Cavendish Test for Greater Autonomy


The question concerning the interpretation of the liquidated damages (LD) clause has often been a point of dispute before an Arbitral Tribunal. Such provisions have been classified into two types by the courts in the UK: The unenforceable penalty clauses and the LD clause, which are enforceable.1 In India, Section 74 of the Contract Act2 explicitly bars any LD to be paid which is in the nature of penalty. However, the Act does not define “penalty”, and therefore, Indian courts usually rely on English judgment for its definition.

A clause is considered to be in the nature of penalty if it provides for “a payment of money stipulated as in terrorem of the offending party”3 or, if the clause’s contractual nature is “deterrent rather than compensatory”.4 On the other hand, a clause is said to be a LD clause if it is a genuine endeavour by the parties to stipulate the loss arising out of the breach in advance. The nature of the clause also depends on its construction and the encompassing circumstances during the time of forming the contract or at the time of doing the material variation in the contract.5

In the UK, the rules developed by Lord Dunedin in Dunlop Pneumatic Tyre Co. Ltd. v. New Garage & Motor Co. Ltd. (Dunlop)6 were the guiding test for deciding whether a clause is in the nature of the penalty or not. It focused on the question of “whether the clause represents the genuine pre-estimate of loss or not”. Over time, the contracts have evolved and have become more complex, which questions the relevancy of the test. More recently, in Cavendish Square Holding BV v. Makdessi (Cavendish)7, the UK Supreme Court (UKSC) has emphasised that where a clause does not represent the genuine pre-estimate of loss, it cannot be regarded as penalty if there is “commercial justification” for it.

The courts in India are still reluctant to incorporate the new test propounded in Cavendish8, which respects the party’s autonomy. In arbitration, the freedom of the parties to define their relationship is the most fundamental principle. This piece argues that since the relationship between Arbitral Tribunals and national courts oscillates between forced cohabitation and true partnership9, it is the right time to adopt Cavendish10, which will be proven to be a pro-arbitration approach. For convenience, this piece analyses the penalty rule by bifurcating it into two periods i.e. pre-Cavendish and post-Cavendish development. Then, this piece argues why India should adopt the rules propounded in Cavendish11.

Pre-Cavendish development

The sources of the penalty rule in the common law can be traced in the medieval penal bond, firstly in the equity courts and thereafter in the royal courts.12 However, the most venerated test for determining penalty clause was propounded by Lord Dunedin in Dunlop Pneumatic Tyre Co. Ltd. v. New Garage & Motor Co. Ltd.13 where he formulated the following four rules for the construction of LD. “A clause is said to be in nature of a penalty if:

(a) The sum pre-estimated is unconscionable and extravagant compared to the greatest loss that could conceivably be proven to arise from the breach.

(b) The breach consisting only of not paying a certain amount, and the sum stipulated is a sum greater than the sum which ought to have been paid.

(c) A single lump sum is made payable on the occurrence of one or more or all of several events, some of which may occasion serious and other but trifling damage.

(d) The sum stipulated is not a genuine pre-estimate of damage in cases where it is impossible to make a precise pre-estimation.”

This test may be relevant for simple damages clauses in standard contracts, however, it is difficult to apply this test in complex contracts. In complex cases where technical expertise is needed to understand different aspects of a contract, it is very subjective and complex to determine what amounts to “genuine pre-estimate of loss”.

Later, Lord Woolf, in Privy Council in Philips Hong Kong Ltd. v. Attorney General of Hong Kong Co.14 said, “The Court has to be careful not to set too stringent a standard and bear in mind that what the parties have agreed should normally be upheld because any other approach will lead to undesirable uncertainty, especially in commercial contracts.” Lord Colman in Lordsvale Finance Plc. v. Bank of Zambia15 was to examine a simple form of provision in a syndicate loan agreement which provided for interest to be paid at a greater rate during any period in which the borrower was in default. He observed that simply because the provision for the payment of a sum in case of breach was not a “genuine pre-estimate of damages”, it cannot be said to be a penalty clause. He further observed, … no reason in principle why a contractual provision the effect of which was to increase the consideration payable under an executory contract upon the happening of a default should be struck down as a penalty if the increase could in the circumstances be explained as commercially justifiable, provided its dominant purpose was not to deter the other party from breach…..16

