The best business model? (with a big risk)

Today, I'm digging into Copart (CPRT).

They operate online auctions for wrecked, totaled, and end-of-life vehicles. When your insurer declares your car a total loss, there's a good chance it ends up in a Copart yard, photographed, titled, and sold to one of 100k+ buyers around the world who want it for parts, repair, or resale.

It started as a single salvage yard in California in 1982 who now facilitates the sale of approx. 4 million vehicles per year.

A few things that stood out to me:

• This is a powerful two-sided marketplace. Insurance companies hand Copart a steady river of totaled cars, and and this attracts a significant amount of buyers. More cars pull in more buyers, more buyers lift prices, and higher prices pull in more sellers. That flywheel is hard to replicate.

• Getting paid by both sides of the market without taking inventory is a cash flow holy grail!

• They have strong margins. Gross margin sits around 45% and operating margin around 37%, on a business that is mostly land, tow trucks, and a website.

• They sit on a mountain of cash and basically no debt. Net cash is over $3 billion, and for the first time in awhile they're using some of it to buy back stock.

But it's not all clean and perfect:

• Growth has stalled. Revenue was roughly flat over the trailing year, insurance unit volumes fell, and the stock dropped by more than a third as the market repriced it.

• There is a bear case when it comes to the growing adoption of self-driving. Copart's inventory is, at its core, wrecked cars. If Level 4 and Level 5 autonomy eventually makes driving dramatically safer, you remove the accidents that feed their pipeline. The insurance landscape is likely to change as well which will impact the health and composition of ist customer base.

With that said, Copart has one of the more durable type of moats in business. But it’s incredibly difficult to replicate.

With that said, here are some best practices when trying to start a two-sided marketplace:

• Win the hard side first. Every marketplace has a scarce side and an abundant one. Lock up the bottleneck (for Copart, insurer supply) and the easier side tends to chase the inventory.

• Start narrow. One city, one category. Liquidity is local before it's global.

• Be your own first inventory. Buy and resell on your own account early so you never have empty shelves.

• Embed in the hard side's workflow. Copart wired itself into insurers' claims process. This is a powerful way to help “win the hard side”

It also helps to start with a tool or service that serves both sides of the market independently, and then you can work to try to bring them together.

in past newsletters, I've completely dismissed two-sided marketplaces as being too difficult, but that was short-sighted on my part. They are extremely difficult, but never say never!

With that, I will see you on Tuesday!

What got me nerding out is how boring the core is and how durable that makes it. Nobody wakes up wanting to disrupt the salvage car business. The land is permitted and scarce, the insurance relationships are decades deep, and the operational machine for titling and moving millions of cars has been refined since the Reagan administration. The key takeaway here for me is that owning an unglamorous, hard-to-replicate physical layer underneath a digital marketplace is one of the most defensible positions in business.

With that, I'll see you tomorrow!

Nick

TL;DR

• Copart is the dominant online auction platform for salvage and total-loss vehicles, connecting insurance sellers with a global base of buyers.

• About 85% of revenue is service and fee income from running the auctions, with the rest from vehicles Copart buys and resells on its own account.

• Recent performance stumbled: revenue was roughly flat over the trailing twelve months and insurance volumes softened, partly because the prior year was juiced by hurricane catastrophe cars.

• The financial profile is rare air: ~45% gross margin, ~37% operating margin, over $3 billion of net cash, and no real debt.

• On the stock, CPRT is down about 37% over the past year and modestly lower over five, badly lagging salvage rival RB Global and used-car platform OPENLANE, even though it has compounded around 5x over the past decade.

• Operator takeaway: a scarce physical footprint plus a two-sided marketplace is a moat very few competitors can attack head-on.

The 30,000-Foot View

Here's the basic loop. An insurance company totals a car. Rather than deal with the headache of storing, titling, and selling it, the insurer hands it to Copart. Copart tows it to one of its yards, takes detailed photos (including its 360-degree interior and exterior views), handles the title and paperwork, and lists it on a timed online auction. Buyers from dismantlers to rebuilders to exporters to regular people bid, the car sells, and Copart takes a cut from both sides of the transaction.

