The 70% rule in house flipping
The 70% rule is the most repeated formula in house flipping:
Maximum offer = (ARV × 0.70) − repair costs
On a property with an after repair value of $300,000 needing $50,000 of work:
$300,000 × 0.70 = $210,000
$210,000 − $50,000 = $160,000 maximum offer
It exists to be done in your head while standing outside a house. That is its entire purpose, and most criticism of it is really a complaint that it is not something it never claimed to be.
What the 30% actually covers#
The number people miss is that the 30% is not profit. Profit is the last item in a list, after everything else has taken its share.
| Item | Typical share of ARV |
|---|---|
| Purchase closing costs | 1–2% |
| Holding costs — interest, taxes, insurance, utilities | 4–8% |
| Selling costs — commission, closing, concessions | 6–8% |
| Financing points and fees | 1–3% |
| Contingency for overruns | 3–5% |
| Profit | 6–12% |
On the $300,000 example, 30% is $90,000. Selling costs alone at 7% take $21,000. Eight months of holding on a $200,000 loan takes a substantial further slice. The profit at the end is a good deal smaller than the $90,000 the formula appears to reserve.
This is why the rule feels conservative and often is not. The 30% looks generous until it is itemised.
Run your own numbers in the 70% rule calculator — it gives the maximum offer and stress tests both estimates, so you can see what happens when the ARV is light and the repairs run over.
Where the percentage should actually sit#
70 is a convention, not a constant. The right number depends on your costs.
| Situation | Suggested | Why |
|---|---|---|
| First flip, hard money | 65–70% | Estimates are least reliable and money is most expensive |
| Experienced, private capital | 75% | Lower financing cost, better repair estimates |
| Very competitive market | 75–80% | Thinner margins or no deals at all |
| Slow market, long days on market | 65% | Holding costs run longer |
| Heavy renovation, permits, structural | 60–65% | Longer timeline, higher overrun risk |
| Low-value property under $100k | 55–65% | Fixed costs consume a larger share |
| Light cosmetic, fast turnaround | 75% | Little can go wrong and it goes quickly |
Experienced buyers consistently outbid newer ones and it is not courage. Their financing is cheaper, their repair estimates are tighter, and their contractors show up. All three let them work at a higher percentage on the same property and still make money.
Why it fails on cheap properties#
The single most common misapplication.
Fixed costs do not scale down. A permit costs what it costs. A contractor has a minimum charge. Closing costs, utilities during renovation, a dumpster, an inspection — broadly the same on a $60,000 house as a $300,000 one.
Worked through on a low-value property with an $85,000 ARV and $25,000 of work:
$85,000 × 0.70 = $59,500
$59,500 − $25,000 = $34,500 maximum offer
The 30% here is $25,500. Selling costs at 7% take $5,950. Closing on both ends takes perhaps $4,000. Six months of holding on a small hard money loan with points takes several thousand more. Add a dumpster, permits and utilities and the profit is close to nothing before anything goes wrong.
On low-value property the percentage has to come down, not stay the same. Many experienced buyers use a minimum dollar profit instead of a percentage below a certain price point, precisely because the ratio stops working.
Why it fails on long projects#
Holding costs scale with time. The 30% does not.
| Project length | Holding cost on a $200k loan | Share of a $90k buffer |
|---|---|---|
| 4 months | Modest | Small |
| 8 months | Double the above | Meaningful |
| 14 months | Triple or more, plus extension fees | Substantial |
| 20+ months | Compounding, and the market has moved | Often the whole margin |
And long projects have a second problem. The ARV was estimated against a market that existed when you bought. A twenty-month renovation finishes into a different one, and the comps you priced from are two years stale.
Structural work, foundation repair, additions and anything needing multiple permits belong at a lower percentage — not because the arithmetic changes but because the time does.
The rule is only as good as two estimates#
This is the real weakness, and it is not in the formula.
The 70% rule takes two inputs and both are estimates:
ARV, which is a forecast of a future sale price built from comparable sales. Get it 10% wrong and every downstream number moves.
Repair costs, which on a property you may not have been able to inspect are partly a guess.
The errors compound. On the $300,000 example, a 10% optimistic ARV and a 20% repair overrun:
| As estimated | As it turned out | |
|---|---|---|
| ARV | $300,000 | $270,000 |
| Repairs | $50,000 | $60,000 |
| Maximum offer the rule gave | $160,000 | — |
| What the rule would have given | — | $129,000 |
You paid $160,000 for a property the rule says was worth $129,000. The arithmetic was never wrong. The inputs were, and the confident output disguised it.
This is why the rule is a filter, not a decision. It tells you whether to spend an hour on a property. The hour is where the estimates get tested.
The 70% rule and distressed acquisitions#
Three complications when the property comes from a foreclosure, an auction or a tax sale.
You often cannot inspect. At a foreclosure auction there is no interior access, so the repair estimate carries far more error than usual and the contingency inside the 30% has to absorb it. Buyers working at 70% on properties they have never entered are applying a rule built for the case where you can see what you are buying.
Comparable sales may themselves be distressed. In a neighbourhood with concentrated foreclosures the recent sales are foreclosure sales, which depress the ARV below the market you will actually sell into. Filtering them out is necessary and it shrinks an already small comp set.
