ARV and Comps: The Full Method, Fudge-Factor Chart, and Worked Examples

TLDR

Find three to five sold houses that belong in the same neighborhood and match the house you’re creating. Normalize their prices, make only a supported adjustment, check for deal killers, and try to disprove your number. Never fudge the neighborhood to make the deal work.

Table of Contents

• What ARV Is Measuring
• Three Ways to Think About Value
• What Makes a Good Comp
  • Date sold: use the market you’re buying into
  • Proximity: the actual neighborhood
  • Features: compare the house you’ll actually have
• My Five-Step Comping Process
• 1. Identify three to five good sold comps
• 2. Normalize the prices
• 3. Make a supported adjustment when the match isn’t exact
  • The Full Fudge-Factor Chart
  • Check the Data Before You Trust the Average
• 4. Check for Dead-on-Arrival Conditions
• 5. Try to Disprove the ARV
• Turn the ARV Into an Offer

To figure out what a house will be worth after I fix it, I start with houses like it that have actually sold. Then I work through what makes those houses comparable, what differences I can account for, and what could make my number wrong.

The math is easy. Choosing the right houses is where people get in trouble.

I’ve had more than a hundred appraisals done, and I still get a broker’s price opinion to check my work when I need it. Being careful doesn’t mean you can never get confident enough to make an offer. It means your confidence comes from the houses and the data, not how badly you want the deal.

What ARV Is Measuring

After-repair value is the value of the house in the condition I’m planning to create.

That last part matters. If I’m buying a beat-up house and renovating it into a nice house, I need to compare it with nice houses. A run-down sale doesn’t tell me what the renovated version should bring.

I use the scale of livability to explain this:

The livability line is a useful way to think about condition. It isn’t a promise that a bank will approve any particular house. The point is to know where I’m buying on that scale and where I’m trying to take it.

ARV is also a range. If good evidence puts a house in the low $200,000s, a calculation that spits out $207,050 doesn’t mean I know the sale price to the dollar.

My scope, the quality of the work, and how I present the house will help determine where it lands in that supported range. They don’t give me permission to use the value of a different neighborhood.

Three Ways to Think About Value

There are three appraisal approaches to understand:

Think of $40,000 of land plus $200,000 to build a house. That’s the idea behind the cost approach. The comparable-sales approach asks what similar finished houses actually sell for.

For the houses I’m buying and fixing, I want to know what a buyer is paying for the finished product. That’s why I’m focused on comps.

What Makes a Good Comp

I check three things: date sold, proximity, and features. All three have to make sense.

Date sold: use the market you’re buying into

Start with recent sold houses. Six months is a useful first pass. In a fast-moving market, I may want a shorter window. In a slower, stable area, I may go back twelve months and sometimes eighteen if that’s what it takes to find good comparisons.

But the reason matters. I’m not extending the date because an old high sale gives me the answer I want. I’m asking whether the price still represents the market I’m in.

A house for sale isn’t a sold comp. Somebody can ask whatever they want. Price cuts and the final transaction may tell a different story.

Active listings still help later when I try to disprove my number. They show what my buyer could choose instead. They don’t replace closed-sale evidence.

Proximity: the actual neighborhood

A ZIP code isn’t a neighborhood. A radius on a map isn’t automatically a neighborhood either.

I use three things together:

1. The feel on the ground. Drive and walk it. Watch where the houses, construction era, size, and surroundings change.
2. Physical borders. Major roads, railroads, highways, waterways, and other breaks can separate different markets.
3. Census tracts. These give me a useful starting boundary that I check against the first two.

Sometimes a census tract contains more than one pocket. Sometimes the line I would draw from a map isn’t the line that makes sense when I’m standing there. I need to understand the place.

I saw an out-of-area investor buy a house near Northshore in Chattanooga using Northshore comps. The house was across a neighborhood boundary. You could practically throw a rock from one market into the other, but the prices were very different. That mistake cost tens of thousands of dollars.

If it’s outside the neighborhood, I don’t use it as a comp. There is no proximity fudge factor.

Features: compare the house you’ll actually have

A 1,200-square-foot ranch and a 2,500-square-foot split-level don’t become equal just because I divide both prices by square feet.

• Size: Similar living area. I start around 200 square feet above or below the subject; on larger houses I may widen that.

• Bedrooms and bathrooms: Start with matching counts before considering a limited adjustment.

