Salvage Moat

Salvage Moat

Copart's durability rests on three reinforcing advantages: a global pool of roughly one million buyers that lifts what insurers recover, an owned-yard footprint that is slow and costly to replicate, and deep integration into insurer claims workflows. The clearest proof sits in the margin — 36.5% at the operating line, against near-identical-revenue RB Global's 17.8% [1][2]. The moat looks wide. Its soft spot is that seller ties are relationships, not contracts, and long-run demand depends on total-loss trends.

The Margin Gap

A moat should show up in the numbers, and here it does. Copart and RB Global are the two listed pure-plays in vehicle-auction remarketing, and their most recent fiscal years line up almost exactly on revenue — Copart at $4.65 billion (year ended July 2025), RB Global at $4.28 billion (year ended December 2024). What they do with that revenue is not close.

Sources: Copart FY2025 10-K, Consolidated Statements of Income [3]; RB Global FY2024 10-K, Consolidated Income Statements [4].

Sources: Copart FY2025 10-K [5]; RB Global FY2024 10-K [6].

On roughly the same revenue, Copart earned $1.70 billion of operating income to RB Global's $761 million, and $1.55 billion of net income to RB Global's $413 million [7][8]. The comparison is imperfect: RB Global is a blend of Ritchie Bros' commercial-equipment auctions and IAA, the salvage marketplace it bought in March 2023 for about $6.6 billion [9][10]. But IAA is precisely Copart's nearest salvage rival, and even folded inside a larger group it drags the blended margin to less than half of Copart's. RB Global also carries the cost of that deal: $234 million of interest expense in 2024, against Copart's net interest income of $179 million on a debt-free balance sheet [11][12]. A well-capitalized rival has spent the better part of a decade competing head-to-head and still cannot match Copart's unit economics. That gap is the moat, expressed as a number.

The Network Flywheel

The first source of that gap is a two-sided marketplace that gets more valuable as it grows. Copart maintains a database of roughly one million registered buyers — dismantlers, rebuilders, used-car dealers, and exporters — and because its VB3 platform is entirely virtual, a wrecked car in Texas can draw bids from a buyer in Eastern Europe or West Africa without either party moving [13][14]. Copart states the mechanism plainly: the virtual platform "increases the pool of available buyers for each sale, which brings added competition and an increase in the amount that buyers are willing to pay" [15].

That is the flywheel. More buyers raise the winning bid; a higher bid means the insurer recovers more on a totaled car; higher recoveries make Copart the obvious place to send the next assignment; more assignments deepen the inventory that attracts still more buyers. International buyers are the part a domestic-only rival struggles to replicate — foreign importers "now represent a significant part of our total buyer base," and they bid on cars that a local salvage yard could not clear at the same price [16]. The result is the supply concentration behind the whole model: insurers supplied 81% of the vehicles Copart processed in fiscal 2025 [17]. Crucially, Copart earns fees on the transaction rather than buying most cars for its own book, so each incremental sale adds high-margin revenue without inventory risk — the capital-light half of the 36.5% margin.

The Physical Moat

The second advantage is the opposite of capital-light, and that is the point. A virtual auction still needs somewhere to put the cars. Copart operates 281 facilities globally and, unlike its principal rival, owns much of the land under them [18]. Net property and equipment has climbed from $1.9 billion in fiscal 2020 to $3.6 billion in fiscal 2025 as the company steadily bought yards rather than leased them [19].

Source: derived from Copart reported balance sheets, FY2020–FY2025; FY2025 figure per the FY2025 10-K, Consolidated Balance Sheets [20].

Owning the land matters because the land is genuinely scarce. New sites run into "zoning restrictions or difficulties obtaining and maintaining use permits," and Copart warns that when it cannot secure enough capacity it risks damaging "our relationships with insurance companies" — the salvage yard is not a commodity warehouse but the physical link that keeps insurer volume flowing, especially when a hurricane fills every yard in a region overnight [21]. RB Global runs the leaner-looking version: it carried $1.5 billion of operating-lease right-of-use assets at the end of 2024, renting much of the ground Copart owns outright [22]. Ownership ties up capital and holds free-cash-flow conversion below net income — the cost side of this choice, established earlier (Financials and Estimates) — but it also builds a barrier a new entrant would need years and a permitting fight to cross.

Switching Costs, Stated Honestly

The third pillar is the softest, and worth being precise about. Copart is deeply embedded in how insurers process a total loss: it clears salvage titles across every state — "our know-how in the area of title processing is a competitive advantage" — arranges pickup within roughly a day, and hands adjusters an integrated workflow that would be disruptive to rebuild around a competitor [23]. That is a real switching cost, but it is operational, not legal. Copart itself flags "the absence of long-term contractual commitments between us and our sellers," so nothing contractually stops a large insurer from shifting assignments to IAA [24]. The lock-in is that Copart usually returns more money per car and moves it faster — an advantage it has to re-earn on every cycle rather than one it can invoice for. The stable ~36% margin through years of volume wobble is the evidence that, so far, it keeps re-earning it.

What Could Break It

The moat has held against the obvious attack. A well-funded competitor already exists, has owned the number-two salvage platform for years, and still cannot close a nineteen-point margin gap — which separates Copart's structural advantage from mere good execution. The more serious threat is not competitive but demand-side, and it is genuinely two-sided: safer cars and driver-assistance technology should reduce accident frequency over time, which would shrink the pool of damaged vehicles, while the rising cost of repairing those complex cars pushes more of them past the total-loss threshold, which enlarges it. Whether rising total-loss frequency outruns falling accident frequency is central to whether the network keeps compounding or merely holds — and it deserves its own treatment, which Demand Engine takes up.

On the evidence here, Copart clears a wide-moat bar: the advantage is company-specific, quantified in a durable margin gap, and has already survived a serious competitor's attempt to match it. The read that would change is a demand one — if total-loss volumes fell for several consecutive years even as accident rates normalized, the flywheel would slow and the owned-land barrier would protect a shrinking pie rather than a growing one.