Post-Cavendish development

Cavendish17 consisted of two appeals in the UKSC: Cavendish Square Holding BV v. Makdessi and ParkingEye Ltd. v. Beavis. This piece focuses on the first one since it deals with the commercial contract. Makdessi had agreed to sell Cavendish Square Holding a controlling interest in a business. Under the terms of the agreement, Makdessi would forfeit the right to collect the last two instalments of the money Cavendish was required to pay if certain restrictive covenants were broken. The provision further stated that Makdessi might be forced to sell Cavendish the remainder of his share at a price that did not include goodwill towards the company. Makdessi said that the provisions requesting enforcement were penalty clauses after breaking the terms of the agreement.

Following the facts illustrated above, the UKSC unanimously felt the need for further refinement in the pre-Cavendish position. While rejecting the suggestion of total abolition of the pre-Cavendish position, it provided a reformed test applicable to the clauses which amount to the secondary obligation imposed on the contract breacher. It provided the test in two limbs:

(a) Whether any “legitimate business interest” is protected by the clause (first limb).

(b) If so, is the provision made in the clause “exorbitant, extravagant or unconscionable” or is there some wider “commercial or socio-economic justification” for the clause (second limb)?

Contrary to the strict bar against all covenants of a deterrent nature in Dunlop18, the test in Cavendish19 advocates that deterrence might not compulsorily be considered as penal in the cases in which the party establishes the presence of “legitimate interest” in securing the performance of the contract which goes beyond the mere right of recovering damages. This test in Cavendish20 renders additional protection to the covenants that might otherwise be considered as a penalty under the old Dunlop test but are considered “commercially justifiable” if viewed in the light of the brisk development of present time business and commerce.21

Why should Cavendish prevail over Dunlop?

The Dunlop test is susceptible to three main criticisms:

  1. That there is the absence of consideration appropriated to commercial realities.

  2. After Lordsvale22 , this test is inapplicable in complicated matters pertaining to apparently valid commercial justification with impugned clauses. This led to judicial inconsistency and vagueness.

  3. That the rigid dichotomy created i.e. “genuine” and “non-genuine” pre-estimate of loss, is misleading, artificial and arbitrary.23

This rigid dichotomy has created a dilemma for the judiciary. “Miller” points out that Lord Dunedin’s postulation presumes that stipulated damages can either be a penalty or LD. However, there may be cases where one function may be more dominant than another, and it is not every time the situation that the other function is totally absent. Therefore, both functions are not necessarily mutually exclusive. It is very well possible that a clause may have an element of deterrence, and at the same time it may be a “genuine pre-estimate of loss”.24 This situation may arise due to the under compensatory character of contract damages. A number of damages remain unpaid, such as loss of productivity, lost opportunity, internal cost and non-monetary losses like emotional distress.25

Critics have contended that the Dunlop test’s endurance stems from the court’s reluctance to cede its authority to make decisions.26 Having said that, the circumstances surrounding the contract’s establishment are not completely meaningless. When parties are fairly informed, well-informed, and possess comparable or nearly equal bargaining power in a contract, a strong initial presumption should be that the parties are the best arbiters of what would be reasonable in the event of a breach of the agreement. The core ideas of contract law, “freedom of contract” and “pacta sunt servanda”, are essential to the laissez-faire approach taken by the majority of common law jurisdictions worldwide.27 In order to ensure surety and certainty, this flexibility includes the right of the contracting parties to negotiate and include clauses regarding agreed upon remedies in the event of a breach. It also considers whether the contract may be enforced. Hatzis’s argument that parties in business contexts should be deemed to have considered the benefits and drawbacks of the clause before signing the contract, as well as the court’s refusal to enforce the terms of the agreement whether they are penal or not further bolsters this line of reasoning.28

The test applicable in India

Decades after Dunlop29, the Indian Supreme Court (SC) in Fateh Chand v. Balkishan Dass30 , examined a deed of sale which provided that if the purchaser could not register the deed by the stipulated date, the earnest money and the sale price INR 1000 and 24,000 respectively, paid by the purchaser would be forfeited. The SC applied the Dunlop test and observed that the INR 24,000 stipulation was not a “genuine pre-estimate of loss” and was manifestly a stipulation in nature of penalty.

