Brief Refinance Shift Tests Mortgage Lenders’ Compliance Controls
Critical defect rate jumps 23.9% as math-based compliance findings expose the potential for one systemic error to affect loans across a lender’s book
Mortgage lenders did not need a full-scale refinance boom to expose weaknesses in their compliance controls.
The industry’s critical defect rate rose to 1.71% in the first quarter of 2026, up from 1.38% in the fourth quarter of 2025, according to the latest Mortgage QC Industry Trends Report from ACES Quality Management.
The 23.9% quarterly increase was the largest in four years. The rate was also 30.5% higher than the 1.31% recorded in the first quarter of 2025.
But the more consequential finding was where the defects occurred. Three of the four primary underwriting categories improved, while Legal/Regulatory/Compliance accounted for 26.02% of all critical defects, its largest share in more than five years.
That was up from 24.66% in the fourth quarter and approximately 15% one year earlier.
The increase points to a different kind of mortgage quality problem: not necessarily whether lenders made the correct underwriting decisions, but whether their systems consistently performed the calculations and procedural steps needed to produce compliant loans.
“Most of the testing behind the Legal/Regulatory/Compliance category is math-based,” Nick Volpe, executive vice president of ACES Quality Management, told NMP. “Systemic issues with the calculations or the way the data is inputted into the engine can lead to one systemic issue causing numerous individual findings across a lender’s book.”
That creates the potential for a single faulty calculation or data-entry process to affect multiple loans rather than remaining an isolated underwriting mistake.
A Refinance Test, Not A Refinance Boom
ACES attributed some of the compliance pressure to the brief refinance opening created when the average 30-year fixed mortgage rate fell below 6% in February.
Refinances represented 32.05% of loans reviewed during the quarter, up from 27.37% in the fourth quarter. It was the fourth consecutive quarterly increase and the highest refinance review share since the first quarter of 2022.
Refinance loans also produced 38.57% of critical defects, compared with 36.84% in the previous quarter.
ACES cannot determine exactly how many of the compliance findings occurred on refinance loans because its published benchmarking data does not cross-tabulate defect categories and loan purpose. Volpe said the connection is ACES’ interpretation of the broader trends rather than a directly measured causal finding.
“In our opinion, a mini-surge in refinances in Q1 negatively impacted Legal/Regulatory/Compliance defects,” Volpe said.
Refinances generally involve additional compliance considerations, including points-and-fees calculations, high-cost and higher-priced mortgage testing, rescission timing, and applicable state statutes. Government refinances can also carry net tangible benefit and recoupment requirements.
Compliance testing becomes particularly consequential because a calculation problem can be repeated across a lender’s production.
The quarter’s production shift was also more complicated than a simple increase in refinance volume. Industrywide refinance originations declined 7% from the fourth quarter but remained 24% above their year-earlier level, according to figures cited by ACES. Refinances gained market share largely because purchase originations fell 19% quarter over quarter to their lowest quarterly total in 12 years.
That means lenders experienced a more refinance-heavy production mix without enough additional overall volume to necessarily justify adding staff or compliance capacity.
Underwriting Improved While Compliance Slipped
The report found that the increase in critical defects did not originate in the four core underwriting categories.
Income/Employment defects declined from 21.52% to 20.07% of all critical defects. Assets fell from 15.25% to 10.41%, Credit edged down from 5.38% to 5.2%, and Liabilities remained almost unchanged at 10.78%.
Together, those four categories represented 46.46% of critical defects, their lowest combined share in the report’s 21-quarter comparison period.
Instead, findings became more concentrated in compliance, documentation, and property-related categories. Loan Documentation increased from 7.17% to 8.55% of critical defects, while Appraisal findings rose from 2.69% to 3.35%.
Even within underwriting, however, the nature of the remaining problems changed.
Calculation and analysis findings increased from 17.65% to 42.86% of all critical defects within the Assets category. That represented a 142.8% increase in their share of Assets findings, even though the Assets category improved overall.
Asset documentation findings, by comparison, fell from 73.53% to 53.57% of the category.
Income/Employment showed a similar, though less dramatic, movement. Calculation and analysis represented 45.28% of findings within that category, up from 39.58%, while documentation findings declined.
