FOR IMMEDIATE RELEASE
Washington, D.C. — December 8, 2025 – Polygon Research has released the first transparent, reproducible Qualified Mortgage (QM) flag built entirely from public HMDA loan-level data, covering every originated loan from 2018 through 2024.
The new QM / Non-QM / Out-of-Scope flag is now live in HMDAVision®, Polygon Research’s flagship market-intelligence tool in Polygon Vision. It is accompanied by a technical white paper detailing the full methodology.
Since the Ability-to-Repay / Qualified Mortgage (ATR/QM) rule took effect, the mortgage industry has debated "QM vs. Non-QM" trends without a consistent view of the total market. Until now, estimates of the Non-QM segment relied on RMBS deal disclosures, proprietary PPE/lock data, partial deed or servicing datasets. While useful, these sources are often incomplete, selection-biased, and difficult to compare across lenders or geographies. HMDA offers a loan-level view of all reported originations, but historically lacked a standardized, Reg Z–aligned method to distinguish QM from Non-QM. Polygon Research has bridged that gap.
Polygon Research has implemented a rules-based QM flag for the entire HMDA-reported universe (2018–2024), utilizing only publicly available fields. Each loan is classified as QM, Non-QM, or Out of Scope (loans outside the ATR/QM framework).
The methodology is strictly aligned with 12 C.F.R. § 1026.43, factoring in:
Where HMDA data is ambiguous (e.g., lack of documentation type), the methodology applies a conservative "Out of Scope" label rather than forcing a classification. This ensures the "Non-QM" cohort represents a high-confidence signal rather than a best guess.
"If the market is to have a serious conversation about Non-QM, we need a shared view of where those loans actually are," said Val Buresch, CMB, Founder & CEO of Polygon Research. "We built this framework to be inspected, not hidden. By aligning our flags with the ATR/QM structure and documenting every step, we are giving lenders, investors, and policymakers a definitive starting point for analysis."
With the new flag, users can move beyond high-level averages to size the Non-QM market and track market share by year, lender, channel, and geography with precision. Other applications include:
The Non-QM Flag is available immediately to HMDAVision subscribers. The accompanying white paper, which details the regulatory foundations and logic, is available for download at https://insights.polygonresearch.com/Non-QM-Methodology
Polygon Research is a mortgage data science company based in Washington, D.C. We unlock microdata from over 20 public datasets—including HMDA, Census ACS, RMBS/MBS, and Call Reports—transforming them into HMDAVision®, MBS Pivot, and other interactive market-intelligence tools. We serve lenders, credit unions, investors, and policymakers who require deep, loan-level analytics without the overhead of building proprietary data warehouses.
Learn more at https://www.polygonresearch.com/
Lyubomira "Val" Buresch
Founder & CEO
Polygon Research
14 Ridge Square, NW
Washington, DC 20016
Phone: 202.844-2720
Email: lburesch@polygonresearch.com
Polygon Research introduces an industry-first QM/Non-QM flag on every 2018–2024 HMDA loan, enabling true loan-level Non-QM market sizing. Download the white paper.
FOR IMMEDIATE RELEASE
Washington, D.C. — December 8, 2025 – Polygon Research has released the first transparent, reproducible Qualified Mortgage (QM) flag built entirely from public HMDA loan-level data, covering every originated loan from 2018 through 2024.
The new QM / Non-QM / Out-of-Scope flag is now live in HMDAVision®, Polygon Research’s flagship market-intelligence tool in Polygon Vision. It is accompanied by a technical white paper detailing the full methodology.
Since the Ability-to-Repay / Qualified Mortgage (ATR/QM) rule took effect, the mortgage industry has debated "QM vs. Non-QM" trends without a consistent view of the total market. Until now, estimates of the Non-QM segment relied on RMBS deal disclosures, proprietary PPE/lock data, partial deed or servicing datasets. While useful, these sources are often incomplete, selection-biased, and difficult to compare across lenders or geographies. HMDA offers a loan-level view of all reported originations, but historically lacked a standardized, Reg Z–aligned method to distinguish QM from Non-QM. Polygon Research has bridged that gap.
Polygon Research has implemented a rules-based QM flag for the entire HMDA-reported universe (2018–2024), utilizing only publicly available fields. Each loan is classified as QM, Non-QM, or Out of Scope (loans outside the ATR/QM framework).
The methodology is strictly aligned with 12 C.F.R. § 1026.43, factoring in:
Where HMDA data is ambiguous (e.g., lack of documentation type), the methodology applies a conservative "Out of Scope" label rather than forcing a classification. This ensures the "Non-QM" cohort represents a high-confidence signal rather than a best guess.
"If the market is to have a serious conversation about Non-QM, we need a shared view of where those loans actually are," said Val Buresch, CMB, Founder & CEO of Polygon Research. "We built this framework to be inspected, not hidden. By aligning our flags with the ATR/QM structure and documenting every step, we are giving lenders, investors, and policymakers a definitive starting point for analysis."
With the new flag, users can move beyond high-level averages to size the Non-QM market and track market share by year, lender, channel, and geography with precision. Other applications include:
The Non-QM Flag is available immediately to HMDAVision subscribers. The accompanying white paper, which details the regulatory foundations and logic, is available for download at https://insights.polygonresearch.com/Non-QM-Methodology
Polygon Research is a mortgage data science company based in Washington, D.C. We unlock microdata from over 20 public datasets—including HMDA, Census ACS, RMBS/MBS, and Call Reports—transforming them into HMDAVision®, MBS Pivot, and other interactive market-intelligence tools. We serve lenders, credit unions, investors, and policymakers who require deep, loan-level analytics without the overhead of building proprietary data warehouses.
Learn more at https://www.polygonresearch.com/
Lyubomira "Val" Buresch
Founder & CEO
Polygon Research
14 Ridge Square, NW
Washington, DC 20016
Phone: 202.844-2720
Email: lburesch@polygonresearch.com