Claim No: CL-2022-000527 - [2025] EWCA Civ 1387
Court of Appeal (Civil Division)

Claim No: CL-2022-000527 - [2025] EWCA Civ 1387

Fecha: 05-Nov-2025

The inconsistency ground

The inconsistency ground

42.

The argument as to inconsistency can be summarised as follows. The Certificate referring to “Charterers’ Liability including Liabilities for damage to Hull -Class 1”, and the insuring clause in part 1, providing, as it does, for the Insurer to indemnify the Charterer against its legal liabilities under the Award, are fundamentally inconsistent with the pay first clause 30.13. The separate obligation to indemnify in part 1 fell due when the Award established the Charterer’s liability, yet the pay first clause provides, inconsistently, that the Charterer has no right to recovery under the Policy until it pays the Award. The two clauses cannot sensibly be read together. Moreover, the two clauses are in different documents. The insuring clause is in the Certificate, which refers to part 1. The pay first clause is only in the General Terms and Conditions in part 5, which, applying the hierarchy clause, should not be given effect because it conflicts with the insuring clause.

43.

At [35]-[47] of his judgment, the judge dealt with the cases cited to him on inconsistency. I shall summarise the effect of the four cases that seem to me to be most relevant: Glynn v. Margetson [1893] AC 351 (Glynn), Pagnan SpA v. Tradax Ocean Transportation SA [1987] 2 Lloyd’s Rep 342 (Pagnan), Alexander v. West Bromwich Mortgage Co [2016] EWCA Civ 496 (Alexander), and Septo Trading Inc v. Tintrade Ltd (The Nounou) [2021] EWCA Civ 718 (The Nounou) (together “the four cases”). Again, I am drawing on the judge’s treatment.

Glynn

44.

In Glynn, the bill of lading provided specifically for the carriage of a perishable cargo of oranges from Malaga to Liverpool, whilst the printed terms of the bill of lading gave the shipowner wide powers to deviate. The Court of Appeal unanimously read down the printed deviation clause to allow only calls at a particular port or ports in the course of the stipulated voyage. Lord Herschell LC said at pages 354-5 that it was “well recognised that in construing an instrument of this sort, in considering what is its main intent and object, and what the interpretation of words connected with that main intent and object ought to be, it is legitimate to bear in mind that a portion of the contract is on a printed form applicable to many voyages, and is not specially agreed upon in relation to the particular voyage”. He said that “it would be to defeat what is the manifest object and intention of such a contract” to hold that the contract permitted unlimited deviation. Instead, the court was justified in “looking at the main object and intent of the contract and in limiting the general words used, having in view that object and intent”.

Pagnan

45.

In Pagnan, the contract was for the export of Tapioca from Thailand. A special term provided that the “seller [was] to provide for export certificate”. The contract also incorporated the GAFTA 119 form, clause 19 of which provided, in essence, that, if the export was prohibited by government, the contract should be cancelled. A hierarchy clause provided that the special terms should “prevail in so far as they may be inconsistent with the printed clauses” of the GAFTA form.

46.

The Court of Appeal (Dillon, Woolf and Bingham LJJ) held that the clauses were not inconsistent and that both could be given effect. The special term imposed an absolute obligation on the exporter to obtain the export certificate save insofar as another clause relieved it from liability. In the circumstances of that case, clause 19 relieved the exporter from liability if they could not provide a certificate because of an impediment falling within the carefully defined ambit of clause 19.

47.

Bingham LJ’s judgment includes two passages that are instructive in this case. At pages 349-350, he said that it would be quite wrong to approach the interpretation of the contract “with any predisposition to find inconsistency between the special condition and clause 19”. “By including the inconsistency clause, the parties have acknowledged that there may be. One should, therefore, approach the documents in a cool and objective spirit to see whether there is inconsistency or not”.

48.