• Age and construction era: A much newer house may compete with different buyers than an older one.

• Style and layout: Ranch, split-level, two-story, and unusual builds aren’t automatically interchangeable.

• Condition: Compare with the condition I intend to create, not just the condition I’m buying.

• Garage, carport, pool, and lot: Check what the subject has and what the comp has.

• Basement or ADU area: Don’t assume it contributes the same value as the main above-grade living area.

• Street and surroundings: Traffic, commercial frontage, and other nearby conditions can change the comparison.

I want to start with close matches so the math has less explaining to do.

My Five-Step Comping Process

The process is identify, normalize, fudge where supported, check DOAs, and disprove.

I may catch a deal killer during a quick first look and stop right there. There’s no prize for completing a spreadsheet on a house I already know I don’t want.

1. Identify three to five good sold comps

Use the three tests above. Look at the photos. Read enough of the listing and property record to know what actually sold.

The software suggesting a house doesn’t make it a good comp. I’ve had suggested properties with different bedroom and bathroom counts. Others weren’t renovated. I had to remove them before doing the math.

Three very strong matches can be more useful than five weak ones. If I can’t find enough sound evidence, I don’t turn the closest-looking bad evidence into good evidence by averaging it.

2. Normalize the prices

For each valid comp:

Sold price ÷ comparable square feet = price per square foot

Here’s how the math works for two sold houses:

For the 1,100-square-foot subject in that example:

Using the unrounded average:

[($300,000 ÷ 1,000) + ($325,000 ÷ 1,200)] ÷ 2 × 1,100 = $313,958.33, or roughly $314,000.

Those two rows illustrate the calculation. For an actual purchase, I’m still looking for the three to five good comps discussed above. And I still need to check whether that result fits the neighborhood’s range.

Don’t use a small house’s high price per foot on a much larger house without thinking. Kitchens and bathrooms account for a lot of the value. As houses get larger, more of their square footage is additional rooms and space. The price per foot can change.

3. Make a supported adjustment when the match isn’t exact

This is what I call a fudge factor. It isn’t permission to make the numbers up.

The rules are simple:

• Start by looking for comps that don’t need an adjustment.
• Fudge one variable at a time. If I’m already stretching the date, I don’t also want to fix three feature differences.
• Make the adjustment because the comparison needs it, not because my offer needs a bigger ARV.
• Never fudge proximity.

A comp that needs a bedroom change, a bathroom change, a date change, and a different neighborhood isn’t a close comp. I’m doing too much work to make it fit.

The Full Fudge-Factor Chart

These are the working adjustments I use to think through a comparison. They aren’t a universal appraisal schedule, and they don’t override what good local sales show.

The dollar figures are adjustment amounts, not instructions to always subtract. The direction depends on which house has the better feature.

The chart divides the examples into under and over $500,000. That isn’t a claim that a market flips from one adjustment to another at one exact dollar. It’s a working guide, and the closer the comparison, the less I have to lean on it.

Basement and ADU credit means I don’t automatically count that space like the main house. It doesn’t mean every basement or ADU is worth half. If I can avoid comparing a house with that extra space to one without it, I will.

Add or subtract from the comp first

I’m adjusting the comp to make it more like my subject.

If the comp is better, reduce the comp’s price. If the subject is better, increase the comp’s price.

Then divide that adjusted price by the comp’s square feet.

Take a comp that sold for $300,000 and had 1,000 square feet, but had one fewer bedroom than the subject. Using the under-$500,000 bedroom adjustment:

In another comp search, I looked at a $700,000 sale with three bedrooms while the subject had two. Using the $20,000 bedroom adjustment, the math was $700,000 − $20,000 = $680,000. Divide that by the comp’s 1,440 square feet and you get about $472.22 per square foot. But I still rejected that house as a comp because the size and basement details didn’t fit. Being able to calculate an adjustment doesn’t make a bad comp good.

I don’t subtract $10,000 from a dollars-per-square-foot figure. I don’t apply it twice. Adjust the comp’s total price, then normalize.

What about an older sale?

When I use an older comp, I check whether prices have changed since it sold. For a $369,000 sale and a 1.2% change over the comparison period:

$369,000 × 1.012 = $373,428

At 1,016 square feet, that is about $367.55 per square foot.

