Our data typically reflects HUD's update within 1-2 days after their target. HUD occasionally delays past it.
HUD's Neighborhood Watch has every Compare Ratio, SDQ rate, and servicing portfolio in the country — buried behind a ColdFusion form and Excel downloads. We turn it into a clean REST API plus 41 task-shaped MCP tools you can plug straight into Claude. Including data HUD has but doesn't expose cleanly: the per-lender retail vs sponsored channel split.
Trailing 2-year window. Click any lender for the full profile.
| # | Lender | Originations | SDQ % | Compare Ratio | vs. peers |
|---|---|---|---|---|---|
| 1 | UNITED WHOLESALE MORTGAGE, LLC | 151,732 | 3.46% | 103 | In line with peers |
| 2 | ROCKET MORTGAGE, LLC | 120,896 | 2.73% | 81 | Better than peers |
| 3 | DHI MORTGAGE COMPANY LTD | 61,979 | 2.38% | 71 | Better than peers |
| 4 | CROSSCOUNTRY MORTGAGE, LLC | 49,212 | 3.22% | 96 | In line with peers |
| 5 | LOANDEPOT.COM | 44,781 | 4.31% | 128 | Worse than peers |
| 6 | GUILD MORTGAGE COMPANY LLC | 38,170 | 3.89% | 115 | In line with peers |
| 7 | LENNAR MORTGAGE, LLC | 37,096 | 1.66% | 49 | Top performer |
| 8 | NEW AMERICAN FUNDING, LLC | 32,129 | 4.40% | 131 | Worse than peers |
| 9 | FAIRWAY INDEPENDENT MORTGAGE CORPORATION | 30,346 | 3.64% | 108 | In line with peers |
| 10 | FREEDOM MORTGAGE CORPORATION | 29,732 | 2.46% | 73 | Better than peers |
Compare Ratio = lender's 24-month SDQ+claims rate ÷ peer area's rate × 100. 100 = at par; lower is better. 150puts a lender on HUD's Credit Watch; 200 + above the national rate can trigger Direct Endorsement termination in that HUD field office. Streamline refis are excluded from the public CR. Methodology →
Three question shapes that take five tools and an hour elsewhere — one tool call here.
The published Compare Ratio blends retail and sponsored TPO flow. We split them apart at national, state, HOC, county, and ZIP for every one of the 1,227 FHA-approved lenders. A national CR of 76 with a sponsored side at 130 is a different story than 76/76 — and the channel verdict line tells you which.
Try the lender lookup →One call returns the lender's QoQ + YoY CR delta, the nearest enforcement threshold they're drifting toward, and a one-word verdict (ok / monitor / escalate). Replaces five history lookups and the trend math LLMs get wrong on sparse quarterly series.
See the MCP tools →Given any sponsor lender ID, get: agent count, volume-weighted CR, HHI concentration, count of agents above 150, count on CWT or DE-terminated, named top-10 problem agents. Not exposed anywhere public — built directly on HUD's relationship table.
Try the relationship graph →Per-lender Total / Retail / Sponsored CR split — at national, state, HOC, or ZIP. HUD has it; we're the only public source that exposes it cleanly.
Full institution metadata, branch counts, DE underwriting status, agent / principal relationships, DBA crosswalk, DE-termination history.
Originations, seriously-delinquent count, claims, Compare Ratio. National, state, HOC, MSA, county, ZIP — both 2-year and 1-year cohorts.
County-level SDQ choropleth with click-to-drill ZIP overlay. Force-directed relationship graphs. Both backed by the API for embed.
Per-servicer SDQ + loss mitigation breakdowns for every active FHA servicer. HUD TRSII tier rankings (1-4) with Part B reimbursement.
Structured trend verdicts ("approaching 150"), QoQ CR movers, sponsor-book counterparty risk roll-ups (HHI + weighted-avg CR). 14 quarter-end snapshots.
Federal Register CWT / DE termination notices, parsed and linked to lender IDs. Resolves "affected" rosters to lookup-able lender_ids.
ZIP ↔ county ↔ MSA ↔ HUD field office. Pre-aggregated so the same data slices any way you want without a separate join layer.
41 task-shaped tools. "Compare CMG to PennyMac in Florida". "Is Freedom drifting toward CWT?" "What's Rocket's retail vs sponsor CR in Denver?"
Pricing is live and testers are onboarding. Request access to see the plans and get on the invite list.
View pricing & request access