FinanceRateCalc/fha-denial-rates-top-100

FHA Denial Rates by Lender — Top 100 Lenders, 2025

FHA mortgage denial rates for the 100 largest FHA lenders in the United States, computed from the complete 2025 CFPB Home Mortgage Disclosure Act (HMDA) public loan-level record.

Key facts: the 2025 spread across the 100 largest FHA lenders runs from 1.8% to 78.7% — a roughly 44x difference within a single federal program. The national FHA denial rate in 2025 was 22.1% (forward mortgages; reverse/HECM excluded — see correction note below).

Correction note (please quote the corrected figure)

An earlier version of this card stated a national rate of 21.7%, computed on a universe of 1,217,297 decisioned FHA applications that included 29,691 reverse-mortgage (HECM) records. HECM underwriting is not comparable to forward mortgages; excluding it yields the corrected universe of 1,187,606 decisioned forward FHA applications and a national denial rate of 22.1%. The correction was published at the source site and shifted all universe-level figures slightly. If you are quoting a national rate from this dataset, use 22.1%, not 21.7%.

Known limitation of the table below: it reflects the original top-100 snapshot and still contains two primarily-reverse lenders (Finance of America Reverse, Longbridge Financial). Their per-lender rows are accurate as reported, but they are not forward-mortgage lenders. The continuously corrected lender tables live at https://financeratecalc.com/fha-denial-rates-top-100.html.

Columns

• rank_by_denial_rate — 1 = highest denial rate
• lender_name — institution name as reported to HMDA
• lei — Legal Entity Identifier
• fha_applications_decisioned — FHA applications reaching a credit decision (action_taken 1, 2, or 3)
• fha_denials — applications denied (action_taken 3)
• fha_denial_rate_pct — denials ÷ decisioned applications × 100

Methodology

FHA = loan_type 2. Denominator = action_taken in {1, 2, 3} (originated + approved-not-accepted + denied). Denial = action_taken 3. Top 100 by FHA application volume. Full methodology: https://financeratecalc.com/methodology.html

Denial rates partly reflect applicant mix; a high rate is a historical observation, not evidence of wrongdoing, and this dataset is not a recommendation to use or avoid any lender. How much of the dispersion applicant mix can and cannot explain has now been measured — see the companion research below.

Companion research (variance decomposition)

How much of this lender-to-lender spread is the applicant pool, and how much is the lender itself? A nested-model decomposition on 859,090 decisioned 2025 FHA applications finds that lender identity accounts for 37.97% of all explained variation in denial outcomes ("the door effect"):

• Paper: Yetiş, Z. (2026). The Door Effect: Lender Identity and Unexplained Variation in FHA Mortgage Denial Outcomes. SSRN 7309319. https://ssrn.com/abstract=7309319 · doi:10.2139/ssrn.7309319
• Results dataset: https://huggingface.co/datasets/FinanceRateCalc/door-effect-fha-2025
• Plain-English summary: https://financeratecalc.com/door-effect.html

Source & citation

Underlying data: CFPB HMDA 2025 public Snapshot. Processing: FinanceRateCalc (independent; no lender partnerships, no advertising).

Yetiş, Z. (2026). FHA Denial Rates by Lender — Top 100, 2025. FinanceRateCalc. https://financeratecalc.com/fha-denial-rates-top-100.html — companion working papers: SSRN 7156938 (AI benchmark), SSRN 7309319 (door effect).

License: CC BY 4.0. A denial is a data point, not a verdict on the applicant.