Again, in Maula Bux v. Union of India31, it was held by the Supreme Court that in cases where the parties are unable to determine the reasonable compensation, if the amount decided by the party is a “genuine pre-estimate of damages” it should be considered a reasonable compensation. Further, in Kailash Nath Associates v. DDA32 , the SC held that only those LD clauses which are “a genuine pre-estimate of damages” can be enforced as “reasonable compensation”.

Though several Indian decisions have referred to Cavendish33, none of them have completely relied on the test propounded. In Union of India v. DishnetLtd.34 , the High Court of Tripura, referring to Cavendish35, said, “though it establishes true principles with respect to such clause, the dominant test in India is still a genuine pre-estimate of damages test” and decided on the basis of that only. Again, in Electronics Corpn. of Tamil Nadu Ltd. v. ICMC Corpn. Ltd.36 , the High Court of Madras was asked to decide upon the invocation of the LD clause due to the suppliers’ failure to follow a delivery schedule. In this case also, the Court again referred to Cavendish37 but ultimately relied on “the genuine pre-estimate of loss test” for deciding the nature of the clause. Lastly, in LIC Housing Finance Ltd. v. CST38 the Customs Excise and Service Tax Appellate Tribunal again the Tribunal referred to Cavendish39 however did not discuss its implication on the case. Therefore, though the Indian courts have cited Cavendish40, but abstained from relying on the test.

Cavendish embraced in foreign jurisdictions


In Paciocco v. Australia and New Zealand Banking Group Ltd.41the High Court was asked to decide whether a bank’s credit card late fees qualified as penalties under the applicable Act. The Court determined that the motive for the imposition of late payment fees was to make up for any potential loss that could arise from the failure to pay. Despite the fact that the fee did not accurately estimate the potential loss resulting from a specific violation due to the relatively small amount of late payment, the court determined that the charge did not qualify as a penalty. The court did not entirely distance itself from the punishment rule, though.

New Zealand

The structure established by Cavendish42 was adhered to by the Supreme Court of New Zealand in 127 Hobson Street Ltd. v. Honey Bees Preschool Ltd.43 The Court created a new proportionality standard that weighs the stipulated amount against the non-breaching party’s “reasonable interest”. It acknowledged that damages recoverable under common law might not include broader economic or commercial interest protected. The “legitimate interest” of a non-breaching party is the main focus of this new approach, which rejects the Dunlop test.


In Cubic Electronic Sdn Bhd v. Mars Telecommunication-Sdn Bhd44, the Federal Court, following Cavendish45, recognised the concept of “legitimate interest” and “proportionality” while judging what should be a reasonable compensation. It held that the court must first determine whether a damages clause serves to protect any “legitimate business or commercial interest” in performance that extends beyond the possibility of monetary compensation which may result from the breach, and if so, whether the provision created to safeguard that interest is in proportion to the identified interest.


LD and contractual penalties are distinguished by the German legislation, known as the German Civil Code (BGB). In situations where the prescribed amount is “disproportionate and excessively high”, Article 343 of the BGB requires a judicial reduction; nevertheless, it also states that the evaluation must take “every legitimate interest of the oblige, not merely his financial interest” into account. This evaluation follows the logic presented in Cavendish46.


Under many national arbitration regimes, an arbitral award is subject to very limited judicial interference. This attracts the parties to choose that nation as the seat of arbitration. The Cavendish rule gives greater autonomy to parties to define their relationship in comparison to Dunlop and party autonomy is one of the cardinal principles behind the evolution of the law of arbitration. In cases where the parties are of equal or comparable bargaining power, the interference by the arbitrator or Judges by declaring any clause penalty merely on the basis of a reasonable pre-estimate of loss goes against the fundamental principle of party autonomy. As a matter of practical usage, it is noteworthy that the recently published Scottish Law Commission Report has observed, “the Cavendish test is well-received by commercial law firms and professional bodies for being highly flexible and workable in terms of providing clear guidance as to future contract drafting”.47 Various common law jurisdictions have positively responded to Cavendish and have relied on them. Therefore, it is legitimate to incorporate the test propounded in Cavendish48.

