Analysis

A £17m profit or a £97m loss? The £660m Premier League clubs made by trading with themselves

In 2024/25 Aston Villa reported a £17m profit. The company above it reported a £97m loss. Both are audited. The Orange Ledger rebuilt 20 seasons of Premier League accounts and found £660m of profit in the last three seasons that came from deals inside the owners' own groups, against £52m in the 17 seasons before, booked just as the spending rules began to bite. The rule-makers have since stopped counting it.

By Arthur van de Laak21 October 2026Updated 31 October 2026 · View updates21 min read

In 2024/25, Aston Villa made a profit of £17.0 million.

Or it lost £96.7 million.

Both figures are audited. Both cover the same football club in the same twelve months. The first is in the accounts of NSWE Sports, the company that holds the club. The second is in the accounts of NSWE UK, the company that owns NSWE Sports.

The difference is £113.7 million, and it has a simple explanation. Think of NSWE UK as the parent at the top of the family, with two companies beneath it: NSWE Sports, which holds the club, and NSWE Holding. During the season, NSWE Sports sold Villa’s women’s club and a warehouse company to NSWE Holding, and booked a profit of £113.7m on the sales. The parent owns both the seller and the buyer. In its accounts, the family sold something to itself, so the profit disappears.

Newcastle United’s accounts for the same season tell a similar story. Newcastle United Ltd reported a profit of £34.7m before tax (with no tax charge, it is the same after tax), which included £133.1m of gains on selling St James’ Park leasehold improvements and a subsidiary to other companies in its group. NCUK Investment, the owners’ company above it, reported a loss of £118.1m. There, too, the gain disappears; it accounts for most of the difference between the two figures, and the rest is in the other items of the owners’ accounts.

These sales are disclosed in audited accounts, and we are not suggesting they were false accounting. But they matter, and they matter now. This month England’s new football regulator warned that English football is “standing on the edge of financial disaster”, with the average Premier League club losing £44m in 2025/26. The losses football has reported for years were themselves made smaller by deals like these.

Explore the twenty-year record: every season’s figures, with the adjustments you choose.

£660m in three seasons

To measure how much, we rebuilt 20 seasons of Premier League profit and loss, 2005/06 to 2024/25, from the accounts filed at Companies House: nearly 400 club-seasons. We searched every filing for profits on selling property, subsidiaries, women’s teams and stadium interests to the owners or their companies, and for loans or costs that owners wrote off or took over. Every gain of £5m or more was checked on its page by two readers working independently. (How we chose which company’s accounts to use for each club is in the method section at the end.)

The result splits football’s last 20 years into two eras.

• 2005/06 to 2021/22: £52m in 17 seasons. In the filings we hold, and on the criteria above, we found three such items: a £9.5m loan written off in 2007/08 by Fulham’s former parent company, which, like its new one, was controlled by the Al Fayed family; £20.3m of loan interest waived by Aston Villa’s owner in 2011/12; and Manchester City’s £22.5m sale of intellectual property to related companies in 2012/13. Clubs did sell old grounds and training sites in those years, Everton’s Bellefield and West Ham’s Boleyn Ground among them, but the filings do not show them as sold to the owners or their companies, so we did not count them.
• 2022/23 to 2024/25: £660m in three seasons. Every one of these items came from a company in the same group, from the owner or, at Bournemouth, from the former owner on the sale of the club.

Bournemouth’s case is not a sale of assets, but it is the same kind of money. When the club was sold in December 2022, the outgoing owner’s company released the loans the club owed it. The debt did not just disappear: the club booked it as income for that season. A loss turns into a profit, with no money earned from football.

In four of these six seasons, the club reported a profit that, without the deal, would have been a loss: Chelsea 2023/24, Newcastle, Villa and Bournemouth. Chelsea’s 2022/23 accounts also include £17.1m of legal costs recharged to its parent, Blueco 22: costs the club incurred, taken over by the owners’ group.

