A Rigged Benchmark, an Invalid Contract? The Attribution Problem after Livronsa

A rigged benchmark does not necessarily mean an invalid contract. In Case C-60/25 Livronsa (2026), the Court of Justice of the European Union rejected the automatic extension of antitrust nullity to a mortgage term referencing Euribor. The argument developed here is a separate contractual one: a contracting party involved in producing a benchmark implicitly undertakes, as a matter of good faith, not to manipulate it. That undertaking concerns its own conduct. It does not make it the guarantor of every other contributor's honesty.

The Court’s ruling

In Livronsa, the Court considered an Italian mortgage whose interest rate referenced Euribor. The borrower invoked the European Commission’s Euro Interest Rate Derivatives cartel decisions to challenge the interest term. The lender had not participated in the cartel, and the loan neither formed part of nor implemented the prohibited arrangement.

The Court held that the automatic nullity under Article 101(2) of the Treaty on the Functioning of the European Union did not extend to that term: the parties were outside the cartel, and their contract concerned a different market (paragraphs 35–38). The Commission decisions bound national courts only within the scope of the infringement established. The finding of a restriction by object did not require proof of actual effects and did not, by itself, establish specific effects on Euribor’s level (paragraphs 42–45).

This was a ruling about antitrust nullity and the reach of the Commission’s findings, not a general endorsement of Euribor clauses. The Court did not establish an implied contractual duty not to manipulate. That is the distinct proposition defended below.

The contractual argument

A benchmark-linked term distinguishes the rule for determining the price from the data used to apply that rule. The parties agree, for example, that interest will equal Euribor plus a margin. The benchmark supplies a changing input to an agreed formula. This arrangement allows them to adapt a long-term relationship without renegotiating whenever market conditions change.

If the input is distorted, the formula may still have been applied correctly. The problem is not necessarily defective drafting or incorrect arithmetic, but interference with the information on which performance depends. Manipulation does not, therefore, automatically invalidate the price-setting term; a separate legal basis is needed to establish that consequence. An economic connection between contracts using the same benchmark does not establish a shared legal defect.

The more relevant question is what the parties have undertaken towards one another. I argue that a party which helps produce the benchmark governing its own contract assumes an implied obligation not to manipulate it. The obligation protects the agreed method of price determination, rather than any particular price or economic outcome.

Why good faith matters

The point is easiest to see through an analogy. Suppose parties entrust a valuation expert with setting their contractual price. One party then bribes the expert. The objection is not simply that bribery is unlawful. By corrupting the agreed decision-maker, that party has also acted inconsistently with its own bargain. No express undertaking not to bribe the expert should be necessary.

The same reasoning applies when a benchmark contributor uses the index in its own contracts. Agreeing to an external pricing reference while secretly manipulating its production substitutes covert influence for the procedure the parties selected. Good faith supplies the obligation to respect that procedure. The obligation applies throughout the period in which the benchmark determines the contractual price; it is not merely a statement about the contributor’s conduct when the agreement was signed.

This does not require contributors to have no economic influence on the benchmark. Their legitimate transactions or compliant submissions may affect its value. Nor does it require a wholly mechanical benchmark free from judgement. The distinction is between participating according to the benchmark's rules and deliberately corrupting their operation. The contractual undertaking is to refrain from the latter.

The doctrinal route will vary across jurisdictions. Article 5.1.2 of the UNIDROIT Principles of International Commercial Contracts, for example, identifies contractual purpose and good faith as sources of implied obligations. Elsewhere, an implied term must satisfy the governing law’s own requirements. The argument is therefore functional, not a claim that every legal system recognises an identical general duty of good faith.

Attribution without a guarantee

Identifying the obligation also identifies its limits. A bank that uses a benchmark does not breach a duty not to manipulate merely because somebody else rigs it. Even participation in the benchmark’s production is insufficient: membership of a contributing panel is not participation in the misconduct. The alleged manipulation must be legally attributable to the contractual counterparty.

To infer a breach from the benchmark’s corruption alone would silently replace a duty governing the party’s own behaviour with a guarantee of the entire production process. Those are different contractual commitments. Selecting a benchmark does not, without more, show that the selecting party assumed the latter. It may itself be exposed to manipulation by other contributors.

Conversely, where a contracting contributor manipulates the benchmark, its conduct can have an independent contractual significance even if the cartel’s immediate target was another market. The obligation concerns the integrity of the agreed pricing process, not the market in which the manipulator hoped to profit. This contractual analysis is the argument advanced here, not a further holding attributed to Livronsa.

Nor does attribution turn breach into automatic invalidity. The consequences remain governed by the applicable law. Other contractual commitments or mandatory rules may also matter; absence of participation is not a universal defence to every possible challenge. The narrower proposition is that third-party rigging alone establishes no breach of this implied duty.

Livronsa blocks a shortcut from an antitrust infringement to contractual nullity. The contractual argument explains why the inquiry must instead identify the counterparty’s own undertaking and conduct. A party must not corrupt the mechanism it has agreed to use. It does not follow that, by agreeing to use it, the party insured its counterparty against manipulation by everyone else.

Ernesto Suárez Puga is a Teaching Fellow at Esade Law Faculty.