeds in March, driven primarily by the rates rally in February. With rates subsequently moving higher in March, May factor data is expected to show a decline in speeds — a clear convex response consistent with model expectations. 

Figure 2: FN/FH Premium Coupon Back-Testing — Model CPR vs. Observed CPR 

GNMA — FHA and VA Segmentation 

GNMA performance showed a similar pattern to Fannie/Freddie across discount and premium coupons. A notable enhancement this month: the team has introduced the ability to split GNMA back-testing results by FHA vs. VA segments on the Vertex report, providing additional analytical granularity. 

Two segment-level observations: 

  • FHA: FHA: The model has shown a slight drift in prepayment speeds over the past year. This is primarily attributed to the FHA trial modification policy change in 2025, under which servicers are no longer required to buy out delinquent loans — a policy shift that has meaningfully reduced prepayment speeds relative to historical levels. 
  • VA: VA: A discrepancy between modeled and actual speeds reflects the default VA-to-PMMS spread assumption being too tight. Users can adjust the VA spread to current market levels within the platform to bring results into closer alignment with observed speeds. 

Figure 3: FHA Segment Back-Testing 

Figure 4: VA Segment Back-Testing 

New Non-QM Credit Model: CM 7.1 

Guanlin Chen from the Quantitative Modeling Group presented an overview of the upcoming Non-QM Credit Model, version 7.1, expected to be available to all users near end of Q2 2026. 

Model Structure 

CM 7.1 follows the same three-component framework as RiskSpan’s agency credit model: 

  • 1. Transition Model — generates a time-varying transition matrix estimating transition probabilities across delinquency states and time periods 
  • 2. Liquidation Timeline Model — applied once a loan enters default 
  • 3. Severity Model — estimates final losses on the defaulted balance 

A key design decision: the model is built with separate sub-models for each documentation type, consistent with the segmentation used in the Non-QM prepayment model: Bank Statement, DSCR, Full Doc, and Other. This segmentation reflects the meaningfully different performance characteristics across these loan types and allows for more accurate, documentation-specific projections. 

Transition Matrix Design 

The transition matrix tracks loans across delinquency states — current (0), one-month delinquent (1), two-month delinquent (2), foreclosure (F), and REO (R) — with additional granularity for delinquency history.

Back-Testing Results 

Initial back-testing of the 30-day delinquency transition demonstrates that the model captures the overall trend in Non-QM credit performance well across all four documentation types. The COVID period was intentionally excluded from model training — including it would have caused extreme unemployment levels to dominate the model and distort sensitivity to other risk factors. 

Figure 5: NonQM Credit Model — Current to 30DPD Transition: Actual vs. Projection by Documentation Type 

A dedicated webinar covering CM 7.1 in detail is planned for end of May or early June. Please stay tuned. 

Housing Turnover Analysis: Rate Sensitivity at Positive Refinancing Incentive 

Shane Lee from the Quantitative Modeling team presented new research on housing turnover behavior in a positive refinancing incentive environment. 

Background 

Total prepayment has two components: housing turnover (prepayment driven by home sales) and refinancing. In current model design, housing turnover is assumed to be weakly rate sensitive — and in the positive refinancing incentive regime, sensitivity is held at zero. The question the team set out to answer: is that assumption correct? 

Data Sources 

  • NAR: National Association of Realtors (NAR) Existing Home Sales — measures the number of homes sold including single-family, condo, and co-op properties (including sales without mortgages) 
  • Equifax ADS: Equifax ADS Data — tracks trade lines per consumer, allowing the team to identify housing turnover by flagging cases where an existing mortgage closes and a new mortgage originates at a different ZIP code for the same borrower 

Figure 6: ADS vs. NAR Data Comparison — Prepaid Mortgages vs. Home Sale Units 

Key Finding 

During the post-COVID refinancing boom, housing turnover activity increased significantly — by almost 50% above the baseline level. This elevated turnover coincided with the period of low rates and high refinancing activity, driven in part by the work-from-home migration wave. 

Figure 7: Housing Turnover CPR — ADS Data (nearly 50% above baseline during COVID refi boom) 

This finding suggests that housing turnover is more rate-sensitive in a positive refinancing incentive environment than current models assume — a potential source of underestimation when projecting prepayment speeds in a low-rate environment. Research is underway to incorporate this into Prepayment 4.0. 

Macroeconomic Update: April 2026 

Federal Reserve — No Rate Cuts Expected in 2026 

CME FedWatch futures currently indicate no Federal Reserve rate cuts this year. The overall expectation is that the Fed funds rate (currently 350–375 bps) will remain unchanged through year-end 2026. 

Figure 8: Federal Funds Target Range — Upper Limit (Source: FRED) 

Mortgage Rates — Elevated and Volatile 

Mortgage rates hit a six-month high in March, with Freddie Mac’s primary rate reaching 6.45% and Mortgage News Daily data showing rates approaching 6.64%. Rates have since pulled back modestly as some geopolitical uncertainty subsided. The 10-year Treasury rate is expected by market consensus (econforecasting.com) to remain above 4% for the next three to five years — implying mortgage rates are unlikely to fall significantly below 6%. 

Figure 9: 10-Year Treasury Yield — Historical and Market Consensus Forecast 

Figure 10: Primary Mortgage Rate Trend 

Unemployment and Inflation 

  • March unemployment rate: 4.3% — trending upward 
  • PCE (excluding food and energy): approximately 3% — still above the Fed’s 2% target 

Figure 11: Unemployment Rate 

Figure 12: PCE Inflation (ex. Food & Energy) 

Consumers continue to face pressure from elevated gasoline and oil prices. Credit card balances have risen significantly over the past two years, adding to the financial strain on households. 

Home Prices — Stabilizing but Elevated 

Home price growth remains positive but has decelerated substantially from the 20%+ year-over-year peaks observed in mid-2022: 

  • Case-Shiller National Index: approximately 1% year-over-year growth 
  • 10-City Composite: slightly above the national index 
  • FHFA All-Transaction Index: somewhat stronger, indicating variation across market segments 

Figure 13: Case-Shiller U.S. National Home Price Index — Year-over-Year % Change 

Two regional case studies highlight the range of outcomes: Austin, TX and Boise, ID both experienced peak growth of 30–35% in 2021–2022, followed by sharp declines through 2023, and are now returning to modest positive territory. Housing supply remains severely constrained. 

