The Orange Engine · Technical notes →

The disposal basis

One unresolved reading of the rules changes the profit on selling a player whose contract falls under the five-year amortisation cap. It is a question of basis: two defensible readings of the same rules produce different numbers. In the current sample it is worth 0.39–0.51 pts at Chelsea (the lower figure counts only registrations with firm evidence) and less elsewhere, because the cap only began in December 2023.

Two rules that do not quite meet

• Rule A.1.22(c) spreads a transfer fee over the contract, capped at five years, so clubs cannot lower annual squad cost by stretching contracts. It applies only to contracts executed or extended on or after 12 December 2023, the date the clubs voted it in, agreeing it would not be backdated. On this model's reading, an extension on or after that date brings the registration within the cap from the extension date.
• Rule A.1.13 defines profit on a player sale, which enters the ratio through a three-year average: the proceeds less the registration's carrying value at the date of sale.

The question is which carrying value, for a registration the cap applies to. The accounts carry it on the contract term they recognise; the cap would carry it on at most five years. A.1.22(c) contains no express bridge into A.1.13 saying the capped figure applies.

What this model does, and why

This reconstruction uses the accounting carrying value. A.1.22(c) is an amortisation provision, and the cap is described as applying to amortisation in squad cost. Rule A.1.22 also defines estimated squad cost by reference to the club's annual accounts, which points at the accounting basis rather than a synthetic one.

That is a reading, not a certainty. Nothing in the rules says the cap stops at squad cost, and a club arguing the other way would have the simpler position: that one rule cannot value the same registration two ways in the same calculation. For a registration signed before the cap, the argument runs the other way: its amortisation was never capped, so the accounting value is the only consistent one.

Where the two readings differ

Three conditions must hold together:

1. the contract runs longer than five years;
2. it was executed or extended on or after 12 December 2023;
3. the player is sold before the accounting carrying value has run to zero.

Up to the fifth anniversary the gap widens, because the capped basis writes the fee off faster. After five years the capped value is nil while the accounting value can still carry something until the contract ends.

Outside the question: Enzo Fernández, signed in January 2023 on a long contract before the cap existed and never extended since. The simplest case inside it: João Félix, £42.0m in Aug 2024 on a 7-year contract and sold about a year later. The accounts carry £36.1m at that point, the capped basis £33.7m, so the same sale produces £2.4m more profit on the second reading.

The largest case shows how an extension brings a player in: Nicolas Jackson signed in July 2023, before the cap, but his September 2024 extension to 2033 brings him within it from that date. The capped basis carries £16.5m at sale against £21.1m in the accounts: £4.5m more profit. The exact day of the extension is not sourced; he qualifies on any date that month, and the extra profit ranges from £4.2m to £4.5m.

Every registration in the twenty-club sample of 23 September 2026 that meets all three conditions. Together they add £12.0m of profit, £10.4m of it Chelsea's. Rows marked est. rest on an estimated contract length or date (Omari Kellyman, Renato Veiga, Morgan Rogers, Mathis Amougou); without them the total is £9.0m.

What it would do to the ratios

Sale profit enters the ratio as a three-year average, so the extra profit is divided across three years first. Taking only disposals inside the current averaging window, each club measured against the figure it had when the analysis was run:

3 of the twenty clubs moved, all by 0.51 points or less, and no club changed compliance state on either reading. The question is smaller than others the engine publishes, such as the treatment of Omar Marmoush's loan (see the Engine page).

Why it is small, for now

The longest any registration in the sample has been inside the cap before sale is about two and a half years. Every capped disposal so far sits on the rising side of the gap: the cap is less than three years old, and the gap peaks at the fifth anniversary, which no capped registration can reach before December 2028. Chelsea carries most of the current total, £10.4m of £12.0m, because 6 of the 8 registrations are theirs. A registration whose contract does not exceed five years has no exposure to this difference at all.

What would settle it

A club's own accounts cannot, because they are prepared on the accounting basis by definition. The answer has to come from the League: express drafting that carries the cap into A.1.13, or a published decision in a case where the difference mattered.

Until then this model states the basis it uses, publishes the size of the alternative, and does not average the two. If the rules do require a capped carrying value at sale, every affected sale profit in the series rises, by an amount largest around the fifth anniversary of each signing or extension.

Contract lengths are as reported and are themselves estimates at several clubs. The capped carrying value uses the same straight-line basis as the accounting one, with the five-year cap substituted for the contract term. Where a registration was extended before sale (Jackson, Savinho), both carrying values fold the extension. A registration counts as subject to the cap where it was signed, or on this model's reading extended, on or after 12 December 2023, the test the engine applies to amortisation. A 5% direct-cost deduction is applied to gains on both bases. Independent estimate, not an official calculation, and not advice.

← Back to the Orange Engine · Changelog · Technical note: the wage roll-forward