Seller Guide
The three different ways a Los Angeles home gets valued, why they disagree at the top of the market, and how to read the number you are handed.
Almost every Los Angeles seller begins with the same question and receives at least three different answers. A website returns an instant estimate. An agent presents a comparative market analysis. A lender orders an appraisal and produces a third figure that may sit above or below both. The natural reaction is to decide that two of the three are wrong. That is rarely what has happened. The three numbers are produced by different people, working to different standards, answering different questions, for different clients. Understanding which question each one was built to answer is the difference between pricing a house deliberately and pricing it by accident.
This guide explains what an automated valuation model actually does and where it breaks down, what a broker price opinion or comparative market analysis is and is not, and what a licensed appraisal adds that neither of the others can. It covers how comparable sales are genuinely selected and adjusted rather than how people imagine they are, what work does and does not translate into value, and what happens when a lender's appraiser returns a figure below the contract price. It is written for the upper end of the Los Angeles market, where the data is thin and the properties are not interchangeable.
None of this is an appraisal, and none of it is legal or tax advice. The purpose is to let you interrogate the number you have been given: who prepared it, from what evidence, under what standard, and as of what date. Where a figure or a rule has been verified against a primary source, that source is named. Where practice varies by lender, by property or by transaction, this guide says so rather than offering a tidy number that would not survive contact with your particular house. The people accountable for the answer are your agent, your appraiser and your lender, and they should be able to show their work.
An automated valuation model is a statistical estimate. It ingests public records, tax assessor data and, where available, multiple listing service history, and it predicts a number from patterns in prior transactions. No one visits the property. Since 1 October 2025 there has been a federal rule governing their use in lending: the Consumer Financial Protection Bureau, the Office of the Comptroller of the Currency, the Federal Reserve, the Federal Deposit Insurance Corporation, the National Credit Union Administration and the Federal Housing Finance Agency jointly adopted quality control standards requiring institutions that use automated valuation models in credit decisions to adopt policies addressing confidence in the estimate, protection against data manipulation, conflicts of interest, random sample testing, and compliance with nondiscrimination law.
A broker price opinion, or the comparative market analysis most sellers actually see, is an opinion of value prepared by a licensed real estate agent or broker for a client making a pricing or listing decision. It is informed by the same closed sales an appraiser would use, plus current competition and the practitioner's read of demand. It is not an appraisal and does not carry an appraiser's regulatory obligations. The California Bureau of Real Estate Appraisers makes a related point worth knowing: when a California-licensed appraiser develops an opinion of value, that work is an appraisal and must comply with the Uniform Standards of Professional Appraisal Practice regardless of what the report is called or what form it is written on, under Title 10, California Code of Regulations, Section 3701.
An appraisal is prepared by a licensed or certified appraiser, developed under those uniform standards, and in a financed transaction ordered by the lender. The distinction sellers most often miss is that the lender, not the seller and not the buyer, is the appraiser's client. The appraisal exists to tell the lender whether the collateral supports the loan. It is not a marketing document, it is not a negotiation tool, and it is not commissioned to confirm anyone's expectations. That independence is the point of it, and it is also why an appraisal can arrive at a figure no one in the transaction wanted.
Statistical models work where transactions are frequent, recent and genuinely similar. A tract of comparable houses built to three floor plans produces exactly the data a model needs. The upper end of Los Angeles produces the opposite. Sales are infrequent, properties are individually designed, lot topography varies within a single block, and the attributes that drive value at that level, including view, privacy, provenance, street position and the land itself, are either absent from the data or present only as a crude flag. A model cannot see that one house looks at a reservoir and the one behind it looks at a retaining wall.
The input data is also less reliable than it appears. Assessor records and listing histories carry square footage that may never have been measured, room counts that reflect a prior configuration, and improvements that were permitted, unpermitted, or permitted for something other than what was built. In a market where a meaningful share of high-end transactions is negotiated privately rather than through the multiple listing service, a model may also be reasoning from an incomplete picture of what has actually traded. Those private sales still inform what buyers and their agents believe, and they inform an experienced practitioner's opinion, but they are invisible to a system trained on published records.