Bingham LJ explained that it was “a commonplace of documentary construction that an apparently wide and absolute provision is subject to limitation, modification or qualification by other provisions. It does not make the latter provisions inconsistent or repugnant”. Having considered a number of other authorities, Bingham LJ said that “[i]t is not enough if one term qualifies or modifies the effect of another; to be inconsistent a term must contradict another term or be in conflict with it, such that effect cannot fairly be given to both clauses”. That point was echoed by Dillon LJ at page 353 as follows:

What is meant by an inconsistency? Obviously there is inconsistency where two clauses cannot sensibly be read together, but can it really be said that there is inconsistency wherever one clause in a document qualifies another? A force majeure clause, or a strike and lockout clause, almost invariably does qualify the apparently absolute obligations undertaken by the parties under other clauses in the contract; so equally with an extension of time clause, for instance in a building contact. So equally, with a lease, the re-entry clause qualifies the apparently unconditional demise for a term of years absolute, but no-one would say that they were inconsistent.

Alexander

49.

In Alexander, the Court of Appeal (Sir Brian Leveson P, Sharp and Hamblen LJJ) applied both Glynn and Pagnan, but decided that, in the circumstances of that case, there was an inconsistency between two of the special terms of a tracker mortgage, which provided for a fixed rate of interest until June 2010 and thereafter a variable rate of 1.99% above base rate, and terms in the mortgage booklet that had been incorporated into the contract by reference. The contract included a hierarchy clause giving primacy to the special terms over the booklet in the event of any inconsistency.

50.

Hamblen LJ held that two terms in the booklet could not be read fairly and sensibly together with, and were inconsistent with, the main purpose or object of the contract, namely to provide the mortgage product described. Clause 5 of the booklet provided that the variable rate of interest could be varied by the lender at any time for any of 8 listed reasons or by the lender giving advance notice in writing. Clause 14 of the booklet provided that the borrower would be obliged to repay the loan in full if any of a number of listed events occurred, including the lender giving one month’s notice.

51.

Hamblen LJ explained the process at [41] noting that Pagnan had made clear that inconsistency extended to cases where clauses could not “fairly or sensibly” be read together, in addition to cases where there was “a clear and literal contradiction”. He said that the question should be approached “having due regard to considerations of reasonableness and business common sense”.

52.

Hamblen LJ said at [46] that the principle in Glynn depended on being able to identify “the main purpose” or the “main object and intent” of the contract, which itself depended on the interpretation of the contract as a whole considered in its proper context.

53.

At [58]-[62], Hamblen LJ held there were three grounds on which it could be said that clause 5 was inconsistent with the mortgage product description. That description was fundamentally transformed and negated, not merely modified or qualified. Hamblen LJ accepted that “one way of testing whether clauses can be ‘fairly’ or ‘sensibly’ read together is by seeking to put them together in a single clause”. Hamblen LJ also held at [83] that clause 14 was negation of the obligation to provide the loan for 25 years, not modification or qualification.

The Nounou

54.

The Court of Appeal (Moylan, Males and Phillips LJJ) once again endorsed the principles in these cases in The Nounou. The special terms in a “Recap” provided that the quality and quantity of the fuel oil being sold should be ascertained at the loadport by a mutually acceptable first-class independent inspector “[s]uch result to be binding on parties save for fraud or manifest error”. The Recap also contained a hierarchy clause applying the BP 2007 General Terms and Conditions for fob sales: “[w]here not in conflict with the above”. Clause 1 of the BP 2007 conditions provided, in contrast, that certificates of quantity and quality were binding on the parties for invoicing purposes (requiring payment of the invoiced amounts with a right of subsequent adjustment), but not otherwise. Males LJ held that the clauses could not be fairly and sensibly read together. Clause 1 deprived the Recap clause of all practical effect.

55.

Males LJ summarised the state of the law at [28]. I entirely endorse his summary, which was as follows:

Thus there is a distinction between a printed term which qualifies or supplements a specially agreed term and one which transforms or negates it. In order to decide on which side of this line any particular term falls, the question is whether the two clauses can be read together fairly and sensibly so as to give effect to both. This question must be approached practically, having regard to business common sense, and is not a literal or mechanical exercise. It will be relevant to consider whether the printed term effectively deprives the special term of any effect (some of the cases describe this as the special term being “emasculated”, but in my view it more helpful to say that it is deprived of effect). If so, the two clauses are likely to be inconsistent. It will also be relevant to consider whether the specially agreed term is part of the main purpose of the contract or, which is much the same thing, whether it forms a central feature of the contractual scheme. If so, a printed term which detracts from that scheme is likely to be inconsistent with it. Ultimately, the object is to ascertain the intention of the parties as it appears from the language in its commercial setting.