That’s how the adjustment works when the market evidence supports it. A broad trend chart is context; it isn’t proof that this particular house rose by precisely that amount.

It also isn’t the same as guessing what prices will do after I buy. Adjusting an old sale toward today’s market and betting on tomorrow’s appreciation are different decisions.

Check the Data Before You Trust the Average

An out-of-place sale deserves a look. It doesn’t deserve to be deleted just because it’s inconvenient.

I found a house that appeared to have an unusually low price per foot. Looking closer, the listing seemed to include a separate bonus space in the house’s square footage. That larger denominator made the price per foot look lower.

Other things I’ve run into:

• Square footage is surprisingly large or a suspiciously round number: Was it measured, estimated by the owner, or combined with other space?

• Several houses sold together: Was a package price split across the properties in a way that doesn’t reflect each one’s value?

• A house and another parcel sold together: Did the record put the full amount on just the house?

• A very low sale: Was it a related-party transfer or a different kind of transaction?

• A cash purchase below the other sales: Was it a discounted investor purchase, and what was the property’s actual condition?

• An automated estimate appears where a sale price should be: Do I actually have the closed price, or just an estimate or last list price?

Where sold-price data isn’t available on the free sites, I need another reliable source for the actual sale information. An MLS search through someone with access or a data service may be part of that work. I don’t quietly substitute the last asking price.

4. Check for Dead-on-Arrival Conditions

DOA means the house has a problem that puts it outside what I want to take on. Some issues mean I need a much better deal. Some mean I walk.

• Easements, setbacks, or zoning: The property may not support the use or work I have in mind.

• Flood-zone outlier: The subject may carry risk and costs that the other neighborhood sales don’t.

• Wrong side of a neighborhood boundary: My supposed comps belong to a different market.

• Unusual architecture: There may be no sound comparison for the house.

• Irregular, tiny, or steep lot: The buyer may not be getting the same usable property as the comps.

• Awkward layout or low ceilings: Fixing finishes doesn’t necessarily fix the reason buyers don’t want it.

• Busy street, parking problems, or commercial frontage: The setting changes what I’m selling.

• Poor neighboring conditions: The house may be fine while the immediate surroundings hurt demand.

• Power lines, railroads, retention ponds, or highway noise: A map or summer visit may not reveal the whole issue.

• Property stigma: Buyers may react to events associated with the house.

• Lack of good comps: I can’t back up the number I need to buy safely.

Flood risk is a good example of why I compare the actual market. A house with unusual flood exposure in an otherwise unaffected area isn’t the same purchase as houses in a market where that condition is normal and reflected in the sales.

In my area, I generally avoid the outliers. I’m not interested in becoming the expert on every problem just to make one deal work.

And look beyond the photos. I bought a house that looked peaceful next to some woods. When winter came and the leaves fell, the highway behind it became a lot more obvious. That’s the kind of thing that belongs in the decision before I buy.

5. Try to Disprove the ARV

Once I have a number, I go looking for reasons it could be wrong.

• Are there similar sold houses below it?
• Did I exclude one without a good reason?
• Is there a better house sitting for sale below the price I’m counting on?
• Am I above the neighborhood’s proven range?
• Did I use a house from across a boundary because it helped my average?

This is where confirmation bias gets expensive. You work hard to find a lead, and now you want it to be a deal. Suddenly every low sale has an excuse and every high sale seems relevant.

The other trap is what I call the wave. If today’s supported ARV is $300,000, I don’t change it to $320,000 because I expect the market to rise while I’m renovating. If the market does rise, great. I don’t need to spend that possible gain in advance.

When I’m not sure, I get another set of eyes. I still ask my wife, who’s a real estate broker, for a BPO to check my work. An outside opinion is especially useful when I catch myself arguing for a deal instead of assessing it.

Turn the ARV Into an Offer

Once I have a supported range, I estimate what it will cost to get the subject into that condition. I use the scope of work and the Fliporithm work to think through the repairs, then put those figures into the deal calculation.

ARV is one input. It isn’t the offer. I still need the purchase, renovation, financing, holding, and sale numbers to work together.

I want enough detail to see why the number makes sense, and enough discipline to walk away when it doesn’t. I don’t need a fake level of precision that keeps me from ever making a decision.

Before I offer, I should be able to explain which houses I used, why they belong in the comparison, every adjustment I made, and what evidence could prove me wrong. If I can’t do that, I have more work to do.