†4th year student, BA LLB (Hons.), Gujarat National Law University, Gandhinagar.

1. Golden Bay Realty Pte. Ltd. v. Orchard Twelve Investments Pte. Ltd., (1991) 1 WLR 981.

2. Contract Act, 1872, S. 74.

3. Dunlop Pneumatic Tyre Co. Ltd. v. New Garage & Motor Co. Ltd., 1915 AC 79, 86.

4. Lordsvale Finance Plc. v. Bank of Zambia, 1996 QB 752, 762 : (1996) 3 WLR 688.

5. Dunlop case, 1915 AC 79.

6. 1915 AC 79.

7. 2016 AC 1172 : 2015 UKSC 67.

8. 2016 AC 1172 : 2015 UKSC 67.

9. Nigel Blackaby, Constantine Partasides, Alan Redfern & Martin Hunter, Redfern and Hunter on International Arbitration, (5th Edn., Oxford University Press, 2009).

10. 2016 AC 1172 : 2015 UKSC 67.

11. 2016 AC 1172 : 2015 UKSC 67.

12. Larry A. DiMatteo, Civil-Common Law Divergence on Penalties: Is it a Thing of the Past? (2022) 43 Liverpool Law Review.

13. 1915 AC 79.

14. (1993) 61 BLR 41.

15. 1996 QB 752 : (1996) 3 WLR 688.

16. Lordsvale Finance Plc. v. Bank of Zambia, 1996 QB 752 : (1996) 3 WLR 688.

17. 2016 AC 1172 : 2015 UKSC 67.

18. 1915 AC 79.

19. 2016 AC 1172 : 2015 UKSC 67.

20. 2016 AC 1172 : 2015 UKSC 67.

21. Raphael Lok Hin Leung, “In Defence of the Halfway House — The Cavendish Penalty Rule Since 2015”, (2019) 13 Hong Kong Journal of Legal Studies.

22. 1996 QB 752 : (1996) 3 WLR 688.

23. Lucinda Miller, “Penalty Clauses in England and France: A Comparative Study”, (2008) 53 International and Comparative Law Quarterly 79, 82.

24. Lucinda Miller, “Penalty Clauses in England and France: A Comparative Study”, (2008) 53 International and Comparative Law Quarterly 79, 82.

25. Larry A. DiMatteo, Civil-Common Law Divergence on Penalties: Is it a Thing of the Past? (2022) 43 Liverpool Law Review , 426.

26. Mattei, Ugo, “The Comparative Law and Economics of Penalty Clauses in Contract”, (1995) 43 American Journal of Comparative Law 427.

27. Roy Goode, Commercial Law in the Next Millennium in (Sweet & Maxwell, London, 1998) p. 31.

28. Aristides N. Hatzis, “Having the Cake and Eating it Too: Efficient Penalty Clauses in Common and Civil Contract Law”, 22(4) International Review of Law and Economics 381.

29. 1915 AC 79.

30. 1963 SCC OnLine SC 49.

31. (1969) 2 SCC 554.

32. (2015) 4 SCC 136.

33. 2016 AC 1172 : 2015 UKSC 67.

34. 2017 SCC OnLine Tri 90.

35. 2016 AC 1172 : 2015 UKSC 67.

36. 2020 SCC OnLine Mad 244.

37. 2016 AC 1172 : 2015 UKSC 67.

38. 2019 SCC OnLine CESTAT 8290.

39. 2016 AC 1172 : 2015 UKSC 67.

40. 2016 AC 1172 : 2015 UKSC 67.

41. 2016 HCA 28.

42. 2016 AC 1172 : 2015 UKSC 67.

43. 2020 NZSC 53.

44. (2019) 6 MLJ 15 FC.

45. 2016 AC 1172 : 2015 UKSC 67.

46. 2016 AC 1172 : 2015 UKSC 67.

47. Scottish Law Commission, Report on Review of Contract Law: Formation, Interpretation, Remedies for Breach, and Penalty Clauses (2018) 252.

48. 2016 AC 1172 : 2015 UKSC 67.

Must Watch

maintenance to second wife

bail in false pretext of marriage

right to procreate of convict

Criminology, Penology and Victimology book release

Join the discussion

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.