How does an appraiser decide on one final value in California?

Reconciliation is where the appraiser makes the call

By the time an appraiser gets to the end of a report, the hard part should be visible. Every comparable sale (a comp, meaning a similar nearby home that sold) has been adjusted to look like your house. Now there are several adjusted prices on the page, and they never match to the dollar. Real sales are messy. People are messier.

Reconciliation is the step that turns those prices into one opinion of value. Fannie Mae's Selling Guide requires the appraiser's comments to reflect the reconciliation of the adjusted values and to say why the sales given the most weight earned it. The indicated value has to fall within the range of the adjusted sale prices.

In other words, the appraiser has to show the reasoning. "Trust me" is not a reconciliation. It is a fortune cookie with a license number.

Why averaging the comps is the wrong move

The lazy way to reconcile is to add up the adjusted prices and divide. It looks neutral. It is not.

An average treats a sale on your street with almost no adjustments the same as a sale across town that needed $95,000 of corrections. It lets the weakest comp pull just as hard as the best one. That is like letting the uncle who "knows the market" cast the same vote as the house next door that sold last month.

Fannie Mae says it plainly in its section on valuation analysis and reconciliation: reconciliation must never be an averaging technique. The one exception is a weighted average, and only with a proper explanation of the weights.

So the question is never "what is the middle number?" It is "which sales tell me the most about this house, and why?"

Which comps get the most say

Here is what moves a comp up my list when I reconcile:

• Fewest and smallest adjustments. Fannie Mae notes that the best comp would ideally need no adjustment at all. Every adjustment is a place where error can sneak in.
• Most similar where buyers care most. Same pocket, same condition tier, same layout. Condition often beats age with buyers, so a match on condition can matter more than a match on year built.
• Most recent. A sale from three weeks ago knows things a sale from nine months ago does not, even after a time adjustment.
• Bracketing. I want sales above and below your house on the features that matter, so the adjustments are not all pushing one way.
• Verified. A sale I confirmed with the agent or the records beats one with a mystery seller credit.

One caution. Fannie Mae also says the number or size of adjustments should not be the only test of whether a comp is acceptable. A heavily adjusted sale can still be the best evidence for a feature nothing else shares. It just has to earn its place with support.

A hypothetical example: a 1960s ranch in Concord

Say I am appraising a hypothetical 1960s single-story ranch in Concord. It has three bedrooms, two baths and an updated kitchen. After every adjustment, four sales land like this. I list them from least adjusted to most adjusted, so you can check the order on the gross adjustment column:

The adjusted range is $910,000 to $940,000. A straight average is $922,500, and that number leans on Sale 4, the comp with the most corrections and the oldest date.

I would land at $915,000. Sale 1 is on the same street, needed almost nothing and is the freshest read of the market, so it gets the most say. Sale 2 brackets it from below. Sales 3 and 4 confirm the top of the range but carry more guesswork.

For comparison, a weighted approach of 50% on Sale 1, 20% each on Sales 2 and 3, and 10% on Sale 4 comes out to $918,500. Both methods point to the same neighborhood, and the report would say why the final number sits where it does. Print order and weight are separate decisions. The table is sorted by adjustment size. The weights come from the reasoning.

Weighting versus judgment inside the range

Some appraisers assign percentages to every comp. Others reach a value by judgment inside the adjusted range and explain the reasons in words. Both are legitimate. Appraisal is part numbers and part informed judgment, and the rules allow either one as long as the reasoning is written down.

What is not legitimate is starting with a number and working backward. The value has to come from the evidence. The contract price does not set it, and neither does a refinance target. If the evidence lands on the contract, fine. It got there on its own legs.

A good reconciliation reads like a short argument: here is the range, here is the sale that says the most, here is what the others add, and here is why the value sits at this spot and not $20,000 higher. If you finish reading it and still cannot tell why, that is a fair question to ask through your lender or client.

What I see in the field, and how to check a reconciliation yourself

The most common weak spot I see is a reconciliation that is one sentence long. "All comps were considered, with most weight given to Sales 1 and 2." Why? Silence. The second most common is a value that sits exactly where the contract sits, with no sale that actually supports it.

When you read your own report, check these:

1. Find the adjusted range. Is the final value inside it? Fannie Mae requires that for the sales comparison approach.
2. Find the most-weighted sales. Does the report say why they won?
3. Look at the gross adjustments. Do the favored sales have the smallest ones? If not, is there a reason?
4. Check bracketing. Are there sales both above and below your house on size, lot and condition?
5. Ask about the outliers. If one comp pulls far from the rest, the report should explain what it adds.

If you need a value that walks the reader through every comp and every reason, that is how I write my narrative reports at ValuedAudit. You should be able to follow the logic without a decoder ring.