How to Calculate ARV Like a Pro

TLDR

A good comp has three things: matching features, recent sold date, and same neighborhood. You can fudge features and date a little. Never fudge proximity. Forecasting a speculative ARV in a flat market is the fastest way to lose money.

Table of Contents

• What a Comp Actually Is
• The Three Factors That Define a Good Comp
• Features: What Counts, What Doesn’t
• Date Sold: 6 Months, Sometimes 12
• Proximity: The One You Can’t Fudge
• The Fudge Factor
• Common Mistakes That Kill Deals
• Dead on Arrival Conditions
• The Five-Step Process
• FAQ
• Related

What a Comp Actually Is

A comp is a comparable property. When you run comps, you’re finding what your house will sell for after you’ve fixed it up. That number is the after repair value.

Real estate agents call this a BPO (broker’s price opinion). Appraisers call it a CMA (comparative market analysis). Same thing.

You comp the renovated version of the property, not the current version. You’re buying to renovate. You need to know what it’s worth after the work, not before.

Software like Zillow, Redfin, Trulia, and Realtor.com all offer free comp tools. They’ll give you suggested comparables, but “suggested” is not the same as “good.” You need to filter them yourself.

Why Bad Comps Will Kill You

I did a house years ago where I tore the roof off, added a second story, built a three-car garage in the back. Listed at $795,000 based on comps I ran before buying.

Sold for $667,500 after months of price cuts. Six-figure loss. The worst deal I’ve ever done.

The thing was dead before it started. I broke some of the comp rules I’m about to walk through. Wrong comps meant a wrong ARV, which meant the project had no margin of safety from day one.

Forecasting the wrong ARV is the most expensive mistake you can make in this business, and it happens before you ever close.

The Three Factors That Define a Good Comp

Three things matter:

1. Features: size, specs, style, condition
2. Date sold: how recently it actually closed
3. Proximity: same neighborhood, not just same zip code

Miss on any one of these and the comp is noise. Miss on proximity and you’re working with fiction.

Features: What Counts, What Doesn’t

Size

Target comps within 200 square feet plus or minus of your subject property. For houses over 1,800 square feet, you can stretch to 400 plus or minus.

Why 200? That’s what appraisers use. And your end buyer is almost certainly getting a mortgage, which means the bank will send an appraiser. Matching appraiser standards protects your sale.

Basements and ADUs don’t count the same as above-grade square footage. Most appraisers count them at zero or 50 percent. Don’t lump that square footage into the rest when you calculate price per square foot.

Specs

These have to match:

• Bedrooms
• Bathrooms
• Garage vs carport vs none
• Is it on a busy road or commercial property

Style

Not all styles are the same. A craftsman isn’t a mid-century modern. A bungalow isn’t a split-level. A two-story isn’t equivalent to a one-story ranch even at the same square footage.

Finishes

Masonry vs wood siding. Different exterior finishes affect value.

Generation

A 1950s house isn’t the same as a 2000s house. Comps should be similar age.

Lot Size

Similar lot size. A house on a half acre isn’t comparable to one on a tenth of an acre, even if everything else matches.

Condition

This is the one most people get wrong. You’re renovating the house. You need to comp against other renovated houses, not beat-up ones. The comp is what your finished product competes with on the market.

Date Sold: 6 Months, Sometimes 12

Target six months or less for sold comps. In a volatile market (prices moving fast), cut it tighter than six months.

Not for sale. Sold. Anyone can list a house at any price. The list price is fiction until someone actually pays it.

Fudging Date Sold

If you have to stretch, go to 12 months. Sometimes 18 in stable markets. Never further.

When you stretch beyond six months, adjust for market movement. Check the zip-code-level appreciation rate on Zillow’s home value data. If the market’s up 5.5% year-over-year but the specific area you’re buying in is only up 1.2%, use the 1.2% number.

Take the sale price, multiply by the adjustment (like 1.012 for 1.2%), divide by square footage. That gives you the price per square foot adjusted to current market.

Pro Tip

Only fudge one variable at a time. If you’re stretching the date, don’t also stretch the features. You compound error when you fudge multiple dimensions on the same comp.