The practical conclusion is not that automated estimates are worthless. They are a reasonable orientation on a homogeneous property in a liquid price band, and they are free. They are a poor foundation for a listing decision on an architecturally distinctive house, a large lot, a compound, a property with view or access issues, or anything above the price point where the comparable set thins out. Treat the estimate as a question rather than an answer, and note that a wide confidence interval is the model telling you, honestly, that it does not know.
The Fannie Mae Selling Guide, which governs appraisals supporting loans Fannie Mae purchases, requires a minimum of three closed comparable sales and expresses a preference for sales closed within the last twelve months, while acknowledging that the best comparable sales are not always the most recent. It requires the appraiser to state distance in miles with a directional indicator, measured as a straight line between properties. There is no rule confining an appraiser to one mile. Older comparables and comparables from competing neighbourhoods are permitted where better data does not exist, provided the appraiser explains why they were used. Loans above the conforming limits are not Fannie Mae loans, and each jumbo or portfolio lender sets its own additional requirements.
In Los Angeles the straight-line distance is often the least informative fact about a comparable. Market areas here are defined by hillside versus flats, by which side of a boulevard a parcel sits on, by school attendance boundaries, by whether a property lies inside a Historic Preservation Overlay Zone, and by view corridors that change over a hundred feet of elevation. A sale half a mile away across an arterial road can be a weaker comparable than one two miles away in the same pocket. A good appraisal and a good comparative market analysis both defend the market area they chose. If neither explains that choice, that is the question to ask first.
Sellers should also understand what closed sales, pending sales and active listings each contribute. An appraiser develops value primarily from closed transactions, because those are the only ones where price is known and settled. Pending sales indicate where the market has moved since those closings but the terms are not yet public. Active listings show competition and seller expectation, not value, and a list price is evidence of nothing except what someone hoped for. An agent's analysis uses all three deliberately, weighting closings for value and actives for positioning. A presentation built mostly on asking prices is a marketing document.
Once comparables are chosen, each is adjusted to the subject property: added to where the comparable is inferior, subtracted from where it is superior. The adjustments are meant to be derived from the market rather than from a fee schedule, typically by comparing pairs of sales that differ in one respect. Time adjustments matter in a moving market, and the Fannie Mae guidance is explicit that a nine month old sale with a time adjustment may be preferable to a one month old sale requiring multiple adjustments. That sentence contains the whole discipline: fewer and smaller adjustments mean a more defensible conclusion.
The corollary is that heavily adjusted comparables are weak comparables. When living area, lot size, view, condition, parking and location all require correction, the appraiser is effectively constructing the subject property arithmetically out of houses that do not resemble it. That is sometimes unavoidable on a distinctive property, and it is precisely where two competent professionals will diverge. It is also where a seller can help, by supplying the appraiser with the permit history, an accurate measurement of living area, and the closed sales the seller believes are genuinely comparable, along with the reasoning for each.
At the top of the Los Angeles market the grid eventually stops carrying the analysis. On a trophy property, the land, the architect, the privacy and the view can account for the majority of value, and none of those adjust cleanly. Valuation there becomes a reasoned argument about what a small set of qualified buyers would pay, supported by whatever transactional evidence exists. That is not a defect in the method; it is an honest description of a market with few participants. It is also why the seller's own knowledge of who bought what, and why, has real analytical value rather than merely anecdotal value.
The most durable rule in valuation is that cost is not value. What an owner spent is a historical fact about their budget, not a market fact about the property. Value is extracted from what buyers actually paid for comparable features, which is why a highly personal renovation can return a fraction of its cost while an unglamorous structural or systems correction can protect the whole price. Deferred maintenance behaves asymmetrically: buyers routinely discount visible neglect by more than the repair would cost, because the visible problem raises a question about the invisible ones.
