Chapter 7

Capital Allocation

Copart returned nothing to shareholders for a decade — no dividend since its 1994 IPO, no buyback in the six years through FY2025 — while free cash flow of roughly $4.4 billion piled cash and investments up to $4.8 billion. In early 2026 it broke the pattern, repurchasing $1.63 billion of stock, mostly in February and March near the lows. The move is real and founder-driven, but opportunistic: it paused in April, and $4.2 billion still sits idle.

Source: Q3 FY2026 Form 10-Q, Consolidated Balance Sheets and Note 6 — Stock Repurchases [1]; [2].

A decade of retention

For a business a value investor would prize for its balance-sheet safety, Copart's default with surplus cash has been to keep it. The company has not paid a dividend since it went public in 1994, and it states plainly that it intends to retain earnings for use in the business [3]. It also did not repurchase a single share under its buyback program in fiscal 2023, 2024 or 2025 [4], nor in fiscal 2020, 2021 or 2022 [5] — six consecutive years without a buyback or a dividend.

The result is visible on both sides of the ledger. Over FY2021 through FY2025 the business generated roughly $4.4 billion of cumulative free cash flow [6], and returned none of it. Cash and held-to-maturity investments climbed to $4.79 billion by the July 2025 year-end, against zero funded debt [7].

Source: free cash flow derived from reported operating cash flow less capital expenditures, FY2021–FY2025 10-Ks [8] and 9M FY2026 10-Q [9]; buybacks per Note 6 [10].

Retention this complete is a deliberate choice, not an accident of a stretched balance sheet. It is the counterpart to the fortress described earlier (The Franchise Marked Down): the same cash that removes any real bankruptcy risk also sits, year after year, earning a money-market return inside a business whose operations earn far more.

The framework, in management's words

Copart's stated priority order is consistent across years of calls, and it puts owned real estate first. The company launched a "20/20/20" initiative in April 2016 to acquire and expand 20 facilities over 20 months, and has since spent several hundred million dollars a year buying land, developing capacity and buying out previously leased yards, on the logic that "long-term stewardship requires ownership, as leasing limits our control over how we service the insurance industry" [11]. The CFO frames capacity, real estate, logistics and technology as "our number one priority from a capital allocation perspective," with acquisitions held to "a very high threshold for how we think about generating returns long term" [12].

Where the buyback sits in that order is explicit. Management describes evaluating investments over "30-, 40-, 50-year horizons," electing to "invest productively in the business long-term" first, and treating repurchases as the use of "residual capital" — bought "aggressively in real volume" but "episodically as opposed to via a predictable routine buyback program" [13]. That is the key to reading the record: the absence of buybacks through FY2025 was not neglect, and the FY2026 buyback is not a new standing policy. Both are the same opportunistic framework, expressed at different prices.

The first buyback in six years

The break came in the January–March 2026 window, as the stock fell through the high $30s. Copart repurchased 43.4 million shares in the first nine months of FY2026 at a weighted-average price of $37.63, for $1.63 billion — its first material repurchase in six years — leaving 282 million shares still authorized [14]. Net of shares issued for equity compensation, the count fell 4.3%, from 967.5 million to 925.8 million [15].

Two features of the timing matter. First, the buying was concentrated and then it stopped: only 5.5 million shares through the end of January, then 37.9 million across February and March, then nothing in April [16].

Source: Q3 FY2026 Form 10-Q, Issuer Purchases of Equity Securities [17].

Second, the $1.63 billion exceeded the roughly $0.99 billion of free cash flow Copart generated in the same nine months [18], so the company dipped into its accumulated hoard rather than merely spending the period's cash. It funded part of this by letting held-to-maturity securities run off — that balance fell from $2.01 billion to $0.85 billion over the nine months — which is why cash-plus-investments declined only about $0.6 billion, to $4.20 billion, even after the repurchase [19].

For the value-oriented reader, the honest reading cuts both ways. On the constructive side, this is the founders acting on the cheapness the report has documented — deploying more than a year's free cash flow at a decade-low multiple (Margin of Safety), the strongest evidence in the record that they treat the share price as a real allocation input. On the cautionary side, the buying was executed at $37.63 against a $27.28 quote today, so those repurchases are marked roughly 27% underwater so far, and the program went quiet in April even as the price fell further. Being opportunistic on the way down is not the same as being committed: the per-share compounding that every upside scenario leans on is a management decision that can pause, and just did.

The cost of the fortress

The idle capital carries a measurable return cost. Return on capital employed slipped from about 28% in FY2022 to 18% in FY2025, even though the underlying auction business remained highly profitable — the compression is amplified by a growing pile of low-returning cash and securities.

Source: derived from reported financials, FY2020–FY2025 10-Ks [20].

The scale of the drag is straightforward. FY2025 net income of $1.55 billion on $9.19 billion of equity is a 16.9% return [21]. Strip out the roughly $4.8 billion of cash and investments and the after-tax interest it produced — net interest income was $178.9 million pretax in FY2025, up from $65.9 million two years earlier [22] — and the operating business earns on the order of 30% on the capital actually deployed in it. The reported return is roughly halved by capital that is not working. That interest stream is also rate-sensitive rather than operating, a point the earnings-quality discussion develops further (Financials and Estimates).

None of this argues for abandoning the fortress. For an investor whose stated calibration is that the chance of bankruptcy be near zero, a debt-free balance sheet with $4.2 billion of net cash is exactly the margin of safety the business is built to provide, and management is right that it let Copart keep serving insurers through the pandemic without layoffs or capex cuts — a competitive asset with clients who "have a long memory" [23]. The point is narrower: safety at this scale is not free, and the gap between the ~30% the operations earn and the ~17% the equity earns is the price of holding several billion dollars in reserve. Notably, the company also arranged a new 2026 Credit Agreement it may draw on to fund expansion [24], an unusual step for a company sitting on this much cash and one worth watching for what it signals about future deployment.

What the founder inherits

The capital-allocation question sharpens precisely as control changes hands. Co-founder Willis Johnson's long-time partner Jay Adair returns as CEO effective July 31, 2026 (Founders and Alignment), inheriting a $4.2 billion cash position, an unused 282-million-share authorization, and a mix that has already begun to shift: capital expenditure ran at just $259 million over the first nine months of FY2026 [25], well below the $569 million spent in FY2025, even as the buyback ramped [26].

What Adair does with that cash — sustain the buyback into a genuine per-share compounding program, deploy the hoard into land or M&A at the returns the framework demands, or revert to accumulation — is the clearest forward read on whether the founder reset is a change of substance or of nameplate. The corpus ends before his first capital-allocation decisions as CEO are disclosed, so the record here is the setup, not the outcome. The signal to track is simple and observable: whether repurchases resume at the current lower price, or the February–March 2026 burst stands as a one-time opportunistic strike that lapsed the moment the stock got cheaper.