How Are Cartel Fines Calculated? EU, US and Türkiye Compared

How cartel fines are calculated is knowable in advance — the methodologies are published — and understanding the arithmetic changes decisions: what a marker is worth, what settlement saves, why parental liability matters, and why the same cartel costs different participants wildly different amounts. This guide walks the EU, US and Turkish methodologies with worked logic, as part of the cartel pillar of our Competition & Antitrust hub.

How does the European Commission build a cartel fine?

Four steps under the 2006 Guidelines. Basic amount: a gravity percentage of the value of affected sales — for cartels, typically 15–30%, in practice clustering near 15–19% — multiplied by the number of years (and fractions) of participation, plus an “entry fee” of 15–25% of one year’s sales added regardless of duration, purely for deterrence.

Adjustments: aggravation for ringleaders, coercers and — heavily — recidivists (repeat offenders have seen uplifts of 50–100%); mitigation for marginal roles or authority-encouraged conduct is rare and small. The cap: the total is trimmed to 10% of group worldwide turnover. Reductions: leniency and settlement applied last. The structure explains observed outcomes: duration is close to linear (fourteen years of trucks coordination produced enormous basic amounts), and the sales base dwarfs everything — which is why defence economists fight hardest over what counts as “affected sales” (which products, which geography, internal sales or not) rather than over the gravity percentage.

How does the US criminal system differ?

Fundamentally: cartels are felonies. Corporate fines start from the statutory maximum of $100 million or twice the gross gain or loss — the alternative that produced the record $925 million AU Optronics fine and half-billion-dollar forex pleas. The Sentencing Guidelines then compute a range from 20% of affected commerce adjusted by culpability scores that reward compliance programs and cooperation and punish leadership and obstruction.

The sharper edge is individual: prison is the norm for convicted executives — sentences have lengthened to a two-year average, foreign nationals are extradited or arrested in transit, and border watchlists make cartel indictments travel bans for life. This is the fact that reorganises deterrence: European fines are a corporate finance problem; American liability is a personal one. Any cartel touching US commerce — and “touching” is read broadly through import commerce and effects — converts every participating employee into a defendant-in-waiting, which is why global cartels are usually confessed in Washington first.

How does Türkiye calculate cartel fines?

Within Law No. 4054’s 10% ceiling, the Rekabet Kurulu applies its fining regulation: a base rate set for cartels as a proportion of the relevant turnover, escalated for duration bands and adjusted for aggravating factors (recidivism, continuing after the investigation opens, coercion) and mitigating ones (cooperation beyond legal duty, minor role, public-authority encouragement). The 2024 revision of the fining regulation widened the Board’s discretion — moving away from rigid arithmetic toward EU-style individualised assessment — while keeping the statutory cap.

Distinctive Turkish features: personal fines on managers and employees who played a decisive role (a real, applied power); the interaction with settlement (10–25% reduction) and leniency (up to full immunity) introduced into routine practice since 2021; and the banking, cement and labour-market decisions showing the Board’s willingness to fine at scale across sectors. For groups with Turkish revenues, the planning message mirrors the EU’s: the cap runs on Turkish turnover of the undertaking, and repeat findings escalate quickly.

What do settlement procedures add?

A trade of procedure for money. In the EU settlement track, parties acknowledge the infringement on a streamlined record for a flat 10% reduction — stackable with leniency — and the Commission saves years of decision-drafting and appeal exposure. Most modern EU cartel decisions are hybrid or full settlements. Türkiye’s settlement mechanism, active since 2021, offers 10–25% and has been used in a steady stream of Board decisions; the UK, France and Germany run analogous procedures, and the US plea system is settlement by construction.

The strategic decision is about the record: settling forfeits most appeal rights and creates admissions that damages claimants read closely. Companies with large civil exposure sometimes litigate precisely to delay and to keep the file thin — Scania’s calculus, which failed on the fine but bought years. The general counsel’s equation is total-cost: fine saved now versus damages evidence created and optionality lost. There is no universal answer; there is a universal mistake, which is deciding by reference to the fine alone.

Why do damages now often exceed the fines?

Because every direct and indirect purchaser can claim the overcharge plus interest, and collective mechanisms have matured. The EU Damages Directive gave claimants binding infringement findings, disclosure tools and a presumption of harm; the UK’s opt-out class regime, Dutch foundations and German bundling vehicles industrialised recovery. Trucks alone generated claims dwarfing the €3.8 billion fines; interchange, forex and air cargo each built decade-long litigation industries.

For planning, treat public fines as the visible half of a roughly one-to-two ratio, jurisdiction depending — and note the asymmetries: interest often runs from the infringement date (decades of compounding), passing-on defences shift but rarely eliminate exposure, and umbrella damages (customers of non-cartelists who priced under the cartel’s shelter) expand the claimant pool beyond your own customers. The full lifecycle cost of a major cartel — fines, damages, counsel, remediation, debarment, distraction — is the number the compliance budget should be argued against.

What does the arithmetic look like on a worked example?

Take a supplier with €200 million of annual EEA sales in the cartelised product, six years of participation, no leadership role. Basic amount at 17% gravity: €34 million × 6 = €204 million, plus a 20% entry fee (€40 million) — €244 million before adjustments. No aggravation or mitigation; the group’s worldwide turnover is €2 billion, so the 10% cap trims to €200 million.

Now the levers: second-in leniency at 40% takes it to €120 million; settlement’s further 10% (applied to the reduced figure) lands near €108 million. The identical participant without leniency or settlement pays €200 million — an 85% swing determined entirely by post-discovery choices. Run your own exposure with your affected-sales figure and honest duration estimate: the model is coarse but calibrated, and boards that see the range respond differently to compliance budgets than boards that see an abstract ‘10%’.

How much do appeals actually change?

More than folklore suggests, on identifiable grounds. The EU courts hold unlimited jurisdiction over fines and have recalculated for duration errors, sales-base overstatement, unequal treatment among participants and inadequate reasoning — reductions of 10-30% recur, and whole decisions occasionally fall (the AdSense annulment shows the evidentiary standard has teeth). Turkish administrative courts annul and remand Board decisions with real frequency, often on procedural grounds, and the Board’s recalculations on remand have produced materially different figures.

The strategic content: appeals are won on records built during the administrative phase — sales-data challenges, duration evidence and equal-treatment comparisons must be laid early, not invented before the court. And the interest asymmetry matters: in the EU, annulled fines are repaid with interest, while pending appeals require payment or guarantees up front. Litigation is a real option with a real price; it is bought at the settlement fork, which is why that decision deserves the full total-cost analysis rather than reflex.

Which mistakes inflate fines unnecessarily?

Recurring, avoidable ones. Continuing after the raid — treated as aggravation everywhere and as fresh infringement at worst. Obstruction: deleted chats and briefed witnesses convert procedural risk into add-on fines and, in criminal systems, separate charges that outlive the cartel case. Uncoordinated defence: subsidiaries answering information requests inconsistently hand the authority contradictions that harden gravity assessments.

Sales-base passivity: affected-sales figures are contestable — product perimeter, geographic scope, intra-group exclusions — and defendants who concede the authority’s first cut overpay structurally. Recidivism blindness: groups acquiring companies with prior decisions inherit escalation exposure they never price. Each error is a governance failure more than a legal one; the remedy is a rehearsed investigation-response protocol owned by the board, the case for which closes our compliance-program guide.

Frequently Asked Questions