A home appraisal is a number produced by a licensed professional, using a methodology authorized by Congress, that tells you what your house is worth. It is the largest financial measurement in most Americans' lives. And it is structurally prohibited from seeing the thing most likely to change what it measures.
For decades, that prohibition was the methodology's greatest strength. I want to trace how it became something else.
How the number gets made
A licensed appraiser visits your property. They walk the rooms, note the condition, measure the square footage. Then they turn away from your house and look at other houses. Specifically, at what similar homes in your area recently sold for. These are called comparable sales, or comps, and they are the foundation of residential appraisal in America.
The methodology is governed by the Uniform Standards of Professional Appraisal Practice, or USPAP, authorized by Congress in 1989. USPAP defines market value as an opinion that "presumes the transfer of a property... as of a certain date." The number is, by professional and legal definition, a portrait of what has already occurred.
For decades, this made sense. The past was a reasonable proxy for the present. A house worth $400,000 last spring, in a neighborhood where similar houses sold for similar prices, was almost certainly worth something close to $400,000 this spring. The methodology's conservatism was its virtue. It kept valuations tethered to actual transactions rather than to speculation. After 2008, when forward-looking optimism nearly destroyed the housing market, the backward-looking discipline of the appraisal process looked like exactly the kind of brake the system needed.
I want to hold onto that, because what's happening now is a methodology that worked, that was right to work the way it did, gradually becoming the thing that breaks the transactions it's supposed to support. The conservatism that protected people is now preventing the number from registering the single most disruptive force in residential real estate: the withdrawal of the insurance market from climate-exposed geographies.
The comps describe a world that no longer exists
In January 2025, the Palisades and Eaton fires destroyed roughly 16,000 structures in Los Angeles, impacting $51.7 billion in residential property value. By October 2025, average sale prices in the Palisades burn scar had dropped 33 percent from pre-fire levels. In the Eaton fire zone, 62 percent.
Now think about what an appraiser faces in the San Gabriel Valley six months after the fires. Sales volume has collapsed. The comps that exist are from before the fires, before insurance carriers pulled back, before premiums surged 35 to 50 percent, before many properties were pushed onto California's FAIR Plan. Those comps are technically "available," which is what USPAP requires. They are also artifacts of a market that no longer functions the way it did when those sales closed.
Four federal agencies temporarily paused certain appraisal requirements in LA County wildfire areas, a relief measure extending through January 2028. That pause is an implicit acknowledgment: the standard methodology cannot handle post-catastrophe conditions. But pausing the process without replacing it with anything is its own kind of answer. The number just doesn't get produced for a while. The question of what the property is actually worth hangs in administrative suspension.
LA is the most visible version of this, not the only one. HousingWire reported in April 2026 that across multiple markets, climate volatility, thin comp pools, and insurer withdrawal have converged to produce appraisals that:
"reflect an older version of the market."
Appraisers working with fewer, older comps that don't capture current conditions. The gap between what the number says and what the market is actually doing widening in real time.
An older version of the market. I keep coming back to that phrase. The number is accurate, in its way. It answers a question about a place that has already changed.
Two clocks, one transaction
The structural bind sounds like an abstract methodology problem. It kills real transactions.
A mortgage lender requires two things before closing a home purchase: an appraisal confirming the property's value, and proof that the borrower has adequate homeowners insurance. These two requirements evolved independently and operate on different temporal logics. The appraisal looks backward at comparable sales. The insurance market increasingly looks forward, at catastrophe projections, climate models, and loss ratios. For a long time, both pointed in roughly the same direction. A house the appraisal said was worth $350,000 could be insured for roughly that amount at a cost the buyer could absorb.
That convergence is breaking apart. A survey cited by the Levy Economics Institute, a progressive policy research center, found that nearly half of homebuyers and sellers encountered insurance problems during their transactions, and 21 percent reported that a deal fell through entirely. In Louisiana, the failure rate reportedly runs between 30 and 40 percent of mortgage loans. Eighteen states have introduced insurance reform bills in 2026, a legislative response to a crisis that is, underneath the policy language, a collision between a backward-looking number and a forward-looking market.