Summary 

Topic Key Takeaway 
Prepayment Model Performing well overall; March speeds driven by February rate rally and day count effects 
GNMA Segmentation New FHA/VA split available in Vertex; FHA drift tied to 2025 trial mod policy change 
NonQM Credit Model CM 7.1 on track for end of Q2 2026; dedicated webinar coming end of May / early June 
Housing Turnover Rate sensitivity confirmed in positive refi regime; ~50% above baseline during low-rate period; research underway for Prepayment 4.0 
Mortgage Rates Hit six-month high of 6.64% in March; expected to remain above 6% through 2026–27 
Fed Policy No rate cuts expected in 2026; Fed funds rate at 350–375 bps 
Home Prices Growth slowing (~1% nationally); supply constraints persist 

We continue to add additional analytics reports on the Platform. Please visit www.riskspan.com/request-access to request free access. 

As always, please feel free to contact us to discuss or learn more.

Models & Markets Update: March 2026 

Register here for next month’s call: Thursday, April 16th, 2026, 1 p.m. ET.

Key takeaways from this month’s call: 

  • Non-mortgage credit is deteriorating more rapidly than mortgage credit 
  • BNPL usage may be masking underlying financial strain 
  • Macroeconomic conditions are likely to remain restrictive, reinforcing current trends 
  • Prepayment models remain well-calibrated, even as borrower behavior begins to shift 

You can read the recap below or click here for the entire 20-minute recording.  

Credit Performance by Asset Class 

The data shows a clear divergence between mortgage and non-mortgage credit: 

  • Mortgage delinquencies remain relatively low, supported by tighter underwriting standards 
  • Credit card delinquencies have increased meaningfully since 2022 
  • Auto loan delinquencies are approaching levels observed during the Global Financial Crisis, particularly among younger borrowers  

The following charts from NYFed illustrate how younger age cohorts are consistently exhibiting higher delinquency rates across credit types (mortgages, credit cards, and autos). 

BNPL Usage as a Potential Blind Spot 

Buy Now, Pay Later (BNPL) usage continues to expand. 

Adoption is highest among younger borrowers. 

A meaningful portion of usage is for essential expenses such as groceries  

Because BNPL obligations are not consistently captured in traditional credit metrics, they may obscure underlying levels of consumer leverage and stress. 

Macroeconomic Outlook: Rates Expected to Remain Elevated 

The macroeconomic environment continues to support a “higher-for-longer” rate outlook. 

  • Market expectations suggest no Federal Reserve rate cuts through 2026. 
  • The 10-year Treasury rate is expected to remain above 4% over the next several years. 
  • Mortgage rates, after declining earlier in 2026, have risen again and are expected to remain near or above 6%. 

At the same time: 

  • Inflation remains above target levels 
  • Unemployment is trending upward  

These conditions suggest a continued tightening backdrop for borrowers, with limited relief from monetary policy in the near term. 

Housing Market: Moderation Continues 

Home price growth remains positive but has slowed: 

  • Case-Shiller index shows modest annual growth (~1.3%) 
  • FHFA index indicates somewhat stronger growth (~3.3%)  

Differences between indices suggest variation across market segments, with relatively stronger performance in more affordable segments and geographic differences on home prices. 

Against this macro and consumer backdrop, prepayment behavior continues to evolve. 

  • Prepayment models remain closely aligned with realized speeds across FN/FH and GNMA collateral, as shown in the coupon-level comparisons.  
  • Refinance behavior is well captured, including sensitivity to changes in mortgage rates.  

There are, however, early indications of shifting borrower behavior: 

  • Prepayment speeds increased in February despite fewer collection days, suggesting a gradual weakening of the mortgage rate “lock-in” effect.  
  • Short-term rate increases may moderate this trend, but the directional change is notable. 

GNMA Segmentation Enhancements 

The introduction of FHA and VA segmentation in GNMA back-testing provides additional analytical detail. 

FHA performance shows some divergence, likely reflecting recent policy changes affecting delinquent loan buyouts. 

VA results are more sensitive to spread assumptions and can be adjusted to align with market conditions. 

We continue to add additional analytics reports on the Platform. Please visit www.riskspan.com/request-access to request free access. 

As always, please feel free to contact us to discuss or learn more.

From Household Debt to Non-QM Credit: February Models & Markets Recap 

Register here for next month’s call: Thursday, March 19th, 2026, 1 p.m. ET. 

In this month’s Models & Markets call, RiskSpan’s quantitative modeling team tackled: 

  • The record debt levels now carried by U.S. households (and the consumer stress that is building beneath the surface); 
  • The likely persistence of higher rates; 
  • RiskSpan’s forthcoming non-QM credit model, and; 
  • (as always) how RiskSpan’s prepayment model is performing 

You can read the recap below or click here for the entire 20-minute recording.  

$18 Trillion in Household Debt (and Growing) 

 U.S. household debt reached $18.8 trillion at the end of 2025 and continues to climb. Mortgages account for the largest share at roughly $13 trillion, with auto loans, student loans, credit cards, and HELOCs making up the balance. 

Using conservative assumptions for average interest rates for each category, we estimate that these balances equate to roughly $1.1 trillion in annual interest payments and $2.5 trillion in total annual debt service payments – the approximate cost required each year just to keep households current. 

A Distributional Problem

The aggregate debt figure masks meaningful stress at the lower end of the income spectrum: 

  • A median household (~$80K gross income) may devote roughly 37% of disposable income to debt service. 
  • Bottom-quartile households (~$32K gross income) may spend 40–55% of disposable income servicing debt. 

Lower-income households are disproportionately exposed to higher-rate revolving credit and subprime auto loans, as opposed to 3–4% fixed-rate mortgages. The averages therefore understate the severity of strain on the more vulnerable segments. 

Are Reported Delinquencies Understating Stress?

