Five Numbers for One House

I spent five years on cargo ships, and one thing you learn at sea is that every vessel carries several values at once. There's what the owner insured it for, what the cargo's worth, what a buyer would pay for the hull, what the bank holds against it, and what it would cost to build another one. Those numbers don't match. Nobody in the wheelhouse expects them to, and the ship goes on floating either way.

A house in a climate-stressed part of America works the same way now, with one difference. The house is where you live, and the person stuck reconciling the numbers is you.

A single property can carry five official prices produced by five different institutions: the county tax assessor's value, the insurer's replacement cost, the mortgage lender's collateral figure, FEMA's buyout offer, and whatever the open market will bear on a given Tuesday afternoon. Each institution follows its own logic, answers its own question, and has no obligation to check its work against anybody else's. For a long stretch this didn't matter, because the numbers stayed close enough that the gaps were pocket change. They aren't pocket change anymore.

What each number actually measures

The tax assessor estimates what the property would sell for, then in many states applies caps and exemptions that drag the taxable figure lower. In Florida, the Save Our Homes amendment limits annual increases in assessed value to 3% or inflation, whichever is less. A property's tax record can hold three distinct numbers — just value, assessed value, taxable value — and none of them necessarily reflects what a buyer would pay today. The assessed value can keep climbing after the market drops, or lag a boom for years. The system was built to keep long-term homeowners from getting taxed out of their houses. It was not built for a market that moves twenty percent in a year because a wildfire rewrote the neighborhood's future.

The insurer estimates what it would cost to rebuild the physical structure at current materials and current labor, and deliberately leaves out the land. That's replacement cost. When Citizens Property Insurance writes a policy in Florida, an agent runs a replacement-cost estimator that prepopulates the coverage amount. That estimator's output isn't published anywhere, doesn't appear in any county report, and can't be compared to the assessor's number, because the two are measuring different things. The assessor prices what stands there. The insurer prices what it would take to put it back.

The mortgage lender got a market-value appraisal when the loan closed. Fannie Mae defines that as the most probable price under competitive conditions. After closing, the lender mostly tracks the unpaid balance, which is a record of debt and not an opinion about the house. When disaster hits, Fannie Mae's servicing guidance tells the servicer to assess damage and manage insurance claims. It doesn't tell servicers to order fresh appraisals across a disaster zone. So the lender may be holding an origination appraisal from 2019, a balance from this month, a damage inspection from last fall, and an automated estimate spat out by a model, with no single one of those constituting an official collateral value.

FEMA's buyout program offers current or pre-disaster market value, depending on which method the local project picks. After the sale, the government knocks the structure down and holds the land as permanent open space. Nothing in the offer guarantees the check covers the mortgage or buys a comparable house somewhere drier. It's what the government will pay.

The open market produces a number only when somebody actually buys. That number swallows everything the other systems try to isolate: structure, land, location, flood history, insurance availability, and the buyer's private guess about the next twenty years. It's the most complete figure of the five and the least reliable.

Where the numbers are coming apart

Buncombe County: the debt outlasts the house

After Hurricane Helene hit western North Carolina, Buncombe County adjusted property values to reflect storm damage. The tax record for 27 River Knoll Road shows a 2024 value of $237,600 and a 2025 value of $88,500. The record for 19 River Knoll Drive went from $230,500 to $86,300. Both properties sit in an approved FEMA acquisition group. Local reporting found owners still making mortgage payments on uninhabitable houses while waiting on buyout checks.

Here is the full public record for 27 River Knoll Road, and the rest of it:

• Tax value, 2024 (pre-storm): $237,600
• Tax value, 2025 (post-storm): $88,500
• Insurance replacement cost: not public
• Mortgage balance: not public
• FEMA buyout appraisal: not public
• Market value (current): not public

Kern's reporting on a comparable Buncombe property in the same buyout process found a mortgage balance of roughly $270,000, better than three times the post-storm tax value. The buyout appraisals themselves aren't public. Neither are the payoff amounts on these parcels. The GAO's finding, published in March 2026, confirms the general condition: homeowners can owe more on a mortgage than a FEMA acquisition would yield. GAO also reported that acquisition projects typically run two to three years and often longer, with states averaging sixteen months just to get the application filed.

I wrote about that carrying cost in an earlier piece for this publication. The household finances the gap between institutional recognition and delivery, paying mortgage, insurance, and taxes on a property it can't sleep in, waiting on a check that may not clear the debt.

FEMA's acquisition rule requires clear title, which means the lienholder has to release the mortgage before the sale closes. If the offer comes in under the balance, somebody has to eat the difference. The regulation is silent on who.

Los Angeles: four numbers between the fire and the rebuild

The LA fires produced a different kind of gap. The houses are gone. The question is what it costs to put them back, and who's holding the money while that gets decided.

Los Angeles County estimated rebuilding at $300 to $800 per square foot in a June 2025 report, and called that range conservative. On a 2,000-square-foot house, that's $600,000 to $1.6 million before contingencies. AIA Los Angeles put preliminary ranges at $350 to $900 per square foot in Altadena and $650 to $1,200 in Pacific Palisades, and recommended tacking on another 15 to 25 percent for contingency.

United Policyholders, a policyholder-advocacy nonprofit, surveyed 692 fire-affected respondents and found that 69% of those with total losses said they didn't carry enough insurance to rebuild. What they had against what rebuilding costs:

The survey is voluntary and not representative, but the direction it points matches the county's own numbers.