Across the league, the effect on the last five seasons is this:

• Reported loss, 2020/21 to 2024/25: £2.94bn.
• Without the owner-related gains and credits: £3.60bn. Another £660m, or 22% more.
• In 2023/24 the league’s reported loss of £56m becomes £254m, more than four times as large.
• 2024/25: a reported loss of £786m becomes £1.08bn.

These totals are not the ones in Deloitte’s Annual Review of Football Finance, the usual reference. Deloitte’s 2026 edition puts the clubs’ combined loss over the same five seasons at £3.07bn, against our £2.94bn. Both add up the clubs’ audited accounts. They differ in whose accounts: a club usually sits in a chain of companies, each filing its own result, and a deal between two companies in the chain is a profit in one set of accounts and cancels in the other. We count the nearest company that holds only the club, every season. Deloitte does not publish its choice, and no single rule reproduces its figures in all five seasons. The largest gap is 2024/25, where Deloitte reports a loss of £948m against our £786m. Of the tens of thousands of ways to combine the clubs’ filed accounts, every one that comes within £5m of Deloitte’s figure counts Villa at NSWE UK, where its £113.7m gain cancels, and Bournemouth at its multi-club parent; together they make £153m of the £162m. So Deloitte’s 2024/25 loss most likely already leaves out Villa’s gain, while its own text keeps Newcastle’s and Everton’s. Our basis is the one that shows each gain where it was booked. Deloitte’s figure for each season is beside ours in the full record.

Why it mattered: the spending rules

The difference between making money and recording a profit has rarely mattered more in Premier League football.

The Premier League’s Profitability and Sustainability Rules (PSR) took effect from 2013/14 and were first applied in 2015/16, to the three seasons 2013/14 to 2015/16. They capped how much a club could lose: £105m over three seasons, less for clubs that had spent some of those seasons in the Championship. Go over it, and a club faced sanctions up to points deductions. Everton and Nottingham Forest both lost points in 2023/24. PSR was replaced by the squad cost rules in 2026/27.

The test started from the club’s own result. A gain on selling a hotel, a women’s team or a stadium lease to a company owned by the same people counted towards it, as long as the Premier League accepted the price as fair. So these sales did something specific: they did not bring in money from outside the owners’ groups, but they reduced the loss the rules started from, in the three-year windows that decided whether a club was in trouble. And they arrived in exactly the seasons when those windows began to cost clubs points.

The table below needs one warning. It shows each club’s reported loss over each three-year window, with and without the owner-related gains, beside the limit. That is not the PSR calculation itself: the rules allow deductions for spending on academies, women’s football, community work and facilities, which clubs do not publish. So the table shows how far the gains moved the starting point, not whether any club passed or failed.

Reported results over each three-season window, at the nearest company that holds only the club, before the deductions PSR allows. Not a PSR calculation and not a finding of breach. The rules let a club submit its own company’s accounts or its group’s, and which each club used is not public. The limit is £105m over three seasons, less £22m for each of those seasons a club spent in the Football League.

* Under the Covid-era rules, the window ending 2022/23 counts 2022/23 and 2021/22 in full, and the average of 2019/20 and 2020/21 as its third season. Covid-related costs could also be deducted.

† Bournemouth’s averaged pair mixes a Premier League season with a Championship season, and the League has not published how that counts. Its limit for this window is £83m or £61m; we show the higher. The loss without the write-off is beyond either.

Read the table with the warning in mind. Several windows show reported losses above the limit even with the gains included. That does not mean a breach: the deductions the rules allow can be large. What the table does show is how much the gains moved the starting point. In four windows, the reported loss is inside the limit and the loss without the gains is beyond it: Bournemouth’s windows ending 2022/23 and 2023/24, Chelsea’s ending 2023/24, and Newcastle’s ending 2024/25. The gains are shown as booked; where the Premier League’s fair-value process adjusts a price, as Newcastle says it will, the amount accepted may differ.

The rule-makers reached the same conclusion

Two regulators have since decided this kind of profit should not count.