Delinquencies are rising across income levels, particularly in lower-income areas. Lenders, however, may be quietly modifying or re-aging loans, particularly in consumer credit categories (e.g., auto loans). Such modifications can: 

  • Push missed payments to the back of the loan 
  • Reset accounts to “current” status 
  • Avoid immediate charge-offs 

While this suppresses reported delinquency statistics, borrower balances may continue to grow. This implies that reported delinquency rates may really be more of a floor, as aggregate DSCR and household stress may be greatly understated. 

Consumer strain is real and potentially worse than what is suggested by the headline metrics. 

Coming in Q2: RiskSpan’s Non-QM Credit Model! 

RiskSpan’s forthcoming non-QM credit model will feature four distinct documentation categories: 

  1. Bank Statement 
  1. Full Documentation 
  1. DSCR/Investor 
  1. Other (e.g., VOE, asset depletion) 

Each segment is modeled independently through a transition-matrix framework covering: 

  • Current 
  • 30-day DQ 
  • 60-day DQ 
  • 90+ DQ 
  • Termination (voluntary and involuntary) 

Prior delinquency figures prominently in the model, with clean loans having relatively low base transition rates from current to delinquent, while loans with prior delinquency history can experience transition probabilities up to 10x higher. Capturing this conditional risk dynamic is central to the model’s design. 

Back-testing (shown below for the Full Doc segment) indicates the model is tracking historical delinquency transition rates reasonably well, though development remains ongoing.

Macro Considerations

Consistent with prior months, the macro backdrop continues to reinforce a “higher-for-longer” rate environment. 

Fed and Policy Outlook 

  • Fed Funds expectations imply limited cuts in 2026. 
  • No immediate expectation of a March rate cut. 

10-Year Treasury 

  • Consensus forecasts suggest the 10-year Treasury will remain above 4% for the next 2–3 years. 
  • Recently, it has hovered around ~4.1%, down slightly from prior highs. 

Mortgage Rates 

Primary mortgage rates are approaching 6%, but not sustainably breaking below it. In our view, mortgage rates are likely to remain around or above 6% through 2026, possibly into 2027. 

Labor, Inflation, and Home Prices 

  • Unemployment ticked down slightly. 
  • Job creation surprised to the upside. 
  • Inflation remains sticky in the 2.5–3% range  
  • National home prices showed modest year-over-year growth (~1.4%). 

Traditional seasonal adjustments may be less reliable in today’s inventory-constrained housing market. Turnover seasonality appears to be shifting earlier in the year, with implications for both pricing and prepayment dynamics. 

Prepayment Model Performance: Stable & Improving

Despite macro headwinds and rising consumer stress, RiskSpan’s prepayment models continue to perform well. 

GSE Discounts (WAC 5.5 and Below)

Prepayments declined slightly in the most recent month, primarily due to fewer collection days and normal January seasonality (lower turnover). Overall model fit remains strong. 

One identified refinement: the model’s seasonal peak appears slightly delayed (June/July shifting toward August). This will be addressed in the next version update 

GSE Premiums (WAC 6 and Above)

Refinance speeds have been largely unchanged over the past two months. S-curve comparisons between December and January show no material differences once recount adjustments are made. A modest ~1.5 CPR change in recent data appears driven by turnover rather than refi activity.

Ginnie Mae: FHA vs. VA Enhancement

Performance across Ginnie segments remains solid, with recent prepayment dips again attributable to fewer collection days. However, we have observed divergence between FHA and VA: Modeled FHA speeds tend to be overestimated, while modeled VA speeds tend to be underestimated compared to recent historicals.

To address this, RiskSpan is adding a loan guarantor filter to the back-testing report, enabling FHA and VA splits (expected early March). This enhancement will improve transparency and precision in Ginnie performance analysis.

We continue to add additional analytics reports on the Platform. Please visit www.riskspan.com/request-access to request free access. 

As always, please feel free to contact us to discuss or learn more. 

AI Isn’t Coming to Structured Finance. It’s Already Here.

At SFVegas 2026, RiskSpan had a front-row seat to one of the most consequential conversations happening in our industry right now. I moderated a session on Agentic AI and the Securitization Lifecycle, while our CEO, Bernadette Kogler, participated in a panel on AI applications in structured finance. Across both rooms, the message was the same: we are done experimenting. We are redesigning workflows. 

The distinction that framed everything was simple but important. Copilots assist. Agents orchestrate. We are moving from AI that enhances what humans do to AI that executes complex, multi-step workflows — with humans supervising the process rather than driving every step. That shift changes everything about how we think about structured finance operations. 

And structured finance, it turns out, is almost perfectly suited for this moment. It is document-heavy, multi-party, process-driven, and data-intensive. From origination through surveillance, the lifecycle is fundamentally workflow-based. Panelists across both sessions shared real examples already in production: legal document review compressed from days to hours, AI-powered loan tape scrubbing before cash flow calculations, prompt-driven scenario generation replacing manual model configuration, and surveillance scaled across hundreds of deals per month. One striking observation: the future interface for structured finance may not be a UI at all — it may be entirely driven by prompts. 

But the panels were equally clear-eyed about what has to come first. Sixty-five percent of financial services firms are actively using AI, yet only 13% have deployed it across production processes. The gap between those two numbers is largely a data problem. The most sophisticated AI cannot overcome poor inputs, inconsistent loan tapes, or legacy system constraints. Firms that want to lead need to fix the foundation before layering on orchestration. 

Explainability matters just as much. With the EU AI Act and US fair lending enforcement raising the stakes, auditable, transparent models are not optional. And governance is shifting from “human in the loop” to “human over the loop” — a subtle but meaningful difference that requires defined accountability, model drift monitoring, and clear operational guardrails. 

The one-year outlook from both panels was notably concrete. Expect AI agents managing defined surveillance workflows, deeper cross-platform integration, and a sharper divide between early movers and everyone else. The competitive advantage will go to firms that clean their data, build explainability into their models from day one, and embed AI into operations — not just into pilot programs. 

