The Appraisal: Who Decides What a House Is Worth

A buyer and a seller agree on a price. That agreement — two informed parties, at arm's length, transacting voluntarily — is as close to a definition of market value as economics offers. Then a third party who has never met either of them visits for forty minutes, and if their number is lower, the deal changes or dies. The appraisal is the point where an individual transaction is checked against a professional opinion, and it exists because the lender is putting up most of the money. Understanding that single fact — you pay for it and it isn't for you — explains almost everything else that confuses people about the process. This report examines the methodology, the automation reshaping it, and the recourse borrowers rarely use.

The core thesis

Our thesis is that the appraisal is best understood as a collateral control that consumers experience as a verdict on their purchase, and that most of the friction in the process comes from that mismatch of expectations.

The lender's question is narrow: if this borrower stops paying and we sell the property, will we recover the loan amount? Everything about the appraisal follows from that — the emphasis on recent comparable sales rather than on future prospects, the conservatism, the requirement for independence from anyone with an interest in the transaction closing, and the fact that a value above the contract price produces no benefit while a value below it produces a problem. The appraisal is a floor check, not a valuation service.

The second part of the thesis concerns methodology. The dominant approach values a property by reference to what similar nearby properties recently sold for. That is sensible, defensible, and structurally backward-looking: it encodes existing price relationships between areas into the valuation of every new transaction. In a stable market that's accuracy. Where historical price relationships reflect something other than the properties themselves, the method transmits it forward — which is the mechanism underneath the valuation bias concerns that prompted federal attention, and it operates without any individual doing anything wrong.

Third, automation is quietly reallocating who bears valuation risk. Automated valuation models and appraisal waivers reduce cost and time, and they remove a physical inspection that was doing work nobody had priced — condition assessment that protected the buyer incidentally while serving the lender.

A value above the contract price does nothing. A value below it creates a problem. That asymmetry tells you exactly what the document is for.

Whose document it is

The practical consequences of lender ownership, which surprise nearly every first-time buyer:

• You don't choose the appraiser. Independence requirements exist specifically to separate appraiser selection from parties with an interest in the transaction closing — a reform that followed the valuation failures of the last major housing cycle.
• You generally can't discuss value with them. Contact intended to influence the outcome is prohibited, which is why the appraiser may seem unhelpfully remote.
• You're entitled to a copy. Lenders are generally required to provide it, and reading it is worthwhile — it contains the comparables used and the adjustments made.
• You pay for it, as part of closing costs, regardless of outcome.
• It isn't an inspection. An appraiser notes condition as it affects value; they aren't examining the furnace or the roof structure. A separate inspection does that, and conflating the two is a common and expensive error.
• It's a point-in-time opinion with an effective date, and it can expire before closing on a slow transaction.

One consequence worth drawing out, because it reframes the anxiety: the appraisal is checking the collateral, while the rest of the file — income, obligations, and the debt-to-income capacity our readiness guide identifies as the usual binding constraint — is checking you. A low appraisal is not a judgment about your creditworthiness, and it doesn't touch your file the way an application does. It's a statement about a building.

The comparable sales method

The dominant approach for residential property: identify recently sold properties similar to the subject in location, size, age, and characteristics, then adjust for differences to estimate what the subject would sell for.

The judgment sits in three places, and understanding them is what makes an appraisal readable:

Comparable selection. Which sales are chosen — how recent, how near, how similar. This is the most consequential choice in the report and the one most often contested. A property valued against comparables from a different submarket produces a different number, and the boundary between submarkets is a judgment.

Adjustments. Each comparable is adjusted for differences — an extra bathroom, a larger lot, a finished basement, a garage. The adjustment amounts are estimates derived from market evidence, and reasonable appraisers differ.

Reconciliation. Weighting the adjusted comparables into a single conclusion.

Two structural features worth naming. The method requires transactions to reference. In areas with few recent sales, unusual properties, or rapidly changing markets, the comparable pool is thin and the estimate is correspondingly less reliable — which means valuation quality is systematically worse in exactly the places where property is least liquid.

And the method is inherently lagging. Closed sales reflect contracts signed weeks or months earlier. In a market moving quickly in either direction, appraisals trail — producing appraisal gaps in rising markets and, less discussed, supporting values that the market has already left behind in falling ones.

The appraisal gap

When the appraised value comes in below the contract price, the lender lends against the lower figure. The difference is the appraisal gap, and it is the buyer's problem to resolve.