• UEFA, July 2025: Chelsea agreed a settlement of up to €80m and Aston Villa up to €20m, plus further amounts under UEFA’s squad cost rule. UEFA said profits from asset sales and related-party deals cannot count as relevant income in its calculation. Those settlements covered accounts for periods ending in 2023 and 2024, so they are not about the 2024/25 sales in this article; they show how UEFA treats this kind of profit.
• The Premier League, from 2026/27: its new squad cost rules exclude any profit on the disposal of investments, hotels included, whoever the buyer is. Among disposals, only the sale of players counts.

For a period, then, a Premier League club could improve the figure its rules started from by selling a hotel to a company it already owned. That period is now over.

A different and much larger question: Manchester City

The sales above are lawful and visible in the accounts. The Premier League’s case against Manchester City is about something else: income that the league’s independent commission found was not what City’s accounts said it was.

In September 2026 the commission found that of £949.9m City recorded as sponsorship from Abu Dhabi partners between 2009/10 and 2017/18, £830.7m was in fact paid by the club’s owner, ADUG, rather than by the sponsors. City dispute the findings and have appealed (2 October 2026), and no sanction has been decided.

The decision prints the owner-paid sums season by season. The seasons that matter most are the five in which break-even rules applied: UEFA judged clubs from 2013/14, and the Premier League’s PSR applied from 2015/16.

In those five seasons City’s accounts showed a club close to break-even, from a £23m loss to a £20m profit. Without the owner’s money described as sponsorship, every one of them is a loss of between £97m and £134m. The full record below holds each season’s figures: select a season there to see City’s own result with and without the owner’s payments.

We keep this separate from the sales above, because it is a different thing: not a sale between companies in the same group, but a finding, under appeal, that income recorded as sponsorship came from the owner.

A different kind of missing cost: Chelsea

In March 2026 Chelsea accepted that £47.5m of payments made between 2011 and 2018, mostly to agents and in connection with transfers, had been made by companies linked to its former owner rather than through the club. The Premier League fined the club £10m and said that “in no scenario” would Chelsea have exceeded the PSR limit.

These were costs missing from the accounts, not income added to them. Because most were costs of signing players, their effect on Chelsea’s results would have been spread over the players’ contracts. Our reconstruction, which rests on estimates where the league’s document gives no contract lengths, puts it at up to about £9m a season, highest between 2014/15 and 2017/18. Our headline figures leave it out; the full record below lets you include it, as an estimate.

What the warning means

“It will demonstrate how it is possible for English club football to simultaneously be the richest, best and most deeply competitive in the world while also standing on the edge of financial disaster.” David Kogan and Richard Monks, Chair and Chief Executive, Independent Football Regulator, State of the Game 2026 (draft for consultation), page 2

The regulator’s draft report puts the league’s 2025/26 losses at about £880m across 20 clubs (page 22). It notes in passing that financial rules “could have unintended consequences, such as stadium or asset sales” (page 30).

Our reconstruction shows what that looked like in the years before the warning. Hundreds of millions of pounds of reported Premier League profit came from deals inside ownership groups, not from money earned outside them. The deals were reported, and the rules allowed them. But they made some clubs’ losses look smaller at precisely the time football was supposed to be bringing its losses under control. And if the commission’s findings stand, for nine seasons one club’s results were also lifted by its owner’s money described as sponsorship.

We have measured only the Premier League. The regulator found clubs losing money in every division of the EFL, and whether the same pattern runs further down the pyramid remains unanswered. The 2025/26 accounts are not filed yet either, so we cannot say how much of this season’s losses is softened in the same way.

The regulator’s consultation is open until 8 December 2026. The Orange Ledger will submit this reconstruction, with every figure traced to the filing and page it comes from.

The full record

Below you can explore the complete 20-season dataset, inspect every transaction with the filing and page it comes from, and read exactly how we did it.