Then the servicer walks in. Even where the insurance proceeds exist, the mortgage holder controls when they get released. Fannie Mae's servicing rules authorize an initial release of up to the greater of $40,000 or 33% of proceeds on a current loan, with the balance paid out in stages after repair inspections. On a delinquent loan the initial release drops to 25%, capped at $10,000 or whatever exceeds the unpaid balance. The Los Angeles Times documented cases where servicers sat on tens of thousands of dollars while quoting completion benchmarks that kept moving: 33%, then 37%, then 46.99%. By July 2026 the California Department of Financial Protection and Innovation had taken in more than 120 complaints about lenders handling insurance money.

Maya Jiménez's reporting for this publication traced the same setup through debris removal, with the homeowner wedged between government and insurer claims on one pool of coverage while recovery arrives in pieces across agencies and forms.

So you've got the insurer's replacement-cost estimate, the policy's coverage limit, the actual claim payment, and the servicer's release schedule. Four figures, and none of them is yet the contractor's bid. The household is expected to make all of them work together. Nobody else is required to.

Coastal Florida: the bet nobody measures

A buyer in Cape Coral today is pricing something neither the tax assessor nor the insurer formally accounts for: whether insurance will still be available, and affordable, in five or ten years. That bet rides inside the purchase price. It isn't in the tax assessment, which estimates current market value and then applies constitutional caps that can shove the taxable figure thirty percent below it. It isn't in the replacement-cost estimate, which prices lumber and labor at today's rates and says nothing about next year's premium or whether a carrier will write the policy at all.

GAO found that inflation-adjusted homeowners premiums rose at least 25% in parts of the southern coast between 2019 and 2024, and that otherwise similar homes in high-wind-risk areas carried premiums roughly 58% above medium-risk areas. Buyers absorb those figures and make a guess about where the line goes next. The assessor's schedule of values makes no such guess. Neither does the replacement-cost calculator. The market is pricing a variable that two of the five systems were never built to see.

What the market is saying depends on who's doing the reading. Florida Gulf Coast University reported inflation-adjusted median prices across Southwest Florida down 7% between the fourth quarters of 2024 and 2025. Redfin's Cape Coral data through June 2026 showed a nominal median of $369,799, up 0.7% year over year. Up or down depends on your window, your geography, and whether you adjust for inflation.

Florida's tax system, meanwhile, generates its own internal spread before any other institution shows up. For Lee County in 2025, total just value ran $219.9 billion while county taxable value came in at $149.6 billion. Sixty-eight cents on the dollar. That $70 billion gap is constitutional design, not error.

Citizens Property Insurance listed 5,586 Lee County homeowner policies at a current average premium of $3,322, plus 1,219 wind-only policies averaging $4,550. The replacement-cost estimates behind those policies — the numbers that would let you compare what the insurer thinks a rebuild costs, what the assessor has recorded, and what a buyer would pay — don't appear in any public report. In coastal Florida, the comparison that matters most can't be assembled from public data at all.

The report nobody has written

I went looking for a federal study — GAO, FHFA, anybody — that puts all five numbers on one page for the same properties and calls the divergence a structural condition. There isn't one.

The pieces exist separately. GAO has studied buyout delays, the mortgage-over-buyout problem, and rising insurance costs in disaster-prone areas. FHFA has surveyed the literature on how floods and wildfires move home prices and mortgage performance. FHFA's own climate-scenario work admits to incomplete property-level data and substantial model sensitivity, and notes that existing credit-loss models leave out land-value changes, insurer exits, migration, and declining local economies.

Each institution studies its own number and has no operational reason to look at anybody else's. The assessor doesn't need the replacement cost, the insurer doesn't need the mortgage balance, FEMA doesn't need the tax roll. No regulator has either the authority or the mandate to compare them. But the household carries all five — it owes the mortgage, pays the premium, pays the property tax, takes or refuses the buyout, sells or doesn't sell. Those numbers land on the same kitchen table, and nobody is required to make them agree.

Questions worth asking

I'm not qualified to tell anybody what to do with a specific property, and anyone who claims to be without knowing your specific numbers is selling something. But the divergence points to a few things worth working out before a disaster forces the issue.

What are your numbers right now? "What's my house worth" has at least four answers, so ask them separately. What does the tax assessor say? What does your policy's replacement-cost estimate say? What's the mortgage balance? What would a buyer pay this month? Most people have never seen those four figures next to each other. The spread can be startling.

Which way is each one moving? Assessment caps can make the tax number lag reality in either direction. Replacement costs follow construction inflation, which has been outrunning general inflation in a lot of markets. The mortgage balance declines on a schedule that pays no attention whatsoever to what's happening outside. And market prices are increasingly absorbing information about insurance availability, flood history, and fire risk that the other systems haven't caught up to.

If the house burned down tomorrow, what would each system produce? The insurer pays replacement cost up to your limit, less the deductible, less depreciation until repairs get verified, less whatever the servicer holds back pending inspections. The assessor lowers the assessed value. The lender still expects the payment on the first. A buyout program, if one ever materializes, offers market value that may or may not cover the debt. And the market value of a burned lot in a disaster-prone area is whatever somebody will pay for the dirt and the right to build on it, assuming they can find a carrier.

Where do the experts disagree? FHFA's own researchers say their models omit major variables. The insurance market is moving fast enough that this year's replacement-cost estimate tells you little about next year's premium or whether coverage will be offered. Appraisers and automated valuation models routinely disagree about the same property. Whether climate risk is already priced into a given market is unresolved in the academic literature, and the answer probably varies by geography, by how sophisticated the buyers are, and by what sellers are required to disclose.

The five numbers were never the same number. For a few decades they landed close enough that the gaps didn't matter much. They matter now, and the institutions producing them have no plans to reconcile them. That work falls to the homeowner.