Discussion of the Inconsistency Ground

56.

The judge dealt with the inconsistency ground at [47]-[59]. The arguments were put rather differently before the judge, but on the inconsistency ground itself, the judge held at [47] that Glynn, Pagnan, Alexander and The Nounou “concerned parties who had agreed a set of main or bespoke terms setting out the individualised terms of their bargain, and incorporated a set of pre-printed terms drafted for general use”. He said that that was “a context which [was] particularly favourable to “privileging” the bespoke term over the boilerplate”, because (i) the boilerplate was not drafted with the particular bargain in mind, and (ii) boilerplate terms were likely to be less intensively reviewed during negotiation. At [48]-[49], the judge held that the court required a stronger case to read down or read out a clause appearing in the same document as the allegedly inconsistent clause, because of these two features. The court was reluctant to read out one of two clauses in a contract (see Lewison on The Interpretation of Contracts 7th edition at [9.101] ff and Leggatt J in Scottish Power UK Plc v. BP Exploration Operating Co Ltd [2015] EWHC 2658 (Comm) at [80]).

57.

On the specific inconsistency arguments before him, the judge held first at [52] that the Owner and the Club could not bring themselves within the more favourable context for establishing and resolving an inconsistency, because the Certificate incorporated the entirety of the Booklet and “[n]o sensible reader … could have imagined that all the terms (or even all of the significant) terms of the Policy were to be found on the face of the Certificate”. In other words, the insuring clause referred to in the Certificate and the pay first clause were to be regarded as being of equal status, since they appeared in the same document.

58.

Secondly, the judge held at [53]-[54] that there was no inherent inconsistency between the insuring clause and the pay first clause. Pay first clauses traditionally made enforcement of the obligation to pay the indemnity conditional on prior discharge of that liability by the insured (see Lord Mustill at page 387 in Charter Re v. Fagan [1997] AC 313). The Fanti and Padre Island and section 9 of the 2010 Act made clear that pay first clauses could co-exist with the main purpose of liability insurance in the context of marine insurance and P&I Clubs. The clauses could be read together under Hamblen LJ’s “single clause” test (see [53] above). Apostolos Konstantine Ventouris v. Trevor Rex Mountain (The Italia Express (No 2)) [1992] 2 Lloyd’s Rep 281 (per Hirst J) did not support the inconsistency argument. The judge was not minded to follow Rein J in the Supreme Court of New South Wales in Lambert Leasing Inc v. QBE Insurance Ltd (No. 2) [2016] Lloyd’s Rep IR 163, [15]-[16], who held that a pay first clause as part of the insuring clause was “inherently inimical to the concept of insurance”, for the reasons he had given.

59.

Thirdly, the judge held at [55] that the fact that the insuring clause gave rise to a liability to indemnify when the Award was pronounced did not give rise to inconsistency with the pay first clause, because, in essence, the pay first clause was dealing with a “right to recovery”, rather than a “right to indemnity”. Coburn v. Colledge [1897] 1 QB 702 at page 705 and Rolls-Royce Holdings Plc v. Goodrich Corporation [2023] EWHC 1637 (Comm) at [236] demonstrated that there was no sensible inconsistency between a clause providing for a particular obligation to accrue at one point in time, and another clause giving a defence to enforcement until some further requirement is met. There were anyway several provisions in part 5 of the Policy that had the effect of preventing a right to indemnity accruing or rendering it unenforceable after it had arisen or even after it had become enforceable (sections 28.1.2, 30.1, 30.2, 44(a) and 44(b)).

60.

