REVERSE MORTGAGE · HUD FHA HECM SNAPSHOT
FREE · PUBLICHECM reverse-mortgage origination
Who endorses FHA-insured reverse mortgages (Home Equity Conversion Mortgages), how the monthly endorsement count trends, and how each originating lender's book splits across traditional / HECM-for-Purchase / HECM-to-HECM refinance, retail vs. sponsored (wholesale/TPO), and fixed vs. adjustable. Sort the board by any metric. Built from HUD's public FHA HECM Snapshot. Dec 2021 – May 2026.
HECMs endorsed
157,949
53 months
HECM originating lenders
4,756
ranked below
HECM average note rate
5.63%
reported loans
HECM average initial principal limit
$248.7K
at origination
HECM-for-purchase share
5.0%
42.2% wholesale
Monthly HECM endorsements
FHA-insured reverse-mortgage endorsements per month · hover for the average note rate
Break in the series: Oct 2025 is absent at source (HUD publication pause) — shown as a gap, never a fabricated zero.
Traditional HECM
72.9%
115,174 equity-draw loans
HECM-for-Purchase
5.0%
7,932 H4P loans
HECM refinance
22.1%
34,843 HECM-to-HECM
Adjustable-rate
98.4%
155,380 ARM · rest fixed
Originating-lender rankings
4,756 lenders · sort any column · aggregated across the loaded window
21 of 4756 lenders open a dossier — those whose FHA mortgagee id also appears in the FHA servicer file. The rest are HUD-published originators this estate holds no entity page for, so their rows are not links to a dead end.
HUD PUBLICReverse-side complement to the FHA forward book — FHA servicing rankings cover the forward (traditional) side.
Source: HUD FHA HECM Snapshot (Office of Housing / Office of Risk Management), the public, no-login monthly loan-level reverse-mortgage endorsement disclosure. Figures aggregate the per-originating-lender monthly rollup across the loaded window (Dec 2021 – May 2026, 53 months present). “Endorsed” = HECMs FHA insured that month; H4P = HECM-for-Purchase; wholesale = sponsored (TPO) originations. Averages cover reported loans only; a lender with no reported rate/limit renders “—”, never zero. Coverage gap: Oct 2025 is absent at source (HUD publication pause) — shown as a break in the trend, never back-filled with a fabricated zero. Public HUD data, not endorsed by HUD or FHA. See data attribution.
What the loans became: the HMBS book
The rankings above are HUD endorsements — who WROTE the loan. This is the securities side: how fast Ginnie Mae HMBS pools pay down, how much borrowers draw on their lines of credit, and how much of the book is approaching the 98%-of-maximum-claim-amount threshold that forces the issuer to buy the loan out of the pool. That buyout, not a refinance, is the dominant HMBS termination and has no forward-mortgage analogue.
HMBS pool prepayment speed
14.3%
-0.53 pp vs prior month
HMBS mandatory-purchase speed
15.9%
-1.58 pp vs prior month
HECM line-of-credit draw rate
5.5%
-1.64 pp vs prior month
HMBS balance to maximum claim amount
65.0%
buyout forced at 98%
HMBS 12-month buyout pipeline
$2.1B
projected, not forecast
Approaching the 98% buyout trigger
Loans by proximity to the maximum claim amount — the bands NEST, they do not partition
A loan at 98% is also counted at 95% and 90% — read the bars as thresholds, not buckets. The twelve-month row projects each loan's own contractual accretion (note rate + annual MIP) forward with draws held at zero; a further draw pulls the date forward and a loan can terminate first for any other reason. It is a pipeline, not an expectation.
Borrower draws on the line of credit
56,730 loans with an open line · measured against contractual growth
Line at month start
$3.9B
undrawn commitment
Grown, if untouched
$3.9B
at note rate + annual MIP
Line at month end
$3.9B
as disclosed
Drawn this month
$18.1M
0.5% of the opening line
A HECM line of credit GROWS every month at the note rate plus the annual MIP whether or not the borrower touches it, so the drawdown is not a balance difference — the line usually ends the month LARGER. The draw is contractual growth that did not materialise, floored at zero per loan. 1,872 loans grew faster than the identity predicts (intra-month rate/MIP timing); they are counted here rather than netted against a real draw somewhere else.
Pool speed against the buyout speed
UPB-weighted HMBS pool prepayment speed vs. the 98%-of-MCA mandatory-purchase speed, by disclosure month
The two lines are on DIFFERENT UNIVERSES and the gap between them is not a spread. Pool speed covers the whole HMBS book (15,109 pools, $56.2B); the buyout speed is loan-level and covers only the fixed / annual-adjustable run-off book (74,831 loans, $16.7B). Both are annualised from a single month. The series starts at Jul 2024: one earlier month has no prior month to measure a buyout speed against and is left out rather than drawn as zero.
HMBS issuers
10 issuers reporting in May 2026 · sort or filter any column
Source: Ginnie Mae MBS Disclosure Data — the public HMBS pool disclosure (whole book: 15,109 pools, $56.2B) and the HMBS loan-level disclosure. Coverage scope: the loan-level file held is the FIXED and ANNUAL-ADJUSTABLE split — the seasoned run-off book (74,831 loans, $16.7B across 10 issuers). The monthly-adjustable split, where essentially all post-2022 HECM production sits, is not held, so every draw rate, buyout figure and termination speed on this panel is scoped to the run-off book and is not the whole HMBS market. Pool and loan measures are never mixed: a pool holds one month of participations — a slice of each loan — so the pool file's own balance-to-MCA percentage is not buyout proximity for anything, and the loan-to-pool cross product is not published. Speeds are annualised from a single disclosure month. Public U.S. Government work, not endorsed by Ginnie Mae. See data attribution.