FHA lender-risk outliers: reading Neighborhood Watch compare ratios
HUD's Neighborhood Watch compare ratio benchmarks each FHA lender's early-default and claim experience against its geographic peers — here's how to read the outliers without over-reading them.
HUD's Neighborhood Watch Early Warning System is the public face of FHA lender monitoring. Its central metric — the compare ratio — is one of the few risk benchmarks in U.S. mortgage that is both public and peer-adjusted. For counterparty, compliance, and FHA production teams, it's the first-pass screen for which originators are running hot relative to the book they actually wrote. This piece walks the national originator ("orig") slice and shows how to read the outliers without over-reading them.
What the compare ratio actually measures
The compare ratio indexes a lender's rate of seriously delinquent loans plus insurance claims (SDQ + claims) against the same rate for all other FHA lenders originating in the same geography over the same two-year window, with 100 set to the peer average. The geographic adjustment is the point: a lender concentrated in a high-default metro isn't penalized for the metro, only for beating or trailing its neighbors there. A ratio of 200 means roughly twice the peer early-default-and-claims experience; 150 means about half again as much.
Across the national originator slice used here — roughly 1,612,355 FHA originations across 905 originating lenders, a two-year window — the baseline SDQ + claims rate is 3.42%. That 3.42% is the "100" against which every lender below is indexed.
The outliers
Restricting to lenders with 2,000 or more FHA originations (enough volume that the ratio isn't noise), the top of the distribution:
| Lender | Compare ratio | SDQ + claims | FHA loans |
|---|---|---|---|
| Top Flite Financial Inc | 217 | 7.43% | 3,595 |
| United Wholesale Mortgage (UWM) | 191 | 6.53% | 2,020 |
| Geneva Financial | 165 | 5.64% | 3,671 |
| Cardinal Financial | 164 | 5.61% | 7,848 |
| Leader One Financial | 155 | 5.30% | 3,058 |
| SecurityNational Mortgage | 154 | 5.27% | 4,727 |
| Ixonia Bank | 149 | 5.08% | 4,256 |
| Ruoff Mortgage | 148 | 5.06% | 4,882 |
| loanDepot.com | 146 | 5.01% | 30,436 |
| Lower, LLC | 135 | 4.60% | 7,181 |
Top Flite Financial tops the list at 217 — a 7.43% SDQ-and-claims rate against the 3.42% national baseline, or roughly twice the peer early-default experience on 3,595 loans. That's the zone HUD watches most closely.
Two names deserve separate attention because of scale. UWM, the largest wholesale lender in the country, posts a 191 on a smaller FHA slice (2,020 loans) — a high ratio on modest FHA volume, worth watching but thinner. loanDepot is the more striking read: a 146 compare ratio on a 30,436-loan book. A ratio that far above peers, sustained across a book that large, is not a small-sample artifact — and because the metric already controls for geography, it isn't explained by where the loans were written either.
What HUD does with it
The compare ratio is the operational backbone of FHA's lender-monitoring program, not just a disclosure. HUD's Quality Assurance Division and the Lender Insight reporting built on Neighborhood Watch use it to triage the servicing and origination population. A ratio sustained above 200 is the conventional line that draws heightened scrutiny — expanded post-endorsement review, Credit Watch termination of a lender's origination approval in an area, or referral to the Mortgagee Review Board for sanctions. Investors and warehouse counterparties read the same public numbers when they set FHA overlays and pricing.
Caveats — this is a signal, not a verdict
A high compare ratio flags where to look, not what you'll find:
- Borrower/product mix. Down-payment-assistance concentration, higher-DTI approvals, and first-time-buyer segments lift early default for reasons the geographic adjustment doesn't capture — it normalizes place, not credit box.
- Servicing and transfers. SDQ is a servicing-observed outcome; loss-mit execution and mid-life transfers move it without touching how the loan was underwritten.
- Vintage. A two-year window is dominated by recent loans seasoning into peak early-default; fast growth into a soft vintage looks worse.
- What it isn't. Early default plus claims — not fraud, not a defect finding, not a legal conclusion. These are statistical outliers, nothing more.
The disciplined read pairs the ratio with volume (is the sample large enough to trust?) and trend (is it rising, stable, or a single bad vintage rolling off?). On those two axes, Top Flite's 217 and loanDepot's 146-on-30k-loans are the entries that most warrant a second look — the first for magnitude, the second for the size of the book behind it.