Credit impact can extend beyond the FEMA flood line

Flood risk is not bounded by the FEMA 1% (1-in-100-year) floodplain (see Exhibit 3). Observed flood depths frequently extend beyond the mapped FEMA boundary because of precipitation-driven flooding, complex local drainage dynamics and interactions between riverine and surface-water processes that are not fully represented in regulatory maps. As a result, a meaningful portion of exposure and loss can occur well outside mapped FEMA flood zones. 

FEMA flood maps are designed to delineate areas used to determine where federally regulated or backed mortgage lenders require flood insurance, using a binary, threshold-based approach that classifies locations as inside or outside a designated flood zone. By contrast, the Moody’s modeling framework represents flood hazard as a continuous, depth-based phenomenon across the entire landscape. Instead of relying on a single cutoff, it models how flood depths and associated damage vary gradually from one location to the next, capturing variation in flood risk both within and beyond FEMA-mapped zones.

 

Exhibit 3: Flood risk exists beyond FEMA-mapped zones

Estimating a major event: the 1-in-100-year flood scenario

Analyzing different time periods as well as a future climate scenario shows growing risk exposure for counties as the flood footprint expands.
 

In a modeled 1-in-100-year flood scenario, total uninsured residential potential loss exposure could exceed $375 billion nationwide, with a protection gap of more than 65% (see Exhibit 4). This scenario does not represent a single flood event occurring simultaneously across all counties. Rather, it reflects an aggregate, probabilistic view of risk in which each location has a 1% annual chance of experiencing losses that exceed modeled thresholds, with extreme impacts that can occur in different places or years. The analysis incorporates fluvial, pluvial and storm-surge flooding, and indicates loss potential can extend beyond the FEMA 1% floodplain.
 

In such a scenario, less than 2% of counties, across 11 states, carry 65% of uninsured loss exposure nationwide, with potential losses above $1 billion each. Counties with potential losses above $5 billion are clustered in the four coastal and Gulf states of Florida, Louisiana, South Carolina and Texas, with protection gaps in the 45%-75% range. All of these states have moderate-to-high credit exposure to physical climate risk.
 

Although large protection gaps in percentage terms can help signal pockets of high risk exposure, it is ultimately the magnitude of such uninsured losses and ability to absorb such shocks through federal disaster aid, state and local resources, liquidity and revenue, and insured loss proceeds, and the strength of governance frameworks, that will determine credit impact.
 

Exhibit 4: Potential uninsured residential loss exposure aggregated across counties exceeds $375 billion under a 1-in-100-year flood

Counties with highest potential uninsured losses are concentrated in four states with moderate-to-high credit exposure to physical climate risk

Estimating an extreme event, the 1-in-500-year flood scenario

A more extreme 1-in-500-year flood could triple total uninsured risk exposure to more than $1 trillion with the nationwide protection gap surpassing 70%, assuming static insurance levels, compared to the 1-in-100-year flood scenario (see Exhibit 5). Counties with potential uninsured losses of more than $5 billion would extend beyond the Gulf and Atlantic coasts to 11 states, some of them inland, and protection gaps widen to 50%-100% under this scenario.
 

As rivers exceed their primary channels, secondary floodplains activate, and pluvial flooding spreads into neighborhoods that rarely flood under moderate conditions. As a result, the flooding footprint expands to states with low physical climate risk exposure, such as Pennsylvania and Illinois.
 

As the magnitude of losses increases in this more extreme scenario, the share of counties with losses above $1 billion rises to 5%, across 25 states, with their total loss exposure rising to 80%. Meanwhile, the proportion of counties likely to incur relatively modest losses of $150 million or less declines to 75%.

Exhibit 5: The magnitude of potential uninsured loss by county in a 1-in-500-year flood triples compared to a 1-in-100-year flood 
 

Counties with highest potential uninsured losses extend beyond the Gulf and Atlantic coasts to 11 states, including states with low credit exposure to physical climate risk

Case study: Hurricane Helene’s catastrophic impact in Asheville, North Carolina

Rainfall during Hurricane Helene, a Category 4 storm that made landfall in September 2024, exceeded the 1-in-1,000-year rainfall return period, placing it in the extreme tail of historical expectations (see Exhibit 6). This divergence highlights the risks and limitations of relying solely on backward-looking statistics to characterize flood risk in a changing hydroclimate, particularly for short‑duration, high‑intensity rainfall events.

 

Exhibit 6: Observed rainfall intensity of Hurricane Helene was well beyond historical expectations

As flood hazard intensifies, total losses rise sharply, while insurance coverage often increases only marginally. This dynamic leaves communities exposed to a persistently high share of uninsured losses across a wide range of event severities.
 

The City of Asheville and Buncombe County, North Carolina, experienced some of the most severe flooding during Hurricane Helene. The event demonstrates how extreme precipitation and runoff can significantly exceed historical levels.
 

In March 2025, Moody’s Ratings revised the outlook on the county’s Aaa rating to negative, reflecting expected decreases in fund balances, partly related to Hurricane Helene. After the hurricane, county management reduced some spending to offset the revenue shortfall. Initial cleanup efforts were primarily funded by federal dollars but as the county was able to contract with the Army Corps of Engineers, which is paid directly by the federal government, it managed to return most of the clean-up funds received. On 15 April 2026, Moody’s Ratings revised the county’s outlook back to stable and affirmed the Aaa rating, reflecting recently balanced general fund operations in line with county policy targets – despite challenges related to Hurricane Helene – and an expectation that the county’s financial position will remain stable, among other factors.
 

Despite elevated flood risk across multiple modeled scenarios, insurance penetration in the county remains low, resulting in most losses being uninsured. Moody’s RMS model results indicate a flood insurance protection gap of around 88% in Buncombe County across a range of severities, from 1-in-100 to 1-in-250 and 1-in-500 loss events.
 