Structured finance has always rewarded process discipline and deep domain expertise. That doesn’t change in an agentic world — if anything, it becomes more critical. The quality of an AI agent’s output is only as good as the prompts and parameters guiding it, and designing those well requires people who understand cash flow waterfalls, covenant structures, and credit risk at a fundamental level. Agentic AI doesn’t replace that expertise. It amplifies it. The firms that understand that will define the next chapter. 

At RiskSpan, we have spent years building at exactly this intersection — combining deep structured finance domain knowledge with purpose-built analytics infrastructure. That foundation is what makes it possible to deploy AI that actually works in production, not just in demos. The opportunity in front of our clients right now is significant, and we are focused on helping them capture it. 

Why AI Won’t Kill Asset-Backed Finance Software — and Why the Last Mile is the Moat

Every wave of financial technology innovation brings the same prediction: software will be commoditized. Today, that prediction is being applied to AI. If AI models can reason, summarize, and generate code, the thinking goes, B2B vertical SaaS becomes unnecessary. 

That conclusion is inherently wrong. ABF platforms are not feature layers, they are governed systems.  

The last mile of AI deployment isn’t friction—it’s the moat. 

ABF Is Not a “Promptable” Problem 

ABF platforms sit directly in the flow of capital allocation, risk management, and regulation. For asset managers deploying institutional capital, this creates a very high bar for reliable data, validated models and domain-specific workflow.  

The real question isn’t whether a system can produce answers. It’s whether it can produce results that are: 

  • Consistent over reporting periods and market cycles 
  • Explainable under stress and investor scrutiny 
  • Defensible and robust enough for LPs, investment committees, and regulators 

That high bar changes everything. It explains why technology adoption in financial markets moves cautiously and why legacy systems persist. These systems embed decision rights, controls, and institutional logic that can’t simply be recreated with better prompts. Any platform that ignores this reality will struggle to scale beyond pilots. 

Which leads to the obvious question: if AI is so powerful, where does it actually help? 

AI Accelerates Workflow — Not Accountability 

Applied correctly, AI can materially improve ABF workflows. It can ingest complex credit agreements faster, reconcile data across counterparties, flag covenant breaches, and reduce manual reporting work. In other words, AI increases operational leverage. 

But AI does not remove the need for explicit deployment configuration and governance. Institutions still must define who owns key assumptions, which decisions can be automated, and where accountability sits when outcomes affect capital. These embedded design choices (not prompts) ultimately determine whether a platform is trusted. 

AI compresses timelines, but responsibility remains fixed. Once this distinction is recognized, the broader implication becomes clear: AI does not eliminate the need for software. It raises the bar for it. 

Software Remains the System of Record 

The idea that AI replaces SaaS also misunderstands where SaaS enterprise value lives. Enterprise value in ABF doesn’t live in isolated insights. It lives in controlled systems of record and durable platforms that provide: 

  • Governed data and the system of record 
  • Embedded domain expertise 
  • Repeatable processes that survive personnel turnover 
  • A shared source of truth across counterparties, investment, risk, accounting, and investor relations 

AI without software discipline creates speed without stability. With it, AI becomes force-multiplying. The question, then, is what separates platforms that successfully integrate AI from those that don’t. 

The Real Differentiator: Deployment Intelligence at Scale 

What separates enduring platforms from feature-rich tools is not model sophistication—it’s deployment intelligence — the ability to integrate AI into live production environments without weakening controls. That requires: 

  • Controlled data pipelines designed for real-world imperfections 
  • Configuration layers that adapt to fund-specific structures without breaking controls 
  • AI outputs that are transparent, and auditable 
  • Implementation treated as a repeatable product, not bespoke services 

This is where defensibility emerges. Deployment intelligence compounds with each client rollout. Each successful implementation strengthens the next, deepening institutional trust and operational resilience. AI amplifies this flywheel but cannot replace it. 

The Mispriced Risk of “AI-Only” Narratives 

In private credit, trust is earned slowly and lost quickly. It is built through consistent valuations, defensible reporting, and reliability during market dislocations. 

A system that produces faster answers, but weaker confidence does not displace incumbents. It increases operational and reputational risk. Investors should be wary of platforms that promise instant replacement without acknowledging institutional reality of fiduciary-grade infrastructure. 

The Investment Takeaway 

AI is not commoditizing ABF software solutions. It is widening the moat for platforms that integrate AI responsibly into governed systems. 

The next phase of growth for category leaders such as RiskSpan will be driven by combining deep domain knowledge with AI-native architecture. Leaders will treat the last mile – data integration, workflow configuration, and control design — as a core product capability, not an implementation afterthought. 

In markets where trillions in capital allocation depend on data integrity and institutional trust, the last mile isn’t an implementation detail. 

It’s the moat. 

Rates, Prepays and Consumer Stress: What the Data is Telling Us at the Start of 2026

 Register here for next month’s call: Thursday, February 19th, 2026, 1 p.m. ET. 

In the January Models & Markets call, our quantitative modeling team hosts their first monthly deep dive of the year into prepayment model performance, an updated analysis of second liens and HELOCs using Equifax data, and the evolving macroeconomic backdrop shaping mortgage markets. 

Here’s a quick recap in case you missed it. 

(Click here for the entire 20-minute recording or continue reading for a summary.)  

Revised HELOC and HEL Results Using Equifax ADS Data

  We performed a comprehensive analysis of second liens and HELOCs using Equifax’s Analytic Data Set (ADS), which represents a 10% anonymized sample of U.S. consumer credit data at the tradeline level. 

Following the resolution of data quality issues identified in an earlier analysis, the revised results now align much more closely with economic intuition. Prepayment speeds behave consistently across vintages, credit score bands, and refinancing regimes. 

One key takeaway holds that higher credit score borrowers tend to prepay faster, particularly during refinancing waves, while lower credit score segments remain slower. This pattern is especially evident in post-COVID vintages. Overall credit quality for HELOCs and second liens remains strong, with performance clustering closer to the highest credit score bands. 

Another notable observation is the role of seasonality in newer HELOC vintages. In a high-rate environment with limited refinancing activity, turnover-driven prepayments become more prominent. Baseline prepayment speeds for HELOCs are running around 15 CPR, higher than what is typically observed in first-lien portfolios under similar conditions. These dynamics provide useful signals for understanding how first-lien behavior may differ when second liens or HELOCs are present on the same property. 

