The short version

PSR — the Profitability and Sustainability Rules — was the Premier League’s loss limit. Between 2013 and the end of the 2025/26 assessment cycle, a club could not lose more than £105 million over a rolling three-year period, measured not on its raw accounts but on an adjusted figure that strips out spending the league wants to encourage (academies, women’s football, community work, infrastructure). Break the limit and an independent commission could deduct points — as Everton and Nottingham Forest found out.

Arsenal’s position under PSR was comfortable. Even on headline numbers, their cumulative pre-tax losses across 2022/23 to 2024/25 totalled about £71 million — inside the £105 million ceiling before any of the permitted deductions are applied. After those deductions, their adjusted PSR losses were lower still. The club’s own accounts statements confirm compliance with all applicable financial sustainability regulations.

One important update: PSR has now been replaced. Premier League clubs voted in November 2025 to move to a Squad Cost Ratio (SCR) system from 2026/27, which caps squad spending at 85% of football revenue plus net player-sale profit. This article explains the PSR system that governed English football for over a decade — and that still shapes the squad-building decisions you see today — then covers what replaced it.


How PSR actually worked

1. A rolling three-year window

PSR never judged a single season. Each assessment looked at three seasons together: as a new season’s result came in, the oldest dropped out. A disastrous year could be absorbed if the surrounding years were sound — and a club could drift into trouble gradually, because one bad year stayed in the window for three assessment cycles.

2. The £105 million ceiling — and the £15 million catch

The headline limit was £105 million of losses over three years, equivalent to £35 million a season, for a club that spent all three seasons in the Premier League. But there were two tiers:

  • Up to £15 million of losses over three years could simply be absorbed.
  • Losses between £15 million and £105 million were only permitted if the owner covered them with secure funding — real equity injected into the club, not loans. In other words, the league allowed big losses only when an owner had demonstrably put their own money behind them.

Clubs that spent part of the window in the EFL had a smaller allowance: each Championship season in the window counted at £13 million rather than £35 million. That is why Nottingham Forest were assessed against a £61 million limit (two EFL seasons plus one Premier League season), not £105 million.

3. Adjusted earnings: the add-backs

The loss PSR measured was not the loss in the published accounts. Clubs could add back spending on:

  • Youth development and the academy
  • Women’s football
  • Community programmes
  • Depreciation of tangible fixed assets (stadium and infrastructure)
  • In earlier cycles, certain costs attributed to the COVID-19 pandemic

This is the single most misunderstood part of PSR. A club could report headline losses well above £105 million across three years and still comply, because millions of pounds of “good” spending were removed from the calculation. When you see a headline like “Club X lost £115m in three years but passed PSR,” add-backs are almost always the explanation.

4. Enforcement had teeth

Breaches were referred to an independent commission, which could impose points deductions. The landmark cases:

  • Everton were deducted 10 points in November 2023 for a breach in the cycle ending 2021/22 (reduced to 6 on appeal), and a further 2 points for a second breach.
  • Nottingham Forest were deducted 4 points in March 2024.

Whatever its flaws, PSR was not a paper rule.

5. UEFA ran a parallel system

Clubs in European competition also had to satisfy UEFA’s Financial Sustainability Regulations, which work differently:

  • The football earnings rule limits losses to €5 million over three years, extendable to €60 million if the excess is covered by owner equity contributions.
  • The squad cost rule caps wages, transfer amortisation and agent fees as a share of revenue — phased in at 90% (2023/24), 80% (2024/25) and 70% from 2025/26.

For a Champions League club like Arsenal, UEFA’s 70% squad cost ceiling is in practice the tighter constraint.


A worked example: why Arsenal’s “£115m of losses” was never a breach

Take the three seasons ending 2023/24. Arsenal’s reported pre-tax losses were:

Season Pre-tax result
2021/22 £45.5m loss (as reported from the club’s accounts)
2022/23 £52.1m loss
2023/24 £17.7m loss
Headline total £115.3m loss

On raw numbers, that is above £105 million. Yet Arsenal were never in PSR danger over this period, for two reasons:

  1. Add-backs. Arsenal spend heavily in exactly the categories PSR excluded — one of England’s most productive academies, and a women’s team that is among the biggest operations in the WSL. Add infrastructure depreciation at a 60,000-seat stadium and community spending, and the adjusted PSR loss falls well below the headline figure.
  2. Owner funding. To the extent losses exceeded the £15 million lower threshold, they were underwritten by the ownership (Kroenke Sports & Entertainment), satisfying the secure-funding requirement.

The following cycle was more comfortable still: losses narrowed to £17.7 million (2023/24) and then £1.4 million (2024/25), so the three-year headline total to 2024/25 was only about £71 million — inside the limit before deductions.

Lesson for readers: never assess a club’s PSR position by adding up headline losses. The adjusted figure is what counts, and clubs are not required to publish it — which is why credible estimates (for example from Swiss Ramble) reconstruct it from the accounts’ detail.


What replaced PSR — and why it matters for Arsenal

From 2026/27, the Premier League’s main control is the Squad Cost Ratio:

  • Player and head-coach wages, transfer amortisation and agent fees must not exceed 85% of football revenue plus net profit on player sales.
  • Between 85% and 115% (“the red threshold”), clubs face financial levies rather than points deductions.
  • Above 115%, sporting sanctions — points deductions — apply.

The logic has changed: instead of policing losses after the fact, the league now caps the squad-cost engine directly. For Arsenal, whose wages-to-revenue ratio sits around 50% (see our wage-bill explainer), an 85% domestic ceiling is not the binding constraint — UEFA’s 70% squad cost rule is. The practical effect of the new regime falls hardest on clubs whose wage and amortisation bills run close to their revenue.


FAQ

What does PSR stand for? Profitability and Sustainability Rules — the Premier League’s former financial control system, in force for assessments from 2013 to the cycle ending 2025/26.

What was the PSR limit? Adjusted losses of no more than £105 million over a rolling three-year period for an ever-present Premier League club. Only £15 million of that could be uncovered losses; the rest required owner equity funding.

Why could clubs lose more than £105m in their accounts and still comply? Because PSR used adjusted earnings. Academy, women’s football, community and infrastructure-depreciation costs were added back, so the PSR loss was usually much smaller than the accounting loss.

Did Arsenal ever breach PSR? No. The club’s accounts statements confirm compliance, and their loss trajectory (from £52.1m in 2022/23 to £1.4m in 2024/25) moved them further from the limit over time.

What is SCR? The Squad Cost Ratio, PSR’s replacement from 2026/27. It caps squad costs — wages, amortisation and agent fees — at 85% of football revenue plus net player-sale profit, with levies up to 115% and points deductions beyond that.

Which rule matters most for Arsenal now? UEFA’s squad cost rule. At a 70% ceiling on wages plus amortisation plus agent fees, it binds earlier than the Premier League’s 85% for a club with Arsenal’s cost profile.


Note on figures: PSR assessments use adjusted earnings that clubs do not publish directly. Where this article reconstructs Arsenal’s position from headline results, it says so; the direction of the adjustment (downward, because of add-backs) is a matter of the rules’ design, not estimation.