Flipping US auction cars: the stats, the playbook, and how carpybara helps you find the good ones
More than half of Copart and IAAI lots end up flipped. Here are the numbers — average margins, time-to-flip, the trims that consistently beat the curve — and the tools we built to make finding them less of a guessing game.
If you've watched a few hundred Copart or IAAI auctions, you've already noticed it: the same VINs come back. A 2021 Tesla Model 3 that sold in Newark for $19,400 in February shows up in a New Jersey body shop's lot in April, freshly polished, listed for $26,000. That gap — minus shipping, repair, fees, and a few weeks of the flipper's time — is the entire economics of the US auction-to-resale market, and it's bigger than most buyers realise.
Below is what the numbers actually look like, plus the tactical bits we've baked into carpybara to take some of the guesswork out.
How big is the flip market, really?
US auto auctions move roughly 9 million vehicles a year between Copart, IAAI, Manheim, ADESA, and the smaller regional houses. Of those:
- About 45–55% of clean-title sales end up resold within 6 months without ever leaving the country. Most go to small independent dealers; a chunk go to the curbstoning grey market.
- About 30–35% of salvage and rebuilt-title lots get flipped after partial repair. These are where the big-margin / big-risk plays live.
- Roughly 8–12% of total volume is exported — Ukraine, Georgia, Lithuania, the UAE, and West Africa being the headline destinations. Most of that is also a flip — just one with a cargo ship in the middle.
That's a $12–14B/year informal market on top of the auctions themselves, and the tooling around it is still mostly Telegram channels, hand-typed spreadsheets, and gut feel.
Where the margin actually sits
We pulled three months of carpybara dealer-completed deals (n=187, January–March 2026) and matched each VIN against its sale price on the destination market. The numbers, by buy-now price band:
| Auction price | Avg landed cost | Avg resale (UA market) | Realistic margin* | Avg days to sell |
|---|---|---|---|---|
| $5K – $12K (older sedans, light damage) | +$3,200 | $14,500 | $1,400 | 41 |
| $12K – $22K (mainstream SUVs, AWD trims) | +$5,100 | $26,800 | $3,200 | 33 |
| $22K – $35K (premium / EV) | +$6,400 | $38,500 | $5,800 | 28 |
| $35K+ (luxury / low-mileage rare trims) | +$8,200 | $52,000+ | $8,000–$14,000 | 19 |
* Realistic margin = expected sale price (p25 of UA market, not the listing median) minus all-in landed cost. Days-to-sell is from delivery to wire transfer.
A few things that fall out of those numbers:
- The mid-tier band has the worst margin-to-effort ratio. $1,400 over six weeks of work, capital tied up in shipping, and one paint flaw away from breaking even.
- Premium / EV is where the consistent money is. Buyers in Ukraine, Georgia, and the EU are still paying near-MSRP for late-model Teslas, BMW iX, Porsche Taycans — and the auction discount on a lightly-damaged Long Range trim is large enough to absorb the import friction.
- Days-to-sell is inversely correlated with margin. Cars that sell fast aren't priced lower — they're picked better.
What separates a $5K-margin flip from a $1K-margin flip
We re-ran the same deals broken out by which lots cleared $5K+ vs which cleared less. The differences weren't in price or trim — they were in the pre-bid checklist:
- VIN history was checked, not assumed. ~70% of high-margin flips had a clean accident history despite being sold under a salvage/secondary title. The flipper knew the title designation came from a single fender bender — not a frame hit.
- Photos were inspected for the cheap-to-fix vs expensive-to-fix split. Front bumper + headlamp + airbag deployment is repairable for $1.8K–$2.5K. Frame bend, B-pillar damage, or undercarriage pierce is not — and they look similar in a wide-angle photo.
- The destination market median was checked against the delivery date, not the auction date. Ukrainian used-car prices move 5–8% per quarter on popular trims. Buying for January's prices and selling into April's market is a different deal than buying for April and selling into July.
- Comparable VIN flips were tracked. If the exact same lot sold three weeks ago, you can see what the prior buyer paid AND what they're now asking — that's a much better signal than "auction average."
Each of those checks is doable manually with a browser tab open to NICB, three insurance-history sites, and auto.ria's filters. It just takes 15–25 minutes per lot, which is why most buyers skip them and lose the margin to the ones who don't.
What we built to make this faster
The whole product is built around shortening that 15–25 minute checklist to under a minute per lot:
- AI-scored buy-now feed — every lot on /buy-now gets a 0–100 score that combines damage classification (cheap vs expensive), title risk, photo completeness, and the destination-market premium. The top decile of scores carries about 70% of our completed-deal margin.
- Live UA market price — every detail page now pulls the median, p25, and p75 across auto.ria, rst.ua, and olx.ua for the exact (make, model, year) combo, so you see the destination-side number next to the bid. We just shipped per-make/per-year pages too: try /car-price-ua/tesla/model-y/2022 or /car-price-ua/bmw/x5/2021 — you can scatter, bar, or line-chart the price-vs-mileage spread, and click any dot to open the underlying listing in a new tab.
- Verified import dealers — instead of finding a Telegram broker on faith, /dealers lists the carpybara network with completed-deal counts, average response time, and customer ratings. Most of our high-margin deals close through dealers who've done 50+ similar lots before.
- VIN check + auto-fetched history — /vincheck pulls the same NICB / NMVTIS / accident records that the careful flippers were checking by hand. We don't gate the basic report.
- Customs + delivery calculator — /customs and /delivery bake in the actual line items (auction fee, broker, ocean freight, customs, VAT, excise, last-mile) so the "all-in landed cost" column above stops being a surprise.
None of this turns a bad lot into a good one. It just lowers the cost of the homework that separates the two.
The honest caveats
Flipping US auction cars is a real business with real downside:
- About 1 in 8 deals lose money. Hidden frame damage, parts shortages, auction fee misquotes, market shifts mid-shipment. The dealers who survive treat the median margin, not the mean — outliers cut both ways.
- Capital is tied up for 4–10 weeks. If you're financing, the interest eats a non-trivial chunk of mid-tier deals.
- Regulatory risk is non-zero. Ukraine's customs schedule on EVs has flipped twice in the last 18 months. Last year's playbook isn't this year's.
For a buyer who wants the upside without running their own brokerage, the saner play is browsing /buy-now, shortlisting the top-scored lots, and asking a verified carpybara dealer for a fixed-price quote that includes everything to your driveway.
For a dealer who wants to find more $5K+ flips faster: open the buy-now feed sorted by AI score, cross-reference our UA market price page for the model-year-mileage band, and treat the bottom-of-page "comparable VIN flips" panel as your edge over everyone still doing the homework by hand.
The math hasn't changed — flipping US auction cars is still a margin business — but the tools are finally catching up with the question every buyer is actually asking: "is this lot worth bidding on, right now, given what the destination market looks like today?" That's the question we're trying to answer in under a minute.