  We plan to expand this analysis further, including deeper investigation into correlations between first- and second-lien prepayment behavior. 

Mortgage Rates Remain Likely to Stay Higher for Longer 

The broader economic outlook remains one of persistence rather than relief. Federal Reserve projections point to unemployment stabilizing around the low-4% range and real GDP growth near 2% over the medium term. Meanwhile, expectations for the fed funds rate suggest limited room for significant cuts beyond 2026. 

Longer-term rates tell a similar story. Consensus forecasts indicate the 10-year Treasury is unlikely to fall meaningfully below 4% over the next two to three years, implying mortgage rates are likely to remain near (and potentially above) the 6% level for much of the period ahead. Temporary dips tied to policy announcements or market events have proven short-lived, with rates quickly reverting back toward recent levels. 

Consumer Stress Continues to Build 

While headline spending remained strong during the most recent holiday season, the composition of that spending tells a more cautious story. Consumers increasingly favored lower-cost retailers, suggesting budget sensitivity and selective spending behavior. 

Survey data reinforces this theme. Year-over-year consumer sentiment and expectations have declined meaningfully, and perceptions of job insecurity (particularly among college-educated workers) have become more negative. These dynamics could have important implications for credit performance and housing activity as economic uncertainty persists. 

Prepayment Model Performance: v. 3.7 Continuing to Track Market Performance Well 

RiskSpan’s prepayment models continue to perform well across Agency collateral. 

RiskSpan’s Prepayment Model v3.7 continues to demonstrate strong performance across collateral types. Recent back-testing shows that model projections remain closely aligned with realized speeds, even as seasonal effects and calendar nuances influence month-to-month results. 

For conventional 30-year loans with lower coupons, December’s modest uptick in observed CPRs was largely attributable to four additional collection days relative to November. After adjusting for day count effects, actual prepayment speeds continue to trend lower, consistent with expectations in a higher-rate environment. 

Premium cohorts also remained largely stable. Despite a brief decline in mortgage rates late last year, the move was insufficient to trigger a meaningful new refinance wave. Most refinance-eligible borrowers have already acted, and the refinancing “pull-forward” effect appears largely exhausted. This dynamic is also visible in the S-curve, which has flattened back toward historical averages after October’s temporary acceleration. 

Agency collateral shows similar patterns. Ginnie Mae discount cohorts tracked model expectations closely, while premium cohorts remained flat. One area of ongoing refinement is deep in-the-money, very high-coupon Ginnie Mae loans, where actual speeds have run slightly slower than model projections as refinance incentives flatten out earlier than in prior cycles. 

Looking Ahead 

In summary: 

  • RiskSpan’s Prepayment Model v3.7 continues to perform well across most collateral segments 
  • HELOC and second-lien analysis using Equifax data now shows economically intuitive and stable results 
  • Mortgage rates are likely to remain near 6% in the absence of a major macro shock 
  • Consumer behavior is showing increasing signs of stress and caution 
  • RiskSpan plans to release additional analytics later this year, including a new non-QM credit model in the first half of the year and a next-generation prepayment model in the second half. 

We continue to add additional analytics reports on the Platform. Please visit www.riskspan.com/request-access to request free access. 

As always, please feel free to contact us to discuss or learn more. 

The Data Model That Powers Private ABF: Why Purpose-Built Architecture Changes Everything 

Private asset-backed securities don’t follow the same rules as public securitizations. The structures are more diverse. The triggers are more nuanced. The collateral is more diverse. And yet, most market participants still try to manage these instruments with tools designed for a different world entirely—or worse, with disconnected spreadsheets that multiply risk with every manual handoff. 

RiskSpan’s Private ABF platform was built specifically to solve this problem. Not as an adaptation of existing tools, but as a purpose-built data architecture designed from the ground up for the unique demands of private structured finance. The difference isn’t incremental—it’s foundational. 

Built for the Full Lifecycle 

RiskSpan’s Private ABF platform isn’t just a database—it’s a relational architecture where every entity maintains its identity and relationships across the entire deal lifecycle. The platform connects deals to their tranches, tranches to their collateral, collateral to individual loans, and all of these to the triggers, waterfalls, fees, and reserves that govern cash flow distribution. 

This matters because private ABS transactions aren’t static. Collateral performs. Triggers trip. Waterfalls redirect. Reserves build and release. A platform that can’t maintain these relationships in real time isn’t managing deals—it’s creating snapshots that are stale before they’re finished. 

Loan-Level Depth That Powers Real Analysis 

At the heart of any ABS transaction is the underlying collateral. RiskSpan’s Private ABF platform maintains comprehensive loan-level data with over 200 attributes per asset, tracking everything from origination characteristics through current performance status. This includes credit metrics like FICO scores and debt-to-income ratios, property and collateral details, payment history and delinquency tracking, modification and loss mitigation status, and ARM reset schedules and rate mechanics. 

The platform currently manages nearly half a billion loan records across active transactions. Each loan maintains its full history—not just current state, but the complete trajectory that informs forward-looking projections. When you run a cash flow model, you’re not working from aggregated pool statistics. You’re working from the actual loans, with their actual characteristics, generating projections that reflect real collateral behavior. 

Depth for Esoteric Structures 

RiskSpan’s Private ABF platform currently supports over several collateral types and structures. Each asset class has its own performance characteristics, prepayment behaviors, and loss dynamics. The platform’s architecture accommodates this diversity without forcing artificial standardization. 

But diversity in collateral is only part of the challenge. Private ABS triggers represent some of the most complex conditional logic in structured finance. Unlike standardized agency delinquency tests, private deal triggers can involve multi-step calculations with lookback periods, cure provisions that allow temporary breaches, step-rate adjustments that phase in over time, and early amortization events with distinct severity levels. 

The platform models triggers as executable logic, not static documentation. When a trigger breaches, the system knows what happens next—which waterfall priorities shift, which reserve requirements change, which reporting obligations activate. This is the deal’s immune system, and it needs to function in real time. 