Proximity: The One You Can’t Fudge

This is the biggest mistake I see. People find a house that looks like a deal on paper, but they’re comping it against houses in a different neighborhood.

I’ve seen an out-of-state investor buy a house for $100K that was across the tracks from a $350K neighborhood. They thought they had a steal. The house wasn’t worth anywhere near $350K. Wrong side of a major road they couldn’t see on the map.

Three ways to verify the neighborhood:

Drive and Walk

Walk the neighborhood. Drive it. You start to feel when things change. House styles shift. Ages differ. Sizes differ. Obvious boundaries appear.

Look for Major Boundaries

Big roads. Railroad tracks. Waterways like rivers and creeks. Parks. These physically divide neighborhoods even when the map doesn’t draw a line.

Census Tracts

This is the one most people don’t know about. Census tracts are the official geographic divisions the government uses. Most real estate software uses them internally to define neighborhoods.

Pull the census tract interactive map for your target property. The tract boundaries often match the real neighborhood boundaries better than anything else on paper.

Combine all three: drive the area, identify major boundaries, confirm with census tracts. That’s how you get proximity right.

Zip codes don’t work. There can be three or four neighborhoods inside a single zip code. One might be fine to buy in; another might be a no-go. Zip-code comps will get you in trouble.

The Fudge Factor

You won’t always find five perfect comps. When you have to stretch, here’s how:

Feature Adjustments

For houses under $500K, use roughly $10,000 per bedroom. Over $500K, use $20,000 per bedroom.

A 3-bed comp in a 2-bed market: subtract the bedroom adjustment before calculating price per square foot.

If the comp sold for $700,000 and has a bedroom your subject doesn’t, use $680,000 as the adjusted number. Then divide by square footage to get price per foot.

Same logic for garage vs no garage, pool vs no pool, busy road vs quiet street. Adjust before calculating price per foot.

Watching for Out-of-Place Comps

Sometimes a comp’s price per foot looks wildly different from others in the same neighborhood. Before trusting or discarding it, check:

• False square footage. ADU or basement footage counted into total livable. Skews the price-per-foot number low.
• Round numbers. Square footage listed at exactly 1,200 or 1,400 usually means the seller quoted it, not a measurement.
• Package deals. Someone bought five houses for $1M and recorded each at $200K regardless of real value.
• Related-party sales. Family deals. Gifts disguised as sales. Way under market.
• Cash investor deals. Someone got a great off-market price. That’s a real comp but know what it represents.

Also watch for oddbirds: properties with unusual architecture, irregular lots, weird layouts, infrastructure issues (power lines, ponds, railroads, highway noise), rough neighbors on that specific street, or scene-of-a-crime stigma.

Common Mistakes That Kill Deals

The Wave (Speculative ARV)

The market’s going up 10% a year. You figure the renovation will take eight months. You use a projected future ARV instead of today’s ARV.

That’s the wave. You’re surfing appreciation.

Waves break. I see investors use speculative ARV all the time, and when the market flattens or drops, they get crushed. Use today’s ARV, always. If the market goes up while you renovate, that’s gravy. Don’t count on it.

Confirmation Bias

You did a lot of work to find this deal. You want it to work. So you throw out comps that would lower your ARV. You keep the ones that support your number.

Fight this. If you can’t trust yourself, get a third party. Ask a real estate agent to review your comps. Ask another investor. Pressure-test your own number.

Wrong Side of the Tracks

Don’t accept comps from across a major road. Don’t accept comps from a different census tract. Proximity is the one rule that can’t bend.

I would personally never invest out of state because I know how easy it is to miss borders on a map that are obvious from the street.

Dead on Arrival Conditions

These are walk-away conditions. I don’t buy houses with them no matter how good the deal looks on paper.

Flood Zone

Not if the whole neighborhood is in a flood zone (like coastal Florida). That’s priced in. The problem is a house in a flood zone when the surrounding neighborhood isn’t.