Why it's the buyer's problem: the loan amount is calculated from the value, so a lower value means a smaller loan against the same price — and the shortfall must be covered in cash. A buyer purchasing at $400,000 with an appraisal at $385,000 doesn't get a smaller mortgage on a cheaper house; they get a mortgage sized on $385,000 and still owe $400,000 under the contract.

The available responses:

Option	What it requires
Bring cash	Liquidity to cover the gap on top of the down payment — the most common resolution in competitive markets
Renegotiate	A seller willing to reduce, which depends on whether they have other buyers who would face the same appraisal
Split the difference	Both sides moving, which is frequently where these land
Reconsideration of value	Factual grounds — see below
Second appraisal	Lender agreement, which is not routinely granted absent identified deficiencies
Change loan structure	A different program or a larger down payment percentage, if available
Walk away	An appraisal contingency in the contract — which is why waiving it is a substantive risk rather than a formality

The point that matters most and is decided earliest: which of these you have available was determined when you signed the contract. An appraisal contingency preserves the exit; an appraisal gap clause committing to cover a shortfall up to a stated amount commits you before the number exists. In competitive markets buyers waive contingencies to win offers, which converts an unknown into an unsecured obligation — and the buyers doing it are frequently those with the least cash to cover it.

Reconsideration of value

The formal mechanism for challenging an appraisal, and one most borrowers never hear about.

A reconsideration of value asks the lender to have the appraiser reconsider, on grounds such as factual errors in the report, comparable sales that were available and not used, or property characteristics that were overlooked. Federal agencies issued interagency guidance in 2024 addressing how institutions should handle these requests — including establishing clear, consistent procedures and ensuring consumers are informed the option exists.

What makes a request effective:

• Specific factual grounds — a room count that's wrong, square footage understated, a renovation not reflected, a comparable that closed and wasn't used.
• Better comparables identified, with addresses, sale dates, prices, and an explanation of why they're more similar than those used.
• Documentation of improvements with permits, invoices, and dates.
• Submission through the lender, since direct contact with the appraiser is generally prohibited.

What doesn't work: disagreeing with the conclusion, citing what you believe the property is worth, or pointing to listing prices rather than closed sales.

The reason to raise its profile: this is a real remedy that is underused because of an information gap rather than a merit gap — the same pattern this desk has documented in the arbitration opt-out, the notice of error in servicing, and extraordinary-circumstances provisions in insurance. The right exists; almost nobody knows to ask. A buyer facing a gap and an agent who knows the comparable pool can frequently assemble a credible request in an afternoon.

Automation and waivers

Two related developments have changed the valuation landscape substantially.

Automated valuation models estimate value from property records, prior sales, and market data without a physical visit. They're fast, cheap, and reasonably accurate in aggregate — and their accuracy degrades exactly where the comparable method degrades: unusual properties, thin markets, and areas with poor data. Federal agencies finalized a rule establishing quality control standards for AVMs used in mortgage decisions, requiring institutions to adopt policies designed to ensure a high level of confidence in estimates, protect against manipulation, avoid conflicts of interest, require random sample testing, and comply with nondiscrimination law.

Appraisal waivers offer to proceed without a traditional appraisal on qualifying lower-risk transactions, relying on modeled valuation and existing data.

The honest assessment of waivers, in both directions:

• For: faster closings, a saved fee, and one less scheduling dependency — genuine benefits, particularly on a refinance of a property with substantial equity and known characteristics.
• Against: nobody looks at the house. The physical appraisal was performing an incidental condition check that a buyer benefited from without paying for separately. A waiver removes it, which matters most on an older property or one a buyer hasn't inspected thoroughly.
• And: modeled values can be wrong in individual cases, and the borrower has no report to read or challenge.

Which produces a clean recommendation: a waiver is usually fine on a refinance and worth thinking harder about on a purchase, particularly of an older home. And in either case, a waiver is never a substitute for a home inspection — a distinction that gets lost precisely because both involve someone visiting the property.

There's a further reason to look closely at an older property whatever the lender requires. Condition drives insurability, and roof age in particular determines both what a carrier will write and at what price — the availability constraint our property insurance analysis documents. A waiver that saves a few hundred dollars on a house that turns out to be expensive or impossible to insure is a poor trade, and the two checks are worth sequencing together.

The bias question

Valuation bias attracted sustained federal attention, including an interagency task force examining appraisal equity, and the resulting guidance on reconsideration of value and AVM quality control both reflect it.

The claim is that properties in some neighborhoods have been valued lower than comparable properties elsewhere in ways not explained by the properties themselves, and that this operates both through individual judgment and — more consequentially — through method.