Time-Series Without the Chaos 

Every entity in RiskSpan’s Private ABF platform exists in time. The platform maintains separate performance histories for tranches, collateral pools, fees, and reserves—each keyed by reporting date to enable point-in-time reconstruction of any deal state. 

This architecture solves one of the most persistent problems in private ABF operations: answering the question “what did we know, and when did we know it?” Whether for regulatory compliance, investor reporting, or internal risk management, the ability to reconstruct historical deal states isn’t a luxury—it’s a requirement that spreadsheet-based approaches simply cannot meet reliably. 

The Waterfall as Working Code 

Cash flow waterfalls in private ABF can run to dozens of steps with conditional branches, pro-rata splits, and priority reversions. RiskSpan’s Private ABF platform models these waterfalls as executable payment sequences—not flowcharts, but actual logic that routes cash from sources to destinations based on current deal state. 

Each waterfall step defines its priority in the payment sequence, its source of funds and destination, its allocation basis (pro-rata, sequential, or targeted), and the conditions under which it activates or suspends. 

When combined with the platform’s cash flow engine, these waterfall definitions become working models. You can project payments under any scenario, stress collateral performance, and see exactly how cash moves through the structure period by period. 

An Architecture Built for AI 

The same architectural principles that make RiskSpan’s Private ABF platform effective for traditional analytics make it exceptionally well-suited for artificial intelligence and machine learning applications. This isn’t a coincidence—it’s a consequence of building a data model that prioritizes structure, relationships, and semantic clarity. 

AI systems thrive on clean, well-organized data with explicit relationships. RiskSpan’s Private ABF platform delivers exactly this: normalized entities with consistent identifiers, clear hierarchies from deals down to individual loans, and temporal versioning that distinguishes current state from historical snapshots. When an AI model needs to understand a transaction, it doesn’t have to infer structure from unstructured sources—the relationships are already defined and traversable. 

The platform’s semantic richness enables natural language interfaces that actually work. Because every field has meaning within a consistent schema, AI can translate questions like “show me deals where the senior OC trigger is within 50 basis points of breach” into precise queries without ambiguity. The loan-level depth means AI models can identify patterns across hundreds of millions of records—finding correlations between origination characteristics and performance outcomes that would be invisible to traditional analysis. 

Time-series architecture is particularly critical for AI applications. Machine learning models for credit risk, prepayment prediction, and loss forecasting require historical sequences, not point-in-time snapshots. RiskSpan’s Private ABF platform maintains this temporal context natively, enabling training datasets that capture how loans and deals evolve over their lifecycles. 

RiskSpan is already deploying AI capabilities on this foundation: automated anomaly detection that flags unusual performance patterns, intelligent document extraction that populates deal records from offering documents, natural language querying that makes complex analytics accessible to non-technical users, and predictive models that leverage the full depth of loan-level history. The platform doesn’t just store data—it organizes knowledge in a form that AI can reason about. 

Designed for the Enterprise 

RiskSpan’s Private ABF platform operates as a multi-tenant platform with granular access controls. Every record carries client and user identifiers that enable sophisticated permission model—issuers see their deals, investors see their positions, servicers see their portfolios, and rating agencies see what they’re authorized to review. 

This isn’t just about security (though it is that). It’s about enabling collaboration across the structured finance ecosystem without compromising confidentiality. A single platform can serve all participants in a transaction, each with their appropriate view of the data. 

Built for What Comes Next 

Private ABS is an evolving market. New asset classes emerge. New structures get tested. New regulatory requirements arrive. RiskSpan’s Private ABF platform accommodates this evolution through flexible schema design that allows custom attributes without requiring database modifications. 

When a client brings a novel structure—say, a securitization of an asset class we haven’t seen before—the platform can ingest and model it without waiting for a software release. This extensibility is what allows a platform to stay current with market innovation rather than constantly playing catch-up. 

What This Means in Practice 

The architectural decisions in RiskSpan’s Private ABF platform translate directly to operational capabilities. For deal structuring, you can model waterfall variations and see their impact on tranche economics before going to market. For pricing, you can run scenarios against actual loan-level collateral, not simplified pool assumptions. For risk management, you can monitor trigger proximity and project breach timing under stress. For surveillance, you can track every metric that matters, with full audit trails and historical reconstruction. 

Private ABF deserves purpose-built infrastructure. RiskSpan’s Private ABF platform delivers it. 

A Day of Rest? Explaining November’s Spike in Non-QM Delinquencies

The just-released non-agency performance data (from November 2025) grabbed more than a few headlines.  

Non-QM loans saw a notable jump in early-stage delinquencies, raising understandable questions around the office (ours and others) about whether this move reflects emerging credit stress or something more benign – like, say, bad data. 

We ultimately concluded that the increase, while real, is likely temporary. The most plausible explanation for November’s spike in Non-QM delinquencies points to a calendar effect tied to the month ending on a Sunday. 

Benign, indeed. 

So what happened in November? 

In the latest data release from Cotality (formerly CoreLogic), delinquency rates rose meaningfully across the non-agency universe, driven almost entirely by a surge in the 30–59 days past due bucket. 

For the servicing month ending November: 

  • Total Non-Agency 30-day DQs increased 48 basis points, from 3.07% to 3.55% 
  • Non-QM 30-day DQs increased 41 basis points, from 1.66% to 2.07% 

For Non-QM loans, this one-month increase represents the largest jump in early-stage delinquencies since the COVID-related shock in April 2020, when these rates surged from 2.99% to 12.51%. For the broader non-agency universe, the increase was the largest since June 2024. 

These figures appear alarming. But a closer examination reveals that, in this case, the calendar may be doing most of the work. 

The Sunday Payment effect 

November ended on a Sunday (not just that, but on a Sunday that was, for many folks, the end of a four-day holiday weekend). When the final day of the month falls on a weekend, payments made on that day typically do not post until the following business day (in this instance, Monday, Dec. 1). As a result, loans that were paid “on time” (or less than 30 days late at least) can be temporarily classified as 30+ days delinquent for November reporting purposes, even though the borrower ultimately made the scheduled payment. 

