Ask ten new investors why a deal went sideways and most will point at the rehab, the market, or the tenant. Look closer and the real fault line usually runs back to one number they set on day one: the after-repair value. ARV is the anchor every other figure hangs off, and when the anchor is soft, the whole analysis drifts — quietly, confidently, and in the wrong direction.
The hard part is that a wrong ARV does not feel wrong. It feels like optimism, and optimism reads as confidence. So the discipline here is not mathematical — it is emotional. It is the willingness to write down the number the comps support instead of the number the deal needs. Here is how to build an ARV you can defend, and how to catch yourself in the act of fooling yourself.
why ARV is where deals die
Every downstream number inherits your ARV error, and the errors compound in your favor right up until closing. Your maximum allowable offer is a function of ARV — the 70% rule multiplies it directly, so a 10% overstatement on a $200,000 ARV inflates your ceiling by roughly $14,000 and you bid that much too high without noticing. Your refinance is a function of ARV too: lenders lend against appraised value, and if your appraisal comes in below the number you underwrote, the cash you expected to pull out simply is not there.
Then there is profit, which is what is left after everyone else is paid. Profit is the thinnest slice of the ARV, which means it is the most sensitive to ARV error. Overstate the value by 8% on a flip and you have not lost 8% of your profit — you may have lost all of it, because the overage came straight out of the one line with no cushion. Optimism does not average out. It stacks, and it stacks against you.
the comp filter
A defensible ARV starts with defensible comparables, and the filter is strict on purpose. The single most important rule: sold comps only. Active and pending listings are asking prices, not agreed prices — they tell you what sellers hope for, not what buyers paid. Build the value off closed sales and nothing else.
- Same bed and bath count. A 3/2 comps against 3/2s. A different room count is a different buyer pool.
- Within about 20% of the square footage. A 1,100 sqft subject does not comp against a 1,600 sqft sale, even next door.
- Within one mile — tighter in dense urban blocks — and never across a boundary line. School-district edges and neighborhood lines move value in ways distance alone does not capture.
- Closed within the last six months. Older sales priced a market that no longer exists.
- Same property type. Single-family against single-family, townhome against townhome. A condo comp will not defend a detached ARV.
If you cannot find three to five sales that clear every filter, that is information, not an obstacle. It usually means the value you are reaching for is not actually supported — and stretching the filter to find it is the first step of fooling yourself.
making adjustments like an appraiser
No comp is identical to your subject, so you adjust — and the direction matters. The rule appraisers use: adjust the comp toward the subject. If the comp is superior, subtract from its sale price; if it is inferior, add. You are asking what that house would have sold for if it had been your house.
Start with condition. A comp that sold fully renovated is superior to a subject you are buying rough — but for ARV you are valuing the subject after its rehab, so a renovated comp is the correct anchor and an un-rehabbed comp needs an upward adjustment to reach finished condition. Then the countable features: an extra bathroom might be worth $8,000–15,000 depending on the market, an extra bedroom more, a garage a few thousand, a pool or a materially larger lot each their own line. Use dollar adjustments the local market actually pays, not national averages.
Two habits keep this honest. Round conservatively — when an adjustment could reasonably be $5,000 or $8,000, take the smaller number. And keep your total adjustments modest relative to the sale price; if you are adjusting a comp by 25% to make it fit, it is not a comp, it is a wish with a footnote.
the failure modes
Fooling yourself is a small set of repeatable moves. Learn to recognize them mid-analysis:
- Using actives and pendings. The most common one. Asking prices drift high, and building an ARV on them bakes seller optimism into your own.
- "The Zestimate says." Automated estimates are a starting point for curiosity, never an underwriting input. They cannot see condition, and condition is the whole game in a rehab.
- Cherry-picking the three highest sales. There are always a few outliers that closed hot. Anchoring to them and ignoring the rest is not comping — it is confirmation bias with a spreadsheet.
- Ignoring days-on-market and concessions. A comp that sat for 120 days and closed with $6,000 in seller concessions did not really sell for its headline price. Read the terms, not just the number.
- Comping a rehabbed subject against un-rehabbed sales. If your comps are all tired houses and yours will be renovated, you are undervaluing — but the reverse, valuing a rough subject against finished comps without discounting for its own rehab, is how people justify overpaying.
pressure-testing the number
Once you have an adjusted range, resist the pull to the top of it. Take the middle, not the ceiling. The top comp exists because something about that sale was unusually favorable, and you do not get to assume every advantage lands in your favor too.
Then sanity-check price per square foot against the neighborhood band. If renovated homes in the area trade at $140–170 per foot and your ARV implies $190, you owe yourself an explanation grounded in something real — a premium finish level, a better lot — or the number is wrong. Price per foot is a coarse tool, but it is excellent at flagging an ARV that has floated free of its market.
The final test is the simplest and the most uncomfortable: if the deal only works at the top comp, the deal does not work. A margin that survives only when every optimistic assumption holds is not a margin — it is a hope. Underwrite to the middle, and let the upside be a bonus you did not need.
This matters because both exits eat ARV error directly. On a flip, the 70% rule sets your ceiling off ARV, and you can run yours in seconds with the 70% rule calculator. On a BRRRR, the refinance typically lands at 75% of appraised value — miss the ARV and the cash-out you were counting on to fund the next deal comes up short. Before you commit, pressure-test the rental math too; the rental property analyzer runs rent, expenses, and debt service so you know the property carries itself even after an honest ARV. Get the anchor right, and every number downstream has a chance. Get it wrong, and no amount of execution downstream can save it.