Intensifying risk and potential losses in a future scenario

Under an intermediate-emissions scenario, uninsured flood losses could increase on average by about 25% by 2050 to around $472 billion compared to the 1-in-100-year flood scenario (see Exhibit 7), with a similar protection gap of about 65%.
 

As the flood footprint expands in this scenario, counties with the largest potential uninsured losses of more than $5 billion include one additional state, New Jersey, as well as Florida, Louisiana, South Carolina and Texas – all with moderate credit exposure to physical climate risk. Insurance protection gaps would increase to 45%-90%.
 

The counties with losses above $1 billion expand across 16 states, with nationwide total loss exposure rising to 70% from 65% in the 1-in-100-year flood scenario, representing about 2% of total counties.

Exhibit 7: In an intermediate-emissions scenario, nationwide potential uninsured loss exposure could increase 25% on average by 2050 in a 1-in-100-year flood

Counties with highest potential uninsured losses extend over five states, with moderate-to-high credit exposure to physical climate risk

In a 1-in-100-year flood, less than 2% of total counties have loss exposure of more than $1 billion and account for 65% of aggregate nationwide uninsured loss exposure, or around $245 billion.
 

In a more extreme 1-in-500-year scenario, 80% of the uninsured loss exposure becomes concentrated in counties with more than $1 billion in loss exposure, or 5% of total counties, while in the 1-in-100-year future climate scenario, 70% of the uninsured loss exposure is concentrated in counties with more than $1 billion in loss exposure, accounting for 2% of total counties (see Exhibit 8).

Exhibit 8: Uninsured flood risk exposure is highly concentrated in a small share of US counties

Uninsured loss exposure ratios show that lower loss magnitudes can still pose credit risks in some cases

Uninsured loss exposure ratios, or potential uninsured loss to total exposure value, provide an added lens to analyze risk exposure. Uninsured loss exposure ratios can be a better indicator of credit risks for some counties than the overall magnitude of losses. However, both metrics provide insights into flood risk exposure and concentration relevant for credit considerations. The total exposure value (TEV) by county represents the estimated total value of insurable residential assets within a county, regardless of whether those assets are affected by a given modeled flood event.^v
 

Counties with the largest potential uninsured losses do not always have the highest uninsured loss exposure ratios. This means that even in counties where the magnitude of losses may not be as high, the loss may still be significant for that particular county; conversely, for some counties, a high uninsured loss exposure in absolute terms may reflect a relatively small share of the county’s overall residential base.
 

In a 1-in-100-year flood, most counties’ uninsured loss exposure ratios fall below the 1% threshold, indicating a relatively manageable level of exposure (see Exhibit 9).
 

A small group of counties in six states – Florida, Kentucky, Louisiana, South Dakota, South Carolina and Texas – have uninsured loss exposure ratios exceeding 10%, where even one severe event could impose significant financial strain on households and local governments, absent relief aid. Of the six states, four have moderate-to-high credit exposure to physical climate risk, with the exceptions of Kentucky and South Dakota.

Exhibit 9: In a 1-in-100-year flood, most counties’ uninsured loss exposure is manageable as a share of property replacement cost

In a 1-in-500-year flood event, although most counties would still have uninsured loss exposure ratios of less than 1%, counties with ratios greater than 10% would expand to 16 states from six in the 1-in-100-year flood scenario (see Exhibit 9). Most of the 16 states have moderate-to-high credit exposure to physical climate risks, with the exceptions of Pennsylvania, Kentucky, Illinois and South Dakota.
 

The result is a widening disparity between areas with modest uninsured loss exposure and those where a single extreme event could overwhelm local recovery capacity, with some counties better positioned to absorb shock than others. 

Exhibit 10: In a 1-in-500-year flood, uninsured loss exposure increases significantly as a share of property replacement cost

In a future intermediate-emissions scenario, counties with potential uninsured loss to property replacement cost of more than 10% would expand into two additional states compared to the 1-in-100-year scenario: Mississippi, with moderate credit exposure to physical climate risk, and Pennsylvania, with low exposure (see Exhibit 10). 

 

Exhibit 11: A future 1-in-100-year flood climate scenario shows moderate expansion of potential uninsured loss exposure as a share of property replacement cost into more states, including ones with low physical climate risk

Appendix
 

Analytical approach
 

The Moody’s RMS US Inland Flood HD model was used to simulate losses from coastal storm surge, fluvial and pluvial flooding across multiple flood return periods (1-in-100 and 1-in-500 years) and a 1-in-100-year intermediate emissions climate scenario for 2050 in the contiguous US.
 

Two residential portfolios were modeled: Moody’s RMS insurable economic exposure and view of NFIP-insured residential exposure. Comparing the modeled results provides a consistent view of the insurance gap for each US county.
 

To assess possible future flood risk, the analysis used RCP 4.5 as an intermediate-emissions scenario to 2050. This scenario was selected to account for the possibility of nonlinear responses.
 

The modeled loss estimates are subject to uncertainty and should be interpreted as indicative rather than precise forecasts. Differences between estimated and realized losses may arise from variability in event characteristics, local flood mitigation measures, building attributes, claims behavior, post‑event loss amplification, and evolving insurance take‑up. Nevertheless, the use of consistent exposure definitions and a physically based modeling framework enables robust relative comparisons across geographies, event severities and climate scenarios. 

Exhibit 12: Flood footprint expansion across different scenarios shows higher overall losses and protection gaps, as well as flood risk exposure reaching states with high, moderate and low physical climate risk exposure
 

Estimates in the table below reflect aggregated potential nationwide loss exposure across US counties, rather than losses from a single nationwide flood event

These figures represent aggregate loss exposures from every county, rather than a single nationwide flood event.