This “Sunday month-end effect” is well documented and understood. And both internal discussions and external market commentary point to this being the primary driver of November’s delinquency spike. Among external commentators, ICE’s Andy Walden may have summarized it most succinctly: “While the topline delinquency numbers show a sharp increase, we’ve seen comparable spikes in prior years when November ended on a Sunday and scheduled payments didn’t post until early December.” 

The effect appears to be amplified with Sunday-ending Novembers in particular (perhaps because of the four-day weekend effect). As noted in the ICE piece, this has most recently happened in 2014, 2008, and 2003, when delinquency rates spiked by 61 bp, 112 bp, and 57 bp, respectively. All of those increases exceeded this year’s roughly 50 bp shift. 

Approach 1: A History Lesson

To test whether November’s increase fits a broader historical pattern, we examined the relationship between month-over-month delinquency changes and the day on which the month ended. 

Since 2006, there have been 33 months that ended on a Sunday. Over that nearly 20-year period, overall non-agency delinquency levels are broadly unchanged. And yet, those Sunday-ending months consistently exhibit upward pressure on reported 30-day DQ rates. 

Key observations: 

  • Across those 33 Sunday-ending months, non-agency 30-day DQ rates increased by an average of 37 bp 
  • 30-day DQ rates declined in only 4 of those months 
  • In the remaining 29 months, delinquency rates increased 
  • Importantly, 27 of those 29 increases were at least partially reversed in the subsequent month 

In other words, when months end on a Sunday, reported delinquencies tend to rise mechanically, only to then fall back once payments post and reporting normalizes. 

Chart 1: Month-over-Month Change in Non-Agency 30-day DQ rates (Sunday Month-Ends Highlighted, with green indicating a decline, and red indicating an increase) 

Approach 2: Agency Data as a Leading Indicator 

Non-agency delinquency data are reported with a one-month lag relative to Agency MBS. As a result, we can use Agency performance as a sort of real-time proxy for how non-Agency data may evolve in the following release. 

For the December factor date (corresponding to payments due November 30): 

  • Fannie/Freddie D30 jumped 19 bp, from 0.73% to 0.92% 
  • GNMA D30 jumped 41 bp, from 3.84% to 4.25% 

Crucially, both measures recovered sharply in December, declining back toward their October levels: 

  • Fannie/Freddie D30 fell to 0.78% 
  • GNMA D30 fell to 3.89% 

That represents a recovery of 74% and 88%, respectively, of the November spike. 

If Non-Agency and Non-QM delinquencies follow a similar pattern, a comparable recovery would imply: 

  • Non-QM D30 falling back to roughly 1.77% 
  • Total Non-Agency D30 falling back to roughly 3.20% 

These levels would be broadly consistent with pre-November trends and inconsistent with a narrative of accelerating credit stress. 

Chart 2: Agency vs. Non-Agency 30-day DQ Rate Changes and Subsequent Recovery (the dashed green and blue lines for December 2025 represent extrapolated D30 rates if Non-agency mortgages see similar recoveries to those experienced by Fannie/Freddie mortgages) 

Conclusion 

November’s spike in Non-QM delinquencies looks dramatic, but the weight of evidence points to a calendar artifact, not a structural shift in credit performance. Similar spikes usually occur when any month ends on a Sunday and are particularly pronounced when November does. History suggests 2025’s anomaly will be largely reversed in December. 

Investors should continue to monitor delinquency trends closely, and we will revisit this analysis when the next Cotality data are released in early February. For now, the data argue for caution, not alarm.

What a Year of Building AI in Structured Finance Actually Taught Us 

The lessons nobody puts in the demo. 

In 2025, our team built production AI systems that process billions of performance records for tens of millions of mortgages, develop cash flow models for complex private ABF structures directly from documents, and connect large language models directly to bond analytics APIs. 

We built dashboards, connectors, and credit analytics. Some of them worked. Some of them taught us more by failing. 

This is what we learned—not the polished conference talk version, but the notes we’d share with a peer team starting the same journey. 

The Value Shift Nobody Prepares You For 

A portfolio delinquency analysis that used to take three hours now takes twenty minutes. 

That sounds like a win. It is a win. But it also raises a question that’s harder to answer than any technical problem we solved this year: 

If AI handles in minutes what took us hours, what are we contributing? 

When we started pulling this thread, we realized that a significant portion of what felt like skilled analytical work was actually mechanical labor—data extraction, formatting, applying the same methodology we’d applied dozens of times before. The expertise was real, but it was wrapped in hours of execution that masked how much of the work was routine. 

Here’s where we landed: 

AI handles the “how.” Humans own the “why” and “so what.” 

The value now lives in knowing which questions matter. Understanding what the client really needs versus what they say they need. Recognizing when output is wrong because we understand the domain deeply enough to see the error. 

That’s an entirely different skill set. It requires judgment, contextual awareness, and domain intuition that deepens over years—the kind of expertise AI can’t simply replicate, unlike procedural analytical work.

Not everyone will make this transition comfortably. The analysts who built their identity around being fast and thorough at execution face a harder adjustment than those who always saw execution as a means to an end. 

We don’t have this all figured out yet. But we’ve stopped pretending the shift isn’t happening. 

Stop Asking AI to Write Code—Start Asking It to Think With You 

For years, we used Claude as a coding assistant. “Write a function that does X.” “Convert this data from format A to format B.” “Generate a script that calculates Y.” 

That works. But it captures maybe 20% of the value. 

The shift that changed our results was: treating Claude not as a tool to instruct, but as an analyst to think alongside. 

The difference looks like this: 

Before (instruction mode): 

“Write a Python script to calculate delinquency rates from this loan data.” 

After (thinking partner mode): 

“We need to identify hidden credit risk in this CLO portfolio—issuers that resemble recent defaults but haven’t shown price distress yet. What factors should we consider? What data would we need? Let’s build a scoring model together.” 

That second conversation led to identifying hidden exposure across issuers. Claude suggested factors we hadn’t considered—CLO concentration patterns, industry clustering effects, the relationship between coupon levels and distress signals. We debated the weighting. We refined the methodology. The output was genuinely collaborative. 

