No national figure exists for debt in newly FEMA-remapped flood zones — but GSE-level flood exposure ($2.2 trillion) is the closest measured proxy for concentration

Summary

No source -- not this corpus, not FEMA's own data, not academic or Federal Reserve research -- publishes a national dollar total for mortgage debt in properties FEMA has newly mapped into a flood zone. That specific figure requires matching FEMA's parcel-level map-revision geometry (before/after zone status) to loan-level unpaid balances and servicer identity, a linkage that exists only inside proprietary flood-determination vendor products (Cotality/CoreLogic, First Street Foundation) sold to lenders and servicers, never published as an aggregate.

On concentration, the best available evidence points to an answer, just not for the exact population asked about: Fannie Mae and Freddie Mac alone hold $8.17 trillion in outstanding single-family mortgage balances (as of Jan. 2023), and Federal Reserve Bank of Richmond research estimates $2.22 trillion (27.2%) of that sits in ZIP codes with meaningful flood risk. Whatever slice of that is in newly-remapped properties specifically is concentrated, by construction, at essentially two entities -- because that is where the majority of U.S. conforming mortgage debt already sits. Separate research on a related-but-different population (properties with real flood risk FEMA has never mapped, rather than newly remapped) finds large banks retreat from lending there while nonbanks and small local banks keep lending and pass more of that risk along via sale or securitization.

What the AskAmerica corpus actually has

Checked live (2026-09-26) via search_catalog, list_tables, and describe_table against the disasters and housing schemas:

• disasters.nfip_policies / disasters.nfip_claims carry a flood_zone column (current zone at time of policy or claim), but no map-revision date, no 'previously mapped as' field, and no link to a mortgage or lender.
• housing.hmda_lar, housing.hmda_loans, housing.hmda_lending_by_tract carry loan amounts, lenders, and census tracts, but no flood-zone designation at all, mapped or not.
• No table in any schema carries FEMA Letter of Map Revision (LOMR) or National Flood Hazard Layer (NFHL) map-change records -- the catalog search for these exact terms returned zero matches.

The corpus has flood-zone status and it has mortgage-loan data, but nothing that joins the two at the property level, and nothing that captures a map-revision event (before/after status over time) at all. This is a genuine sourcing gap, not a query that could have been written differently.

The closest real evidence found, and exactly how it differs from the question

• Federal Reserve Bank of Richmond, "Assessing GSE Mortgage Portfolios' Exposure to Past and Future Flood Risk" (Phan, July 2024) -- matches 100+ million Fannie Mae/Freddie Mac loans (1999-2023) to First Street Foundation's flood-factor projections at the 3-digit ZIP level. Of $8.17 trillion in GSE loans outstanding in January 2023, an estimated $2.22 trillion (27.2%) sits in ZIP codes with an average flood factor of 2 or higher ("at risk"), and $0.41 trillion (5.0%) in ZIP codes with a flood factor of 3+ ("higher risk"). This is the single most directly comparable figure to the question's dollar amount -- but it measures forward-looking flood-risk exposure broadly, not specifically debt in properties a FEMA map revision has newly moved into a mapped zone, and it uses 3-digit ZIP (not parcel-level) matching, which the author himself flags likely understates true exposure. A companion finding: Hurricane Irma raised default frequency by 0.40 percentage points in affected Florida ZIP codes over the following 18 months, evidence the exposure has real financial consequence.
• CFPB, "Flood Risk and the U.S. Mortgage Market" (Jan. 2025) -- matches confidential, property-level HMDA records to FEMA Special Flood Hazard Area status and First Street's Flood Factor for 2012-2021 (South Atlantic and South-Central regions). It measures mortgage-application volume and borrower characteristics by flood-risk tier from a single static 2021 FEMA snapshot -- not map-revision events over time, and not outstanding balances.
• NY Fed, "Flood Risk Outside Flood Zones" (Blickle, Perry, Santos, 2024) and its underlying staff report -- the best available lender-behavior evidence, using property-level HMDA matched to CoreLogic flood scores. Studies properties with real flood risk that are never covered by a FEMA map at all -- the inverse population from the question, which asks about properties actively moved into a mapped zone by a revision. Its lender-concentration finding: large banks cut lending roughly 1 percentage point in these risky-but-unmapped areas; nonbanks and small local banks do not pull back and are 1-2+ percentage points more likely to securitize or sell the loans they do originate. In dollar terms, $200 billion in mortgages were originated in 2021 in already-mapped areas versus over $600 billion in high-risk areas with no map coverage at all.
• Cotality (formerly CoreLogic), industry brief on FEMA's June 2026 map release -- the only source found that discusses the newly-mapped population specifically, but as process and scale, not an aggregate dollar figure: 100+ communities across 12 states received new maps (the largest single-day release in 5 years), with roughly 75% of Honolulu County newly mapped. Confirms that under Fannie Mae/Freddie Mac servicing rules, a newly-mapped property must obtain flood insurance within 120 days of the revision or the servicer force-places coverage and adds the cost to the loan -- but this is a vendor marketing piece, not primary data, and reports no dollar total.
• Retraction flagged, not used as evidence: a widely-cited NBER/Review of Financial Studies paper by Ouazad and Kahn ("Mortgage Finance and Climate Change: Securitization Dynamics in the Aftermath of Natural Disasters") argued lenders systematically sell disaster-exposed mortgages to the GSEs after storms. That paper was formally retracted in 2025 after a journal-appointed committee could not reproduce its central results. It is not used as support for the lender-concentration claim in this report.

Bottom line

The dollar amount specifically for newly-remapped properties has no published answer anywhere. The concentration question has a defensible partial answer through a proxy: because Fannie Mae and Freddie Mac alone hold $8.17 trillion of U.S. single-family mortgage debt and an estimated $2.22 trillion of that sits in flood-risk ZIP codes, any newly-remapped subset of that debt is, by construction, heavily concentrated at those two GSEs before even considering which originating bank or nonbank first made the loan. Layered on top, the NY Fed's finding on the (different, unmapped-risk) population -- that large banks retreat while nonbanks and small local banks retain and offload more flood-exposed risk -- is the best available directional evidence for how newly-remapped-property lending is likely to behave, but it was not measured on that exact population and should not be quoted as if it were.

What This Report Does Not Answer

• Dollar amount of outstanding mortgage debt in properties newly mapped into a FEMA flood zone: No dataset in this corpus, FEMA's own OpenFEMA API, HMDA, or the academic/Fed literature links (a) FEMA flood-map revision effective dates/parcels to (b) loan-level unpaid principal balance. This requires proprietary parcel-to-loan matching (the kind Cotality/CoreLogic and First Street Foundation sell commercially to lenders/servicers) not published as a national aggregate anywhere found. Reported the closest adjacent published figures instead, explicitly labeled as not the same measure -- including the one genuinely close figure found (GSE portfolios' flood exposure, Richmond Fed) which measures ALL flood-risk exposure, not specifically newly-remapped exposure.