In redemption states you may not keep the property. Across 326 tracked Minnesota redemption windows, 33.4% ended with the owner redeeming. And the distribution is exactly wrong for someone using this rule:
| Winning bid vs assessed value | Owner redeemed | n |
|---|---|---|
| Under 50% | 58.1% | 31 |
| 50–80% | 44.2% | 77 |
| 80% or more | 20.0% | 50 |
A purchase at 70% of ARV is often well under half of assessed value — which is the band where more than half of Minnesota windows were redeemed out from under the buyer. The rule is optimising for a discount that, in a redemption state, is itself the strongest predictor that you will not keep the property.
The certificate holder receives their money back with statutory interest, which is a return, but it is not a flip.
A full worked example, both ways#
The rule takes seconds. Here is what the hour behind it looks like, on the same property.
The property. Three-bed two-bath, 1,450 sq ft, needing kitchen, one bathroom, flooring and roof. ARV established at $325,000 from three renovated comps at $224–$226 per foot. Repairs quoted at $58,000.
The rule:
$325,000 × 0.70 = $227,500
$227,500 − $58,000 = $169,500 maximum offer
The pro forma, assuming an eight-month project, hard money at 75% loan to cost, two points, and a purchase at $169,500:
| Line | Amount |
|---|---|
| Purchase price | $169,500 |
| Purchase closing costs | $3,400 |
| Renovation | $58,000 |
| Loan points, 2 on ~$185,000 | $3,700 |
| Lender and valuation fees | $1,800 |
| Interest, 8 months | $14,800 |
| Property tax and insurance, 8 months | $3,900 |
| Utilities during renovation | $1,200 |
| Draw inspection fees, 4 draws | $900 |
| Total in | $257,200 |
| Sale price | $325,000 |
| Agent commission at 5% | $16,250 |
| Seller closing costs and concessions | $4,900 |
| Net proceeds | $303,850 |
| Profit | $46,650 |
Cash actually invested, rather than total cost: roughly $46,000 down plus working capital between draws, call it $60,000 at peak.
Return on cash: about 78% over eight months. That is the number that compares this against another deal. The 70% rule never produced it.
Now stress it. A 15% renovation overrun and two extra months:
| Change | Impact |
|---|---|
| Renovation $58,000 → $66,700 | −$8,700 |
| Two more months of interest, tax, insurance, utilities | −$5,000 |
| Extension fee, 1 point | −$1,850 |
| Revised profit | $31,100 |
Still a deal. Now stress the ARV instead — the property sells for $305,000 rather than $325,000:
| Change | Impact |
|---|---|
| Sale price $325,000 → $305,000 | −$20,000 |
| Commission and closing reduce slightly | +$1,050 |
| Revised profit | $27,700 |
And both together: about $12,000, on $60,000 of cash tied up for ten months. The deal has not failed, but it has stopped being worth doing, and the 70% rule gave no warning at any point because both inputs moved after the offer.
That is the argument for stressing the two estimates rather than trusting the buffer. The 30% is designed to absorb one thing going wrong. Two is what usually happens.
Where the rule came from, and what it assumes#
The 70% convention hardened during a period of relatively cheap money, fast sales and cosmetic renovations. Each of those assumptions is doing work.
It assumes a short hold. Four to six months. Holding costs are inside the 30% and they scale with time while the buffer does not.
It assumes cosmetic work. Paint, flooring, kitchen, bath. Predictable scope, few permits, low overrun risk. Structural work breaks it because the distribution of outcomes is wider.
It assumes a functioning resale market. A property that sells in three weeks and one that sits for four months use the same formula and produce very different results.
It assumes you can inspect. The contingency inside the 30% is sized for the surprises you find in a house you walked through, not for the ones in a house you bid on from the kerb.
And it assumes you keep the property, which in about half the states is not guaranteed after an auction purchase.
None of those assumptions is unreasonable. They are just assumptions, and the rule does not announce them.
The alternatives, briefly#
Fixed dollar profit. Decide the minimum profit that makes a project worth doing and work backwards. Better than a percentage on low-value property, where a good percentage can still be a bad amount of money.
Full pro forma. Every line item costed — purchase, closing both ways, financing, holding by month, renovation by trade, selling. Slower, and it is what the 70% rule is a shortcut to. Do this on anything you are seriously considering.
Return on cash. Profit divided by cash invested, annualised over the project. This is the number that actually compares two deals of different sizes and lengths, and the 70% rule tells you nothing about it.
Use the rule to decide what to analyse. Use a pro forma to decide what to buy.
The short version#
- Maximum offer = ARV × 0.70 − repairs. Thirty seconds, in your head.
- The 30% is not profit. It is closing, holding, selling, financing, contingency, then profit.
- Set the percentage from your own costs — lower for a first flip, cheap property, long project or slow market.
- It is only as good as the ARV and the repair estimate. Both are estimates.
- On a property you cannot inspect, widen the contingency rather than the percentage.
- In a redemption state, a deep discount is the profile most likely to be redeemed.
- Filter with the rule. Decide with a pro forma.