The Supreme Court has recently considered penalty clauses in a combined judgment which covers two cases1. Despite speculation that penalty clauses would be abolished or extended, it was held by majority of the Supreme Court that they would remain. However they have set out a new approach to the identification of penalty clauses which moves away from the concept of ìgenuine pre-estimate of lossîwhich has been followed, perhaps too slavishly, since the 1915 case of Dunlop Pneumatic Tyre Company Ltd. v New Garage and Motor Company Ltd2 (Dunlop)

What is a penalty clause?

The parties to a contract may include a clause that provides that, in the event of a breach, the breaching party shall pay a specified sum to the innocent party. If such a clause satisfies the legal definition of a penalty clause, the amount specified is not recoverable. Instead, the innocent party must prove its actual losses in order to claim damages. If the clause is a valid liquidated damages clause, it is enforceable irrespective of the loss actually suffered.

The cases

The first case, Cavendish Square Holdings BV v Talal El Makdessi involved an agreement for a share sale from Talal El Makdessi (Makdessi) to Cavendish Square Holdings BV (Cavendish). It included two clauses to protect Cavendish against competitive activity. The first was in reality a price-adjustment clause, regulating the measure of compensation for breach of restrictive covenants. The second required a defaulting shareholder to sell shares at a substantially reduced price, which did not factor in goodwill. The second case, ParkingEye Limited v Beavis, concerned a charge for overstaying the free period of parking in a car park.

The new approach

The question to be decided in both cases was when a contractual provision should be considered to be penal. The majority of the Lords felt that the law to date had categorised penalty clauses artificially, and it was unsatisfactory to say that a liquidated damages provision was either a penalty or a deterrent which genuinely pre-estimated the loss. In particular, the four tests in a speech by Lord Dunedin in Dunlop and applied repeatedly in subsequent cases had too often been treated as a code when in fact they should be treated as considerations, not rules. That said, they did contain the essence of the test to determine if a liquidated damages clause was penal i.e. whether the provision was a ìsecondary obligation which imposes a detriment out of all proportion to any legitimate interest of the innocent part in the enforcement of the primary obligationî.

To determine whether a liquidated damages clause was penal required an understanding of the nature and extent of the innocent partyís interest in the performance of the relevant obligation. In less straightforward cases, a broader approach had been preferred: it was not confined to considering the compensation awarded by the clause but also any wider interest for which the clause was intended, such as protection of brand, reputation or good will. The true test should be whether the detriment caused to the contract breaker by paying the sum specified was in proportion to the legitimate interest of the innocent party in enforcing the obligation breached.

Lord Mance suggested a two-fold test:

  1. Consider whether a legitimate business interest is served and protected by the clause breached.
  2. Consider whether the provision to protect such interest is extravagant, exorbitant or unconscionable.


It was held that none of the clauses considered were penalty clauses. In Cavendish, the clauses were viewed as a package and intended to protect the business as a whole. The price-adjustment clause was carefully negotiated between informed and legally advised parties at armís length. The second clause, although forcing transfer of assets without adequate consideration was viewed in the context of being part of the mechanism that brought to an end the shareholder relationship, reshaping the primary relationship, and was a ìnaturalî provision to include.

In ParkingEye although there were no ascertainable damages, the charge was seen as an understandable ingredient in a scheme serving legitimate interests.


This case has replaced a somewhat blunt-instrument test with a more sophisticated analysis which makes it likely that fewer liquidated damages clauses will now be considered unenforceable penalty clauses (although it was acknowledged that in the case of a straightforward damages clause, the interest rarely extends beyond compensation for actual losses suffered). The application of the penalty rule will still turn on questions of drafting. Parties agreeing liquidated damages clauses should consider setting out specific details of the legitimate interest which the clause serves to protect and take care to ensure that the amount is not unconscionable or extravagant by reference to some norm.