The results suggest lenders were having fewer problems collecting documentation but more difficulty correctly interpreting the information, including determining which funds qualified and how large or irregular deposits should be sourced and seasoned.
ACES Cannot Isolate The Riskiest Channels
Volpe said rate-and-term refinances accounted for a larger share of the refinance mix than cash-out loans in ACES’ first-quarter data. However, ACES cannot reliably identify whether defects were concentrated in particular refinance products, lender types, or origination channels.
“The data necessary to parse by products and channels is not uploaded by our clients consistently enough to draw any definitive conclusion across this data set,” Volpe said.
Smaller loans generally carry greater threshold risk under federal and state anti-predatory lending requirements because fixed costs represent a larger percentage of the loan balance. But Volpe said ACES could not make a definitive connection between average loan size and the first-quarter findings.
The Legal/Regulatory/Compliance category primarily covers costs associated with a mortgage, including rate spreads, points, and fees. Depending on the loan, those calculations can include financed single-premium credit insurance and fees covered by the Home Ownership and Equity Protection Act, Ability-to-Repay and Qualified Mortgage standards, and state laws.
The risk is not purely technical. ACES defines a critical defect as one that could make a mortgage ineligible for sale or uninsurable. Fannie Mae also identifies violations involving points-and-fees limitations as potential compliance defects and says its quality-control findings can result in repurchase demands, make-whole payments, or other remedies.
Refinance Quality Gap Narrows
Not every refinance result deteriorated.
Although refinances represented a disproportionate share of critical defects, the difference between their review share and defect share narrowed for a second consecutive quarter.
Refinance defect share exceeded review share by 6.52 percentage points in the first quarter, down from 9.47 percentage points in the fourth quarter and 18.3 percentage points in the third quarter of 2025.
That was the narrowest disparity since the fourth quarter of 2024, suggesting lenders were beginning to close the refinance quality gap.
The improvement has not been tested under a sustained refinance increase, however. The February rate window closed within weeks, and higher rates are expected to reduce refinance activity during the second and third quarters.
“This would have been an area we were interested in watching, but with the Iran war driving up interest rates toward the end of Q1 2026, we think the analysis will be short-lived,” Volpe said. “Refinances are likely to show a drop in Q2 and Q3.”
FHA Remains Disproportionately Defective
At the product level, VA loans delivered the strongest result.
VA loans accounted for 12.73% of reviewed files but 9.78% of critical defects. Their defect share declined 20.6% from the previous quarter.
ACES said streamlined Interest Rate Reduction Refinance Loans could have contributed to that improvement, but the data does not simultaneously segment loans by product and purpose. The report characterized that connection as a plausible explanation rather than a confirmed finding.
FHA loans continued to move in the opposite direction. They represented 24.46% of reviews but 32.27% of critical defects, making FHA the only major product category with a defect share materially exceeding its review share.
Conventional loans accounted for 61.31% of reviews and 57.21% of defects.
What The Data Does And Does Not Show
The ACES report is based on post-closing audits selected by lenders for full-file reviews and categorized using Fannie Mae’s loan defect taxonomy.
ACES does not disclose the precise number of loans or participating lenders because of anonymization provisions in its client contracts. Volpe said the quarterly results include “tens of thousands” of comparable records and that the first-quarter sample was larger than the fourth-quarter dataset and substantially larger than the sample from one year earlier.
ACES excludes specialty products such as Non-QM and construction loans, along with adjacent review categories involving early payment defaults, MERS, and Home Mortgage Disclosure Act compliance, to create a more comparable dataset of agency mortgages.
The composition of contributing lenders can also change as clients reach the 12-month seasoning threshold required for inclusion.
The 26.02% Legal/Regulatory/Compliance figure is likewise a share of all critical defects, not a standalone compliance defect rate.
Still, the direction is difficult to dismiss. Compliance has now been the largest defect category for two consecutive quarters, and its share increased sharply from the previous year.
The first-quarter refinance shift was small and brief. The fact that it coincided with this much compliance pressure raises a larger question for lenders: What happens when a sustained refinance market finally puts those systems under a real volume test?