The sequence goes like this. A buyer makes an offer. The lender orders an appraisal. The appraiser pulls comps and produces a number. The buyer then tries to obtain insurance. And discovers that the property the appraisal says is worth $350,000 either can't be insured through the private market at all, or can only be insured at a cost that destroys the buyer's debt-to-income ratio and kills the mortgage qualification.
The appraisal was accurate on its own terms. The transaction is dead anyway.
Preliminary research from LSU, presented at finance conferences in 2025 and 2026, found something that sharpens this further. In counties with stressed insurance markets, banks didn't just charge higher interest rates. They had lower loan approval rates outright. The insurance crisis is making mortgages unavailable in affected markets. The appraisal number still gets produced. It just floats free of any closeable transaction.
The silence where guidance should be
I went looking for formal guidance from the Appraisal Foundation or the Appraisal Standards Board on how appraisers should handle climate risk, insurance availability as a valuation factor, or natural disaster impacts on comp selection. I couldn't find any. The most recent substantive USPAP update doesn't address it. The most recent Advisory Opinion, adopted April 2026, addresses the use of AI in appraisal assignments. Not climate.
The Appraisal Institute's newsletter referenced a forthcoming webinar on "how sustainability issues are playing a role in real estate valuation." The commercial real estate appraisal world is further along. The CCIM Institute published analysis in 2022 arguing that ESG scores may become "more impactful on valuation than the selection of a cap rate," a position that treats forward-looking risk as a core valuation input rather than an external distraction. On the residential side, where the stakes are most personal and the methodology most standardized, no comparable shift has occurred.
USPAP does contain a mechanism called the "extraordinary assumption," which allows an appraiser to presume as fact otherwise uncertain information about external conditions. In principle, an appraiser could use this to flag insurance availability as a factor affecting value. Whether climate-driven insurance conditions qualify as a trigger for extraordinary assumptions remains professionally contested. Which means, in practice, it mostly doesn't happen. This is the fault line within the profession: the tool exists, the authority to use it is ambiguous, and the institutional guidance that would resolve the ambiguity hasn't been issued.
Those 18 state legislatures introducing insurance reform bills? Every bill I can identify addresses the insurance side of the problem. None appear to touch the appraisal methodology that feeds into the same transactions. Eighteen states have recognized the crisis. None have addressed the valuation process producing the number the crisis has already emptied out.
FHFA, the federal agency overseeing Fannie Mae and Freddie Mac, has been researching the intersection of disaster risk, insurance availability, and mortgage performance. In 2024, the Enterprises shared preliminary research on topics including "housing demand impacts from insurance unavailability following wildfire risk." The fact that the agency backstopping most American mortgages is studying this problem is significant. The fact that it hasn't yet translated into appraisal guidance is equally significant.
The contaminated comp pool
For most American homeowners, home equity is their single largest asset, often the majority of their net worth. That equity is denominated, ultimately, in the appraisal number. The appraisal is what the lender uses to determine how much to lend. It establishes the baseline for refinancing. It anchors property tax assessments in many jurisdictions. It is the official answer to the question: what is this worth?
The widening gap between what homes appraise for and what they can be insured for means that official answer is increasingly disconnected from the conditions that determine whether a property can actually be bought, sold, or financed. California's insurance regulation had the same structural problem until recently: insurers were required to base catastrophe factors on historical data, not forward-looking models. The state has since begun allowing probabilistic modeling. The appraisal profession hasn't made a comparable shift.
California's insurer of last resort grew from 140,000 policyholders in 2018 to over 610,000 by mid-2025, with 97% holding home insurance policies. These properties appear in comp pools alongside privately insured homes, but represent fundamentally different market objects.
Those FAIR Plan properties sit in the same neighborhoods as privately insured homes. They appear in the same MLS databases. They may show up as comparable sales. But a FAIR Plan property is a fundamentally different market object than a privately insured one. Its buyer pool is narrower, because many buyers can't qualify for mortgages when the only available insurance is a last-resort plan with limited coverage and high cost. Its resale trajectory is constrained by the same insurance conditions that pushed it onto the FAIR Plan in the first place. The standard comp selection process has no reliable mechanism to distinguish these properties from their privately insured neighbors. Beyond failing to see the future, the methodology is incorporating distorted present-day data without flagging the distortion.