The important nuance is how Copart makes money. The vast majority of revenue is service and fee income, meaning Copart never owns most of the cars it sells. It's a consignment and fee model, which is why a business full of trucks and gravel lots throws off software-like margins. A smaller slice, the "purchased vehicle" or product line, is cars Copart buys outright and resells, which carries lower margins because Copart takes on the inventory risk. That mix has stayed remarkably steady, with services around 85% of the total.

What keeps customers coming back is workflow and liquidity. For insurers, Copart isn't a vendor you swap out casually. It's wired into the claims process: total-loss valuation, title transfer, loan payoff, and disposal all run through it. Ripping that out and rebuilding it elsewhere is painful. For buyers, the draw is simple. Copart has the cars. Liquidity attracts liquidity, the same dynamic that makes a stock exchange hard to dislodge, and it's why a salvage auction and the New York Stock Exchange share more DNA than you'd think.

Revenue mix (FY2025, ended July 2025):

• Service / auction fees: ~85%

• Purchased vehicles (product): ~15%

Geographic mix (FY2025):

• United States: ~83%

• International: ~17%

Key Stats

• Market cap: ~$28.5B

• TTM revenue: ~$4.6B

• TTM gross margin: ~45.5%

• TTM net income: ~$1.55B

• 1Y total return: about -37%

• Employees: ~13,800

• Industry: Online vehicle auctions and remarketing (specialty business services)

Company History

• 1982: Willis Johnson opens Copart as a single salvage yard in Vallejo, California.

• 1994: IPO on Nasdaq, funding a roll-up of regional salvage yards across the US.

• 2003 to 2004: Launches VB2 (Virtual Bidding) and moves the entire business to online-only auctions, walking away from live, in-person sales. This is the bet that made the modern company.

• 2007: Enters the UK, kicking off international expansion that later reaches Germany, Brazil, Spain, the Middle East, and beyond.

• 2017: Hurricane Harvey forces a massive catastrophe response. Storm surges become a recurring swing factor in volumes and results.

• 2021: 2-for-1 stock split. A net-cash balance sheet and high-teens growth turn Copart into a quiet compounding favorite.

• 2022 to 2023: Jeff Liaw steps up to co-CEO and then sole CEO, with founder-era leader Jay Adair moving to Executive Chairman.

• 2024: Another 2-for-1 split. Copart pushes beyond pure salvage through Purple Wave (construction, ag, and fleet equipment) and powersports remarketing.

• FY2026: Growth stalls as insurance unit volumes soften and prior-year hurricane cars create tough comparisons. Copart starts buying back stock in size for the first time, and the market derates the shares hard.

Show Me the Money

Standout financial features:

• Revenue roughly tripled the slope you'd expect from a "boring" business over the last several years, then flattened out hard in the trailing twelve months at about $4.6 billion. The stall, not the level, is the story.

• Margins barely flinched even as growth stopped. Gross margin actually ticked up to ~45.5% and operating margin held around 37%, which tells you the model isn't broken, the volume just paused.

• The balance sheet is a fortress. Net cash grew from about $0.8 billion three years ago to over $3.2 billion, and CapEx pulled back sharply in the trailing year as Copart slowed its land buying. That combination is why they finally turned on the buyback.

Financial Data

(Net debt is negative across the board, meaning Copart holds far more cash and investments than debt. Fiscal year ends July 31.)

Stock Performance

Dividend-adjusted total returns through mid-June 2026:

That's a jarring table. Copart compounded roughly 5x over the decade, then erased five years of price gains in a single rough year as growth stalled and a very rich multiple compressed. You're now paying less for the stock than you would have five years ago, despite revenue and profit being much higher.

The peer comparison makes it sting more:

RB Global owns IAA, Copart's only scaled head-to-head competitor in US salvage auctions, so it's the cleanest comp. OPENLANE (formerly KAR Auction Services) runs a digital wholesale used-vehicle marketplace, an adjacent two-sided auction business. Both rose while Copart fell, which tells you this was a Copart-specific repricing, not the whole category breaking. When the best operator in a space underperforms its weaker rivals by 44 to 61 points in a year, it's usually expectations resetting, not the business losing.