The code that emerged from the second approach was better, but that’s almost beside the point. The thinking was better. The model was better. The insight was better. 

This requires a different posture than most of us learned. You have to think out loud. Admit what you don’t know. Explain your reasoning and invite critique. Treat the AI as a colleague who happens to have read every document and doesn’t get tired—not as a sophisticated autocomplete. 

The developers and analysts on our team who made this shift produce substantively different work than those who are still in instruction mode. And the gap is widening. 

The First Version Will Be Wrong—Plan for It 

We built a benchmark analysis comparing a client’s NonQM loan portfolio against the broader market. The analysis looked solid: the portfolio showed a 1.37% delinquency rate advantage versus the universe. Strong results. Ready to present. 

Then someone asked about DSCR loans. 

In NonQM lending, DSCR (debt service coverage ratio) loans are a category unto themselves—with measurably better performance than other NonQM products. When we segmented the data, we discovered the universe was comprised of 43% DSCR loans while the client’s portfolio had only 30% DSCR loans. 

This changed everything. 

The client’s portfolio had less exposure to DSCR loans (the better-performing segment) yet still outperformed the benchmark. That alone was impressive, but our initial analysis understated the true picture. Once we compared performance within segments (DSCR vs. DSCR, non-DSCR vs. non-DSCR), the client’s edge was even larger than we’d initially observed. 

If we had presented the first version, we would have undersold our client’s own performance. The insight that mattered most—superior underwriting across both loan categories—would have been invisible. 

Lesson: “Wrong” doesn’t mean broken. It means the output doesn’t fully reflect reality. Have a domain expert review the work before drawing conclusions. 

Deploying AI Agents for End Users Is a Security Project 

Building an AI agent that works in a demo is straightforward. Deploying that agent in a production UI where real users interact with real data took us months. 

We built an agent that lets users query our bond analytics platform conversationally. The AI worked. Making it production-ready required solving problems: 

Prompt injection: When users can type anything into a text box processed by an LLM, you inherit a new attack surface. We implemented input validation, output filtering, tightly scoped permissions, and logging that captures every agent action for audit. 

Rate limiting: A single conversational turn might trigger 50 API calls. We built tiered limits—per-user, per-session, per-token—plus circuit breakers for runaway queries. 

Session management: Agent sessions need conversational context across multiple turns, isolated per user, with graceful expiration handling and automatic cleanup. 

Audit trails: Regulated industries need to know what the AI did. Every query, tool invocation, and response needs to be logged immutably. 

The agent itself was 20% of the effort. Authentication, authorization, input validation, rate limiting, session management, and security review were the other 80%. 

Lesson: In production, the agent is the easy part. The security wrapper is the product. 

Post script: AgentCore from AWS and Agent Framework from Microsoft are solving the deployment and security headaches.  

AI Is Good at Finding Information But Sometimes Overstates What It Means 

While building the credit risk analysis, we asked Claude to research distressed issuers—companies that had defaulted or were showing signs of stress. We wanted to understand patterns we could use to identify similar risks in the portfolio. 

Claude surfaced real-time signals we wouldn’t have found efficiently on our own: FTC antitrust actions, rating agency downgrades, refinancing walls, fraud allegations. Information that wouldn’t appear in pricing data for months was available in news coverage and regulatory filings. The research phase that would have taken days was completed in hours. 

But we also caught Claude drawing confident conclusions from weak sources. In one case, it attributed claims to “industry reports” that didn’t exist when we followed the links. The search results were real. The sources were ‘real’. But the synthesis drew conclusions the sources didn’t support. 

The lesson: use AI-powered search aggressively. It’s the difference between stale knowledge and current intelligence, especially in fast-moving situations. But verify specific claims. Click the links. Read the actual sources. 

AI is excellent at finding relevant information across large volumes of text. It is sometimes too confident about what that information means when synthesized. The combination of broad retrieval and skeptical verification is more powerful than either alone. 

Your Org Chart Isn’t Ready for This 

Our AI strategy deck included projections: reduction in onboarding costs, increased client capacity and margin expansion. 

The numbers were defensible. The business case was clear. 

What the projections didn’t address: the organizational implications of realizing the promised efficiencies. 

If analysts can serve five times more clients, do you need fewer analysts—or do you pursue five times more clients? If the answer is “more clients,” do you have the sales capacity? The support infrastructure? The management bandwidth? 

If developers now own adoption metrics for the features they build, then what happens to the product managers who previously owned that? Are product managers freed up for more strategic work, or are they defending territory? 

If AI drafts client communications, who reviews them? What error rate are we willing to accept? Who’s accountable when the AI gets something wrong? 

These aren’t hypothetical questions. We’re navigating them now, and the answers aren’t obvious. 

AI doesn’t just improve workflows. It reshapes roles. And most organizations—including ours—are making it up as they go. 

The companies that figure out the organizational design will outperform those that simply purchase better software. The differentiation in 2026 won’t come from adopting AI. It will come from redesigning teams, incentives, and accountability structures around what AI makes newly possible. 

What We’re Taking Into Next Year 

A year of building AI systems in structured finance clarified a few things: 

AI is more powerful than the hype suggests—once you integrate it into real workflows rather than treating it as a research toy. 

AI is more frustrating than the demos show—the gap between “works in claude.ai” and “works in production” is where most of the time goes. 

AI is more dependent on domain expertise than the automation narrative implies—it generates analyses quickly, but distinguishing plausible from accurate requires human judgement that compounds over years. The “why” and “so what” remain stubbornly human problems.  

AI changes more than technology—it changes job descriptions, team structures, and how people understand their own value. The skill isn’t operating the tool; it’s knowing when the output reflects reality. 

We don’t have all the answers. We’re still learning what this means for how we build software, how we serve clients, and how we organize ourselves. 

But we’re no longer wondering whether AI will change our industry. We’re focused on making sure we’re the ones defining how. 

Company

Products

Security & Compliance

Amazon Web Services
Microsoft

Linkedin

SOC2 Type II Certified

Get Started
Log in

Linkedin   

risktech2024