Permit status is the specific Los Angeles trap. Square footage that was added without permits may be excluded from living area by an appraiser, treated cautiously by a lender, and disclosed to a buyer regardless. A converted garage, an enclosed patio, a basement finished into habitable space or an accessory dwelling unit that was never legalised can be valuable to a buyer and still fail to appear in the valuation as the owner expects. The same applies in reverse: a properly permitted and finalised accessory dwelling unit is a documented, income-capable asset. The difference between those two outcomes is paperwork at the Department of Building and Safety, and it is usually cheaper to resolve before listing than to negotiate afterwards.
A few other items recur in Los Angeles files. Owned solar and leased solar are not the same asset, because a lease or power purchase agreement travels with the property and may involve a recorded financing statement that title will surface. Pools, courts, guest houses and elaborate landscaping are valued by the market rather than by cost, and their contribution varies sharply by neighbourhood and lot. Views and light can be affected by a neighbour's entitlement rather than by anything on your parcel. None of these has a single number attached to it, and any adviser who offers one without market evidence should be asked to produce the paired sales behind it.
In a financed transaction, the lender sizes the loan against the lesser of the contract price or the appraised value. If the appraisal lands below the price, the shortfall does not reduce the price by itself; it converts into cash the buyer must find, or into a renegotiation. This is why the appraisal is a seller's problem as much as a buyer's, and why a seller who has assembled the evidence in advance is in a materially stronger position than one who learns of the gap from an agent's phone call.
There is now a formal channel for challenging an appraisal. In May 2024, Fannie Mae, in collaboration with Freddie Mac and the Department of Housing and Urban Development, published standardised requirements for a borrower-initiated reconsideration of value, with an implementation date later moved to 31 October 2024. Fannie Mae states that a borrower may request a maximum of one reconsideration of value per appraisal report, and that lenders are responsible for providing the form and disclosure. A reconsideration is an evidence-based process: additional closed sales the appraiser did not consider, corrections of fact about the subject property, and a clear explanation. It is not an appeal on the ground that the number was disappointing.
The other routes are ordinary contract mechanics. The parties can renegotiate price, the buyer can bring the difference in cash if the appraisal contingency was waived or removed, the parties can split the gap, or the buyer can cancel if an appraisal contingency remains in effect. A second appraisal is sometimes ordered but is a lender decision, not a party's right. And in the cash transactions that make up much of the upper Los Angeles market, there is no lender appraisal at all, which removes the mechanism entirely but does not remove the underlying question of whether the price can be supported on resale or on a future refinance.
Value and price are different decisions. Value is an estimate of what the property should transact for; price is a strategic choice about how to attract the buyers who will produce that outcome. The same defensible valuation can support a price set to invite competition, a price set to test the top of the range, or a price set to transact quickly, and each carries different consequences for days on market, for the negotiating posture you will hold when an offer arrives, and for what a buyer's appraiser will later see in the listing history.
In the City of Los Angeles that decision now has a tax dimension that did not exist before 2023. Because the Measure ULA transfer tax is charged on the full value of a qualifying conveyance rather than on the amount above the threshold, prices cluster beneath the tier boundaries and the band immediately above each threshold is unusually difficult to transact in. A valuation that lands near a threshold should be modelled at several prices with the transfer tax and your net proceeds calculated at each, before a list price is chosen. That is an arithmetic exercise, and it should be done before the property is on the market rather than after an offer arrives.
Whatever number you are given, ask for it in a form you can test. Which closed sales support it, when did they close, how were they adjusted, what market area was assumed and why, and as of what date is the opinion good. A seller who can answer those five questions can defend a price to a buyer, to a buyer's agent and, later, to an appraiser working for a lender who has never seen the property. A seller who cannot is relying on someone else's confidence, and confidence is not evidence. Ask for the analysis in writing, keep it, and agree in advance when it will be revisited if the market moves or the property does not sell.
Valuation standards, lender requirements and market conditions change, and no guide can value a specific property; confirm current requirements with your lender and appraiser and obtain your own professional advice before relying on any figure here.
Work With Ben
Every property is its own set of facts. Tell Ben what you are looking at and he will tell you what to check first.
Start the ConversationBen will reach out shortly to set up your tour.