Twenty seasons: revenue, reported result and the three adjustments

Figures in £m, profit before tax, at the nearest company that holds only the club. The reported figures never change. The scenario takes out the owner-related gains and credits; it includes the City and Chelsea amounts only if you switch them on (in the fixed table they stand beside the result). Both show the amount by which a club’s reported result was higher than it would otherwise have been: City’s is the commission’s finding, under appeal; Chelsea’s is our reconstruction of the payments it accepted in March 2026.

* 19 clubs: Portsmouth’s accounts for 2008/09 and 2009/10 are not in our files.

Every gain we removed, with its filing and page

Each was read on its page by two readers working independently, and both agree.

Totals are summed from unrounded figures, so they can differ from the rows by 0.1.

Found but not counted:

• Chelsea: a £12.5m settlement fee in 2022/23 and £10.6m of recharged costs in 2020/21. Neither filing says who paid them.
• Manchester United: £26.1m over 2022/23 to 2024/25 for services recharged to other companies in its group. These are payments for work the club does, not support from the owner.
• Loans and interest waived straight into the balance sheet, at several clubs, among them the interest Manchester City’s owner waived in 2009/10. These never passed through profit, so they did not change the reported result.
• Manchester City’s £24.5m sale of intellectual property to third parties in 2012/13. The commission found that, in a year it does not name, City booked as income £24.5m paid under an image-rights arrangement set up in 2012, money it found came from the owner. The amount and timing match, but neither document says they are the same transaction, and the finding is under appeal.
• Sales not shown as made to the owners or their companies, among them Everton’s Bellefield training ground (£8.4m, 2010/11) and West Ham’s Boleyn Ground (£8.7m, 2016/17).

What we did not adjust

This article measures one thing: profit that came from the owners themselves. It is not a measure of “underlying” profit.

In the five seasons from 2020/21 to 2024/25, clubs also booked hundreds of millions of pounds of one-off costs: write-downs of player contracts (which involve no payment), legal and regulatory costs, and restructuring. None of these came from the owners, so we left them in. A full normalisation would treat one-off gains and one-off costs alike, and would answer a different question.

How we did it

• Which accounts: for each club-season, the nearest company that holds the club and no other club. Where a group owns several clubs (City Football Group; Blueco 22 from 2023/24; Black Knight Football Club UK in 2024/25), we used the club’s own group instead. For Villa and Newcastle in 2024/25 we also show the owners’ company one level up, where the gains cancel.
• Coverage: 398 of 400 club-seasons. Portsmouth’s 2008/09 and 2009/10 accounts are not in our files, so those two seasons cover 19 clubs.
• Periods: some accounts cover more or less than twelve months; they are used as filed and listed in the data.
• Which gains and credits: profits on selling property, subsidiaries, women’s teams and stadium interests to the owner or a company in the same group; loans and interest written off by owners and booked as income (at Bournemouth, by the outgoing owner when the club was sold); and costs the club incurred that the owner’s group took over. Not included: payments for services the club provides to group companies, sales to outside buyers, player sales, and waivers booked straight to the balance sheet. A gain counts at the amount in the profit and loss account, not the sale price; for Chelsea’s hotels, where that line covers every fixed-asset disposal in the year, it counts at the gain in note 2.
• How we found them: besides the sales, we searched the text of all 711 filings for recharges, contributions, waivers and costs borne by a parent, and read every candidate of £5m or more twice. The search runs on scanned text, so unusual wording could be missed.
• Checks: every gain of £5m or more, and every group figure used, was read on its page by two readers working independently; only figures on which both agree are used.
• Deloitte: the comparison with Deloitte’s 2026 Annual Review of Football Finance (Chart 9: 2020/21 £707m, 2021/22 £601m, 2022/23 £683m, 2023/24 £135m, 2024/25 £948m) is under the five-season figures above.
• City: the per-season sums are from footnote 8 on page 20 of the commission’s decision.
• Chelsea: the payments are from the Premier League’s published sanction agreement. Transfer-related payments are spread over the players’ contracts, at five years where the length is not known (three years shown as a check in our data); remuneration falls in the year paid.