Buyer needs flood insurance, usually $100-$200 a month. That comes out of their mortgage underwriting capacity, which lowers what they can qualify for, which lowers your sale price. The bank knows this too. Every buyer will get hit with it.

Check FEMA maps for the 100-year flood zone before you make an offer.

Parcel Issues

• Easements cutting through usable yard
• Setback violations (house built inside the required setback)
• zoning mismatch (duplex in single-family zoning, for example)

I walk away from these. Even if I know the rules, enforcement is subject to municipality opinion. I don’t want to gamble.

Low Ceilings

Under 7.5 feet. Too hard to comp, too hard to sell.

The exception: a long-term rental where I’m never planning to sell. Even then, only if the deal is great. For flips, never.

Weird Builds

Octagon houses. Oddly eclectic layouts. Anything I can’t clearly compare to normal houses nearby.

Warren Buffett: never invest in a business you can’t understand. I apply the same logic. There are other houses I can buy that I fully understand.

DOA conditions aren’t negotiable. These are deals where the math looks great but the exit is broken. Walking away from a “great deal” with a DOA condition is easier than walking away from a great deal after you’ve closed.

The Five-Step Process

Here’s how I actually comp a property.

1. Quick Comp

Look up the subject on Zillow, Redfin, Trulia, or Realtor.com. Scroll the suggested comps. Glance at price per square foot. Get a ballpark.

If the ballpark says 200 a square foot on a 1,025 square foot house, I’m figuring low 200s. Good enough to start a conversation with a seller.

2. Quick Google Map View

Drop into Street View around the subject. Make sure nothing is wildly out of place. This is a time-saver: if the neighborhood is rough and I wasn’t going to buy anyway, I stop here.

3. DOA Check

Flood zone, parcels, ceilings, weird build. If any of these hit, stop. Walk away.

4. Three to Five Good Comps With Fudging

Pull sold comps within six months, narrow radius around the subject, matching beds and baths, within 200 square feet of the subject.

Remove comps that don’t match condition. Apply fudge factor on date or features if needed.

Calculate price per square foot for each. Average them. Multiply by the subject’s square footage. That’s your ARV estimate.

5. Neighborhood Scan for Out-of-Place Sales

Pull up all sales in the last 12 months. Look for prices that seem wrong. If everything looks normal, your number is probably right. If something’s wildly off, check the common out-of-place reasons (package deal, related party, false square footage, oddbird).

The Range, Not the Number

Here’s the mindset shift: ARV is a range, not a single number. You use a scale of livability to understand where your renovation lands:

• Bombed-out houses (lowest)
• Barely bankable (livable but outdated)
• Range of comps (renovated, what you’re shooting for)

You’re buying in bombed-out or barely bankable. You’re selling into the range of comps. That range has a floor and a ceiling. Your job is to push toward the ceiling through scope, renovation quality, and digital introduction.

A good listing, photography, and staging can add 5-10% to the sale price. That’s why the range matters more than the exact number.

Don’t freak out about nailing an exact ARV. Don’t get analysis paralysis. Find a defensible range. Underwrite conservatively inside that range. Move.

FAQ

What if my state is a non-disclosure state?

Some states don’t require sold prices to be public. You’ll see list prices but not sale prices on free platforms. In those states, you need mls access through a real estate agent or a paid platform. Free comp tools won’t give you the real numbers.

How many comps do I need?

Target three to five good ones. If three are spot-on matches, three is enough. More is better when features vary.

When does the 400-square-foot tolerance apply?

Houses over 1,800 square feet. Below that, stay within 200 square feet. The bigger the house, the more square footage variance matters in absolute dollars but less in percentage.

What’s the appraiser trick?

If you think the appraisal might come in low, you want to talk to the appraiser at the property. Leave the key out of the lockbox on appraisal day. They’ll call when they can’t get in. Bring the key, walk them through, and highlight what you’ve done and what comps you think support the price. Not always necessary. Useful when the deal is tight.

I’m just starting out. What’s the one thing I have to get right?

Proximity. Wrong neighborhood is the deal-killer that hides the longest. Spend time walking your target neighborhoods until you know the borders by feel. Then verify with census tracts.