I've watched a lot of new investors lose money on deals that looked great in a spreadsheet. Almost none of them lost it on the purchase price.
They lost it on four numbers that are boring, unglamorous, and systematically underestimated.
The four killers
1. Rehab budget
The number one deal killer, by a wide margin.
People budget the visible work — flooring, paint, cabinets, fixtures — and leave out everything else. Then the drywall comes off.
What routinely gets missed:
- Permits and plan review. Real money and real weeks, especially for anything structural or electrical.
- The surprise behind the wall. Knob-and-tube, cast iron drain lines, an unpermitted addition, termite damage, a roof deck that's gone.
- HVAC. In Phoenix, a failing unit isn't optional. Full replacement runs real money and it's not a line you can cut.
- The second dumpster. There is always a second dumpster.
- Landscaping and exterior. Buyers decide in the first eight seconds, from the curb.
- Holding-period damage. Vacant houses attract problems. Copper theft, vandalism, squatters.
Rule I use: build a 15–20% contingency into the rehab budget, and treat it as spent. If you finish under, that's upside. If you budget to the bid with no cushion, the first surprise eats your margin.
2. Holding costs
Everyone remembers the mortgage payment. Fewer people add up everything else.
Monthly, while you own it: loan payment or hard money interest, property taxes, vacant-property insurance (more expensive than a standard policy, and you need the right one), electric and water, lawn and pool service, alarm or monitoring, HOA dues.
Then multiply by a realistic timeline, not an optimistic one. A "three month flip" is almost never three months. Between acquisition, permitting, the work itself, and 30–45 days of escrow, four to nine months is the honest range.
Underwrite to the long end. If you finish early, you made extra money. If you underwrite to three months and it takes seven, four months of holding costs came straight out of your profit.
3. Selling costs
On the way out you'll pay listing commission, buyer-side commission if you're offering it, seller closing costs and title fees, and — reliably — a buyer repair credit after their inspection.
That last one surprises people every time. A buyer's inspector will find something. They always do. Budget for a credit even on a full renovation.
Realistically 7–9% of your sale price leaves on the way out.
4. The ARV itself
The most dangerous number, because it's the one people most want to believe.
Comp discipline:
- Sold comps only. Active listings are asking prices, which are opinions.
- Last 3–6 months. Older data doesn't reflect the current market.
- Same neighborhood — ideally the same subdivision. In Phoenix, values shift sharply across a single arterial road.
- Similar square footage, bed/bath count, and lot. A 4/2 doesn't comp a 3/2.
- Similar finish level to what you're actually delivering, not what you wish you were delivering.
And this: if your ARV is above every sold comp in the neighborhood, your ARV is wrong. You're not going to set a new ceiling for the subdivision. The appraiser won't let you, and a financed buyer's lender won't either.
A worked example
Phoenix single-family, 1,500 sq ft, 3/2. Solid sold comps at $420,000 for renovated properties.
| Line | Amount |
|---|---|
| ARV (conservative, from sold comps) | $420,000 |
| Purchase price | −$255,000 |
| Rehab (incl. 18% contingency) | −$71,000 |
| Holding costs (6 months, all-in) | −$21,000 |
| Financing points & fees | −$9,000 |
| Selling costs (8% of ARV) | −$33,600 |
| Projected profit | $30,400 |
$30,400 on $420,000 of ARV — about 7%. That's a working deal, not a home run. It survives one bad surprise. It doesn't survive two.
Now watch what optimism does. Drop the contingency, assume three months instead of six, and use $440,000 as the ARV because one outlier sold there:
Profit on paper: $71,900. Same house. Same everything. The only thing that changed is which numbers you decided to believe.
Then the electrical panel needs replacing, the timeline runs seven months, and the house appraises at $415,000. The real number lands closer to $18,000 — and that's if nothing else goes wrong.
The deal didn't change. The underwriting did.
What I do differently
Conservative every time, on purpose. Low ARV, high rehab, long timeline. If a deal only works with aggressive assumptions, it's not a deal — it's a bet on everything going right.
County records and permits before I commit. Unpermitted additions, open permits, and lien surprises are cheaper to find in the records than in escrow.
Post-1978 construction only, so lead paint remediation isn't in the scope.
Two exits modeled on every deal — retail sale and rental hold. If the market moves against me mid-project, I need a plan that isn't "panic and cut price."
That's also how I underwrite for investor partners. Nobody's capital should ride on a best-case spreadsheet.
The short version
Most flips don't fail because someone paid $10,000 too much. They fail because rehab ran 40% over, the timeline doubled, and the ARV was a wish.
Pad the rehab. Stretch the timeline. Discount the ARV. If the deal still works, you have a real one.
Want a second set of eyes on a Phoenix deal? Send me the numbers. I'll tell you what I think honestly — including when the answer is walk away.
Questions about your specific situation? Call or text me at (602) 902-8400, or send me a message. I answer these myself.
This post is general information, not legal, tax, or financial advice. Arizona real estate law and lending guidelines change. Confirm anything time-sensitive with your own attorney, CPA, or lender before acting on it.