Finally, at [56]-[57], the judge rejected the alleged conflict between the pay first clause and the Insurer’s right to terminate the Policy for the Insured’s insolvency in sections 31 and 32 alongside the preservation of the Insured’s rights to indemnity for incidents occurring prior to termination. This argument was not renewed before us.

61.

In my judgment, this case is, contrary to what the judge decided, to all intents and purposes, in the same category as Glynn, Pagnan, Alexander and the Nounou. The hierarchy clause 25 (see [15] above) provided that part 1 “shall prevail over the general terms and conditions [in part 5] in the event of a conflict between them, but any terms appearing in the Certificate of Insurance shall prevail above all others”. Whether the insuring clause is regarded as being in the Certificate or part 1, it prevails, in the event of inconsistency, over the general terms and conditions in part 5, including the pay first clause. The judge seems to have lost sight of the hierarchy clause, when he said at [52] and [47] that this was not a case where the “more favourable context for establishing and resolving an inconsistency” applied, because this was not a case where “parties who had agreed a set of main or bespoke terms … [had] incorporated a set of pre-printed terms drafted for general use”.

62.

Whilst I accept that the Booklet included both the insuring clause in part 1 and the pay first clause in part 5, the essential nature of the part 1 insurance was made clear in the specifically agreed Certificate by the reference to the type of insurance as “Charterers’ Liability including Liabilities for damage to Hull - Class 1”. That is not the same as the facts in the four cases, but it means that the principles they adumbrated should be applied here. I would not apply the more stringent test referred to by the judge at [48]-[49] (see [56] above).

63.

Even applying the test of inconsistency adumbrated in the four cases and summarised by Males LJ at [28] of The Nounou, I would still hold that there is, in this case, no inconsistency or conflict (the word used in section 25) between the Certificate and the insuring clause on the one hand and the pay first clause on the other hand. There are the following main reasons:-

i)

The pay first clause does not negate the insuring clause. It qualifies and supplements it, admittedly in a very significant way. But, as Bingham LJ explained in Pagnan, even though the hierarchy clause acknowledges that there may be inconsistency, one should not approach the exercise of interpretation with any predisposition to find inconsistency. Moreover, it does not make the qualifying provision inconsistent or repugnant, just because a wide and absolute provision is subject to limitation, modification or qualification.

ii)

In this case, the indemnity, as the judge said, fell due when the Award was made, but that indemnity could not be enforced until the insured had paid the claim. That is a qualification, not a negation of the indemnity.

iii)

I have also considered, from the point of view of business common sense, the two questions of whether the effect of the pay first clause is: (i) to emasculate the insuring clause or to deprive the insuring clause of all practical effect, or (ii) inconsistent with the main purpose or object of the insurance. It does not seem to me that either question can be answered positively. As I have explained at [31]-[41] above, the legal background to pay first clauses demonstrates that they are a qualification to the insuring clause, not a complete negation of it. They are commonly used in the insurance and reinsurance industry (Lord Mustill recorded the submission that they were “long-established contractual provisions” in Charter Re v. Fagan). Parliament decided not to outlaw their use in contracts of marine insurance save to the limited extent I have outlined above (death and personal injury cases). The insuring clause has its full effect whenever the insured discharges a judgment or award. The mischief of the pay first clause described by Lord Mance and others only kicks in when insolvency of the insured supervenes. As it seems to me, it would be illogical and excessive to hold that the pay first clause emasculated the insuring clause, deprived it of all practical effect, or that the pay first clause was inconsistent with the main purpose or object of the insurance. In Bingham LJ’s words in Pagnan: “[i]t is not enough if one term qualifies or modifies the effect of another; to be inconsistent a term must contradict another term or be in conflict with it, such that effect cannot fairly be given to both clauses”. That is not the case here.

iv)

It follows from what I have said that, in the sense that Hamblen LJ used those words at [41] in Alexander, the pay first clause can be fairly and sensibly read together with the insuring clause. This can be verified by applying the ‘single clause’ test suggested by Hamblen LJ.

64.

Accordingly, applying the principles established in the four cases and summarised at [28] in Males LJ’s judgment in The Nounou, I would dismiss the inconsistency ground of appeal.