Fewer fines account for most of the fall in ESMA's 2025 penalty total

The €18.3 million fall in administrative fines reported to ESMA for 2025 was overwhelmingly associated with fewer fines, rather than a comparable fall in the average fine. Reconstructing the two annual reports shows that the fine count dropped from 615 to 510, or 17.1%, while the arithmetic mean fell just 1.4%, from about €162,904 to €160,581. Holding the 2024 mean constant, the lower count accounts for €17.1 million, or 93.5%, of the decline. This is an arithmetic decomposition of the published totals, not a causal finding about why authorities imposed fewer penalties.[1][2]

That movement is much larger than the 2.1% decline in ESMA's headline count of all administrative sanctions and measures. Non-fine actions increased from 360 to 445 and offset most of the reduction in fines. The two reports therefore describe a change in the composition of reported enforcement, which a comparison of the headline totals alone would largely miss. They do not establish that violations became rarer, that regulators became more lenient or that the same firms received smaller fines.[1][2]

The denominator changes the story

ESMA's 2025 report, published in October 2026, records €81,896,178 in administrative fines. The corresponding 2024 total is €100,186,062. The €18,289,884 difference is an 18.3% reduction. The reports separately identify the number of administrative fines, making it possible to calculate a mean using the matching value and count. Dividing the full headline action count into the fine value would mix monetary penalties with other measures and produce a different, unsuitable measure.[1][2]

The non-fine increase is also material in its own right. Its 85 additional actions offset about 81% of the 105-fine decline in the total count. Orders to cease a specified activity rose from 99 to 162, an increase of 63. The residual category labelled other sanctions and measures rose from 158 to 161, while the remaining non-fine categories together rose from 103 to 122. These figures explain why the all-action count stayed comparatively steady without requiring a steady number of monetary penalties.[1][2]

This is not a count of independent enforcement cases. A single case can produce more than one measure, and measures can have different targets and legal consequences. ESMA's aggregation cannot be converted into a number of offending brokers by relabelling its rows as cases. Nor can a cease order, a public statement and a fine be assumed to have equal effect simply because each contributes a count. The composition needs to remain visible alongside the total.[1][2]

Separating the count effect from the mean-value effect

The bridge uses a count-first convention. The first term takes the reduction of 105 fines and values it at the 2024 mean of €100,186,062 divided by 615. That gives minus €17,104,937. The second term multiplies the change in mean value by the 510 fines reported for 2025, producing minus €1,184,947. Their sum reconciles to the reported value decrease, subject to rounding. All amounts are nominal euros; the comparison does not adjust penalties for inflation.[1][2]

Reversing the order would distribute the interaction differently between the two terms. The headline 93.5% is therefore tied to an explicit convention, rather than being a unique causal allocation. The simpler observations are independent of that convention: the fine count fell 17.1%, and the mean fell 1.4%. Together they show that a large reduction in the mean is unnecessary to reconcile the decline in aggregate monetary penalties. The bridge quantifies that relationship in euros.[1][2]

The mean is not the typical firm's penalty. A small number of large fines can pull it far above the median, and the reports' aggregates do not provide the case-level distribution needed to estimate a typical comparable outcome. Two German fines of €7.4 million each alone account for about 18.1% of the 2025 fine total. That concentration is sufficient to make a simple mean sensitive to a few decisions. It does not invalidate the arithmetic, but it limits what can be inferred about ordinary cases.[1]

A matched-case assessment of severity would need much more: the legal provision, seriousness and duration of the conduct, harm, cooperation, settlement, economic benefit and the relevant sanctioning powers. Changes in those factors can alter average penalties without a change in a regulator's general approach. The current calculation is deliberately an aggregate comparison. It identifies the immediate numerical source of the lower total while leaving the reasons for the change open.

The decline spans the two largest fine categories

Market Abuse Regulation and MiFID II/MiFIR fines together fell by 81, accounting for 77.1% of the overall decline of 105. Their combined count moved from 448 to 367. This supplies a second check on the aggregate finding: the reduction is visible in the two largest fine-count categories, rather than arising solely from a small newly added regime. It still does not identify which institutions, conduct or enforcement decisions produced the change within those categories.[1][2]

The four frameworks in the table together recorded 493 fines in 2025, compared with 595 in 2024. Other regimes contributed 17 and 20 respectively. The reports therefore permit a broad composition test without claiming a fixed population of investigations. ESMA's coverage also evolves as legislation takes effect and reporting begins. Its 2025 account includes the first sanctions under MiCA and SFTR, while recording none under certain other frameworks. A year-to-year total should be read in that changing perimeter.[1][2]

The jurisdictional distribution adds another limit. Germany, France and Spain reported fine values of €23.3555 million, €17.83 million and €12.545 million respectively. Together they account for about 65.6% of the 2025 total. A movement in a few large national decisions can therefore materially affect the European aggregate. Comparing raw national totals as if they measured equal enforcement effort would ignore differences in market size, supervised populations, conduct, powers and case timing.[1]

Keep fines separate from other monetary amounts

ESMA's wider pecuniary totals include amounts outside the administrative-fine category. For 2025, the report separately identifies disgorgement and penalties based on gains or avoided losses. Adding those to fines produces a larger total, €83,753,645. Using that amount with the 510 administrative-fine count would no longer reproduce the fine mean calculated here. The distinction is particularly important when comparing settlement shares, which can be expressed against the wider monetary total.[1]

The headline figures also concern the specified EU financial-market legislation in ESMA's exercise. They are not a complete inventory of every financial penalty imposed in Europe, and administrative and criminal measures should not be merged casually. A national anti-money-laundering case or another authority's penalty can sit outside this particular perimeter. An apparent omission from the ESMA total is therefore not, without further examination, evidence that a sanction went unreported or that a regulator took no action.[1][2]

For a cross-border broker, the most useful distinction is between what changes a compliance obligation and what describes enforcement after the event. A lower annual fine total does not amend rules on market abuse, conduct or transaction reporting. Nor does a relatively stable mean tell a firm the penalty it would face in a future matter. The distribution of actions can help focus research into supervisory priorities, but an operational risk decision still needs the applicable regime, jurisdiction and facts.

What would establish the underlying cause?

Several explanations are consistent with the published comparison. Authorities may have concluded fewer fine-bearing matters, more cases may have resulted in non-monetary measures, or the timing and mix of decisions may have changed. Those possibilities are not interchangeable, and the two annual aggregates cannot choose among them. A proper causal test would track comparable matters through opening, investigation, decision, appeal and settlement, while distinguishing a calendar-year delay from a persistent change in enforcement practice.

A useful next release would therefore make the relationship between cases and measures easier to follow, while protecting confidential material. Comparable distributions of penalty size and clearer case-type breakdowns would show whether the near-stable mean conceals offsetting changes. A longer run of consistently defined data could also reveal whether the 2025 shift toward non-fine measures persists. The present reports provide a starting point for those questions, not a substitute for their answers.

The established finding is narrower and more informative than saying penalties fell. Administrative fines became substantially fewer, their mean value changed comparatively little, and more non-fine measures kept the all-action total close to the previous year. On the stated count-first calculation, fewer fines account for 93.5% of the €18.3 million reduction in fine value. That is a documented change in the composition of reported enforcement, without an unsupported conclusion about deterrence, misconduct or regulatory leniency.[1][2]

Report an error or submit an update