The N.O.O.B. Nine — Competitive Powers

The Nerd Out on Business Nine is made up of Hamliton Helmer's famous "7 Powers" of competitive advantage (Scale Economies, Network Economies, Counter-Positioning, Switching Costs, Branding, Cornered Resource, and Process Power) combined with two of my own (Data Flywheel and Distribution Advantage).

Average Score: 4.0/5 - a moat built on scale and irreplaceable land, hardened by process power, with real two-sided network effects layered on top.

Memorable Marketing

Copart doesn't run flashy campaigns, and that's the point. This is a B2B supply machine where the "marketing" that matters is keeping insurance sellers happy and keeping global buyer demand deep. The brand work is quiet and functional, aimed at trust and reach rather than awareness.

Notable tactics:

• Global buyer access (ongoing): Copart deliberately opened bidding to international buyers and the general public, not just licensed dismantlers. Widening the buyer pool is the whole game, because more bidders mean higher prices for sellers, which wins more seller supply.

• Catastrophe response as a credibility flex (recurring): When hurricanes hit, Copart's ability to absorb tens of thousands of flooded cars fast is its best sales pitch to insurers. Operational reliability under stress is the marketing.

• Member and seller tooling (ongoing): Products like Copart 360 imagery and IntelliSeller are sold as ways to get sellers more money per car. The pitch is outcomes, not features.

Tactical takeaways:

1. In a two-sided marketplace, your best marketing spend is often widening one side of the market, not advertising the brand.

2. Reliability during a crisis is a sales tool. Be the partner who shows up when it's hard.

3. Sell your tools on the dollars they put in the customer's pocket, not on the technology behind them.

4. When you own the supply relationship, you don't need demand-side hype.

AI Uses & Opportunities

Current exposure:

• Copart already leans on machine learning and computer vision in its Co.ai total-loss valuation tools, which help insurers decide whether to repair or write off a car, and in IntelliSeller, which sets minimum bids and re-auction timing to maximize seller proceeds.

• Computer vision is used to capture and assess vehicle condition through its 360-degree imaging, reducing manual inspection work.

Future opportunities:

• Automated damage assessment from photos that grades a car's condition and predicts salvage value the moment it arrives, compressing cycle time.

• Smarter buyer matching and demand forecasting, routing the right cars to the markets and bidders most likely to pay up.

• AI-assisted title and paperwork processing, where a lot of human labor still hides.

• Dynamic catastrophe planning, using models to predict where storms will create surges and pre-positioning capacity.

• A real risk to watch: better AI repair estimates could change how often insurers total a car at all, which would shift Copart's supply at the source.

Bumps in the Road

• Growth has stalled. Trailing-twelve-month revenue was roughly flat, and global insurance unit volumes fell in recent quarters, partly because the prior year included a wave of hurricane catastrophe cars.

• Uncertainty regarding the impact of self-driving.

• Supply concentration. A large share of vehicles comes from a handful of major insurance carriers, so Copart is exposed to their decisions, claims trends, and accident frequency.

• Total-loss frequency is not in Copart's control. If cars get repaired more often instead of written off, fewer units flow into the yards.

• Valuation reset risk is real and just played out. The stock spent years priced for high-teens growth, and the repricing toward slower growth was rough.

• Catastrophe dependence cuts both ways. Storms boost volumes in big years and create ugly comparisons in the quieter ones.

Your Swipe File

• Owning the physical layer can help build and grow a digital marketplace. Permitted land and tow logistics are a moat precisely because they're slow to build.

• Get paid by both sides without taking inventory. Copart's consignment-and-fee model is why a they earn software-like margins. Look for places you can take a clean cut of a transaction you facilitate but don't own.

• Liquidity is a flywheel. In any two-sided market, concentrate on whichever side is harder to get, and the other side will follow the inventory.

• Embed yourself in the customer's workflow, not just their vendor list. Copart sits inside insurers' claims process, which has given them traction with the hard side of the market.

• Cash discipline buys optionality. Years of building cash meant that when the stock fell, Copart could lean into buybacks.