How to estimate ARV from sold comps

An after-repair value you can check by hand starts as a list of sold prices, not as a feeling about the block and not as the seller’s ask. Sort the prices. The median is the middle one. This page uses three invented sales so the arithmetic is visible: $176,000, $198,000, and $241,000. Sorted, they are already in that order. The middle price is $198,000. That is the hand estimate of ARV. It is not a closed sale of the house you are underwriting, and it is not the weighted median a full analysis prints.

Why the average is a different number

Add the same three prices and divide by three. $176,000 plus $198,000 plus $241,000 is $615,000. Divided by three, that is $205,000. The average is $7,000 higher than the median. The entire gap is the $241,000 sale. A single high sale pulls an average up by a third of the amount it sticks out. It does not move the median at all unless you have an even count and that sale is one of the two central prices. If you are estimating ARV because one comp looks hot, the average hides the problem by splitting it across the set. The median leaves the hot sale at the end of the list, where you can see it and decide whether it belongs.

People search “how to estimate ARV” and get told to average the comps. Averaging is easy, and it is the wrong default when the list is short. Three sales is a short list. One unlike house is a third of the evidence. On these invented numbers the unlike house is $43,000 above the middle. A third of $43,000 is about $14,333, and the average only rose $7,000 because the low sale at $176,000 pulls the other way. The two effects do not cancel into a better ARV. They cancel into a number that matches none of the three houses. $205,000 was not a sale. $198,000 was.

Turn the median into a 70 percent offer

The 70 percent rule on this median is the median times 0.70, minus rehab, rounded once to the nearest dollar. Rehab here is an invented $36,000. Seventy percent of $198,000 is $138,600. Subtract $36,000 and the offer is $102,600. You can check that on the ARV and 70% rule calculator by typing the three sold prices and the rehab. The page’s median of an odd count is the middle price, so it will print $198,000 and an offer of $102,600. Nothing you type there is uploaded.

The same ARV and the same rehab on the 25/5/3 formula are a lower offer, and they should be. Profit is 25 percent of $198,000, which is $49,500. Closing is 5 percent, which is $9,900. Holding is 3 percent, which is $5,940. The maximum allowable offer is $198,000 minus $36,000 minus $49,500 minus $9,900 minus $5,940, which is $96,660. The 70 percent offer was $102,600. The gap is $5,940, which is the 3 percent holding reserve the shorter rule does not itemize. Both numbers are invented checks. The MAO calculator prints the 25/5/3 line and the 70 percent line from an ARV you already have. This page is about how you got the ARV.

What you throw away before you sort

The median only helps after you have decided the sales are the same kind of house. A new-construction sale, a different school, a busy road, or a house that was not repaired do not become comparable because you typed them. If you leave the $241,000 sale in, the median ignores it and the average does not. If that sale is actually the best match and the $176,000 sale is the mismatched one, the median is now the one that is wrong, because the middle of a bad list is still a bad middle. Estimating ARV is the decision about membership first and the sort second. The note on which comps to throw out works an even-count list where one unlike sale still nudges the median, which is the case this three-sale example does not show.

Price per square foot is a second column, not a substitute for the sold price. If the $176,000 house is smaller, its price per foot can be higher than the $198,000 house. Sorting by price per foot would pick a different “middle.” The hand estimate on this page sorts sold prices. The ARV per square foot page shows a pair where applying a comp’s price per foot to the subject invents a value that neither sale supports. Do not do that extra step and still call the result this median.

What a full analysis adds

A full analysis does not stop at the middle of three prices you pasted. It adjusts the sold price when bedrooms or bathrooms differ, then weights the adjusted price. A sale within three months is multiplied by 3. A sale within a quarter mile is multiplied by 3. A missing distance is a neutral 1, not a bonus for being next door. Five comps fill the count portion of the confidence score, because that portion is the count times 6, capped at 30. A sixth comp does not add count points. None of that is in the $198,000 median above. The comp weight guide is the table. Use it when the three prices are real candidates and you need to know which one should count more.

Confidence is not a promise that the next buyer pays $198,000. It is a score on the set. A tight set of recent sales scores higher than three sales spread from $176,000 to $241,000. On this invented set the spread is wide, so a confidence score would give back points even though you have a median. That is a reason to read the range, not a reason to switch to the $205,000 average. The average looked more precise and described no sale.

Write the three prices down before you open a calculator, including this one. If you cannot name the street, the sold date, and why the house matches, you do not have an ARV yet. You have three numbers. The $102,600 offer is only as good as the decision that $198,000 is the middle of the right list. Rehab you typed low makes the offer look easier. A $36,000 rehab that becomes $50,000 in the field drops the 70 percent offer from $102,600 to $88,600, because rehab is subtracted once, dollar for dollar. The median did not change. The offer did.

Questions

Should I average the three sold prices?

Not for the hand estimate on this page. The average of $176,000, $198,000, and $241,000 is $205,000. The median is $198,000. The $7,000 gap is the high sale pulling the average up. The middle sale does not move.

Is this median the report ARV?

No. The report weights recency, distance, and similarity, and it adjusts bedrooms and bathrooms before the median. This page is the median of prices you already trust. A full analysis is the weighted figure.

run a full analysis free when the median you can do by hand is not enough. A full analysis weights the comps, adjusts beds and baths, and prices the repair tiers. The sample figures there are invented, the same way every dollar on this page is invented.