In the LA fire zones, investors have purchased 40 percent of lots in affected areas, often paying cash and bypassing the insurance requirements that block traditional buyers entirely. When those cash sales enter the comp pool, they look like market transactions. They're evidence of a market that has bifurcated into those who need insurance to buy and those who don't. The appraiser pulling comps has no standard way to weight that distinction.
Automated Valuation Models, increasingly used by lenders to speed up loan processing, are equally backward-looking. They run on the same historical transaction data. They just run on it faster, with less room for the kind of professional judgment that might flag a comp pool that's gone stale. New quality control standards for AVMs took effect in October 2025, but those standards address accuracy relative to the existing methodology, leaving the temporal limitation untouched.
What I can't resolve
There are reform ideas circulating. HousingWire's April 2026 analysis proposes that "climate risk data needs to be a standard part of the loan origination process, so that lenders can make better decisions upfront on how loans should work in high-risk zones." The redesigned Uniform Residential Appraisal Report, required for all submissions starting November 2026, could theoretically create space for new data fields. The question is whether those fields will include anything about insurance availability or forward-looking risk, or whether the new form will be a more efficient container for the same backward-looking methodology.
But I keep getting caught on something I can't think my way past. The backward-looking methodology was the thing that made the number trustworthy. It was the discipline that kept valuations grounded after 2008. Incorporating forward-looking climate risk into appraisals means incorporating uncertainty, projection, modeling. It means the number becomes, in some sense, speculative. And we have recent, painful evidence of what happens when speculation enters the valuation process.
So the profession faces a bind that goes beyond the technical. Staying backward-looking means producing numbers that increasingly describe a world that no longer exists. Incorporating forward-looking risk means abandoning the conservatism that made the number credible. Both directions carry costs for the people whose largest asset is denominated in this number. The professionals I can find writing about this in trade publications aren't oblivious. They're stuck.
I think about homeowners in markets where insurance carriers have quietly withdrawn over the past three years, where premiums have doubled, where the FAIR Plan is the only option. Their homes still appraise based on comps from before the withdrawal. The number on the form still looks solid. The wealth it represents is, in some meaningful sense, already compromised. But the official measurement can't register it yet, because the methodology that produces the measurement was designed, carefully and for good reason, to look only at what has already happened.
Transactions close on this number. Taxes are assessed against it. People make retirement plans around it. And underneath, the conditions that gave the number its meaning are shifting in ways the number's own methodology forbids it from seeing.
I don't know what replaces it. I'm not sure the profession does either, and I think the honest ones would say so. But we've built the largest store of American household wealth on a number that is, by design, looking in the wrong direction. And the longer that goes unaddressed, the wider the distance grows between what we think we own and what we actually hold.
-
The new appraisal form: The redesigned Uniform Residential Appraisal Report becomes required for all submissions starting November 2, 2026, and whether its data fields create space for insurance availability or climate risk factors will signal whether the profession treats this as a methodology problem or continues to look away.
-
LSU's insurance-mortgage research: Preliminary findings that stressed insurance markets correlate with outright loan rejections rather than higher rates are being developed by an LSU finance researcher using a dataset of 249 million insurance policies from a 2024 Senate Budget Committee investigation, and peer-reviewed publication would make this the strongest evidence yet of how insurance collapse transmits into credit markets.
-
Race and the appraisal gap: A Brookings Metro analysis from April 2026 identified race- and place-based factors shaping homeowners insurance in the climate era as a distinct problem, and the overlap with well-documented racial bias in traditional appraisal practices suggests the backward-looking methodology may compound existing inequities in ways not yet measured.
-
Colorado's disclosure model: Colorado's 2025 law (HB25-1182) requiring insurers to disclose their risk models and discount premiums for mitigation is the closest any state has come to making forward-looking risk legible inside a transaction, and whether appraisal methodology eventually follows the insurance side toward transparency is worth watching.