The methodological channel is the one worth understanding, because it requires no individual intent:

• Comparable selection is geographically bounded. Valuing a property against nearby sales means a neighborhood's price level is largely determined by its own history.
• Historical price relationships between areas reflect the full history of how those areas were treated, including periods when access to mortgage credit itself differed by neighborhood.
• The method transmits that forward. Each valuation references prior sales, which referenced prior sales — a recursive structure in which historical differences persist as current ones.

This is the same structural mechanism our governance report identifies in modeling generally: a method trained on history reproduces history, and doing so is what makes it accurate by its own measure. The remedies being pursued — clearer reconsideration procedures, AVM quality controls with nondiscrimination requirements, and testing — address the mechanism rather than assigning intent, which is the appropriate response to a structural problem.

Two honest qualifications. Disentangling valuation effects from the many other factors affecting neighborhood price levels is genuinely difficult, and the research is contested in its magnitudes. And the appraiser is frequently the last party in a long chain, describing price relationships they did not create.

Appraisals in refinancing and equity

The appraisal governs more than purchases, and the stakes differ.

In a refinance, the value determines your loan-to-value ratio, which determines eligibility, pricing, and whether mortgage insurance is required. A low appraisal can eliminate the benefit of refinancing entirely.

In home equity borrowing, the value determines how much is available — the mechanism our equity extraction report examines, where the valuation is what converts a housing market movement into borrowing capacity.

In mortgage insurance removal, an appraisal can establish that the ratio has fallen enough to cancel it — an actively useful application that homeowners routinely leave on the table for years, at real monthly cost. It's the same category of unexercised right this desk keeps documenting, alongside the notice of error in our servicing analysis: available, valuable, and requiring the borrower to know it exists and ask.

Practical guidance for these: prepare the property, since condition affects the outcome; document improvements with permits and invoices, because an appraiser cannot credit what they don't know about; and know your own comparables before the visit, so a low result can be evaluated rather than simply accepted.

What to do about a low appraisal

1. Get the report and read it. You're entitled to it. Check the basics first — square footage, bed and bath count, lot size, condition rating.
2. Examine the comparables. Are they genuinely similar? Recent? In the same submarket? A comparable across a boundary that matters locally is the most common defensible objection.
3. Look for omissions — closed sales that support a higher value and weren't used.
4. Check the adjustments for anything that looks inconsistent between comparables.
5. Assemble a reconsideration request if you have factual grounds, with specific addresses and documentation, submitted through the lender.
6. Talk to your agent — they have access to sales data and frequently know the local comparable pool better than an appraiser from another area.
7. Check what covering the gap does to your reserves. Bringing extra cash to closing depletes the buffer our savings analysis argues should survive the purchase — and some loan programs require reserves remaining afterward, so the cash may not be as available as it looks.
8. Renegotiate in parallel, since a low appraisal is information the seller also has to reckon with, and their next buyer will likely face the same number.
9. Consider whether to proceed at all. An appraisal below contract price is a professional opinion that you agreed to pay more than the property is worth. Sometimes that's a market timing artifact and sometimes it's correct — and the discipline to treat it as information rather than an obstacle is worth having.

Scenarios and what we're watching

Scenario	Shape of the world	Signposts
Base case — hybrid valuation	Automated models handle routine low-risk transactions; physical appraisals concentrate on complex, high-value, and thin-market properties; waiver share grows gradually	Waiver rates; AVM accuracy testing; appraiser workforce trends
Automation case	Waivers and desktop methods dominate; valuation becomes largely a data exercise; condition risk shifts to buyers and inspections	Waiver eligibility expansion; inspection uptake; post-closing condition claims
Correction case	Bias findings and AVM quality requirements drive methodological change in comparable selection and adjustment	Reconsideration request volumes; AVM testing disclosures; guidance developments

What we're watching: reconsideration of value volumes, which measure whether the 2024 guidance actually reached consumers or remained a procedural requirement; appraisal waiver share and whether condition problems surface later; AVM quality control testing, particularly the nondiscrimination compliance component, which is the first time modeled valuation has carried an explicit fairness obligation; and appraiser workforce trends, since an aging and shrinking profession is itself a driver of automation regardless of whether automation is better.

Two parties agree on a price and a third decides whether the agreement counts. That's a reasonable control on a lender's collateral, and it becomes something else when the number moves a household's largest transaction — which is why knowing that you can question it, and on what grounds, is worth more than any general advice about the process.

Frequently asked questions

---

This report is for general information only and does not constitute financial, legal, or valuation advice. Appraisal requirements, waiver eligibility, reconsideration procedures, and applicable guidance vary by loan program, institution, and jurisdiction and change over time; confirm current requirements with a licensed mortgage professional.