In the current economic climate, with a number of shipbuilders in serious financial trouble, shipbuilding contracts are often amended. This can have a bearing on the refund guarantees provided to the buyer to enable it to recover instalments already paid in the event that the shipbuilder does not complete the building of the ship(s). It is a fundamental principle of the relationship between the buyer, builder and refund guarantor that, without the refund guarantor’s consent, any material variation of the contract that potentially prejudices the guarantor will invalidate the refund guarantee. Commercial and practical considerations, however, often dictate that the refund guarantor’s consent cannot be obtained, leading to the practice of including “anti-avoidance” clauses in refund guarantees. How can the benefit of a refund guarantee be best preserved when guaranteed obligations are amended?
The general principle – Holme v Brunskill
A surety’s obligations under a contract of guarantee are construed restrictively. The case of Holme v Brunskill1 established the principle that, absent the surety’s consent, any material variation of the underlying contract to which the principal is party will discharge the surety from liability under the guarantee. A variation is material where it is not necessarily beneficial to, or otherwise prejudices, the surety and where any benefit or lack of prejudice is not evident without enquiry. If the benefit or lack of prejudice is not self-evident, then the court will not embark on an enquiry as to whether the variation was indeed beneficial or otherwise unprejudicial to the surety.2
Excluding the effect of Holme v Brunskill: “on demand” bond v. traditional guarantee
It is possible to exclude from the beginning the effect of the law on material change if a refund guarantee, properly construed, in fact imposes primary liability on the refund guarantor (generally known as a demand guarantee or performance bond). This means that such a guarantee operates autonomously from the underlying shipbuilding contract and so any amendments to that contract, consented to or not, are irrelevant to the refund guarantor’s liability under the relevant guarantee.
Whether a payment or refund guarantee issued in favour of a buyer or builder will be considered by the English courts to be a traditional “see to it” guarantee or an “on demand” guarantee will depend upon the court undertaking an examination of the “words actually used by the parties”3 and contained within the instrument itself and “construing the instrument in its factual and contractual context having regard to its commercial purpose”.4
In Wuhan Guoyu,the Court of Appeal identified four factors that create a presumption in favour of a demand guarantee. In particular, Longmore L.J. endorsed the statement in the eleventh edition of Paget’s Law of Banking5 that “where an instrument (i) relates to an underlying transaction between the parties in different jurisdictions, (ii) is issued by a bank, (iii) contains an undertaking to pay “on demand” (with or without the words “first” and/or “written”) and (iv) does not contain clauses excluding or limiting the defences available to a guarantor, it will almost always be construed as a demand guarantee...” (the “Paget’s Presumption”).
The fourth factor giving rise to the presumption of a demand guarantee refers to the so-called “anti-avoidance” clauses (also called “anti-discharge” or “indulgence” clauses whereby amendments or variations to the shipbuilding contract are said not to discharge the guarantor from liability). However, case law has suggested that only the first three factors are important, and that the presence or absence of an anti-avoidance clause is neutral. In Gold Coast, a refund guarantee given in the context of the construction of a ship which included an anti-avoidance clause was held to be a performance bond by the Court of Appeal. The Court found that there could be a number of possible reasons for including such a clause; for example, “it could have been inserted simply to ensure that the rule applicable to true guarantees did not apply to this instrument”.6 The Wuhan and Gold Coast approaches were subsequently applied in Caterpillar Motoren GmbH & Co. KG v Mutual Benefits Assurance Co.7, where Teare J. held that, although the relevant bonds contained clauses that failed to satisfy the fourth condition of Paget’s Presumption, as a matter of construction this was not “fatal” to the application of the presumption.
Anti-avoidance clauses and the doctrine of “purview”
In cases where a refund guarantee is not classified as a performance bond, a buyer may invoke the application of any anti-avoidance clause contained in it and argue that the amendments or variations to the shipbuilding contract fall within the protections provided by such clause. Each case must be decided on the basis of the specific terms of the agreements and, if there is ambiguity as to the scope of a clause in a guarantee, that doubt will be resolved in the refund guarantor’s favour.8
Anti-avoidance clauses have been ruled ineffective in a number of cases,9 the leading one being Triodos Bank NB v Dobbs.10 In that case, Longmore J. held that such clauses would only extend to cover variations that fell “within the general purview of the original guarantee” and where the amendments could be seen as a true variation of an existing obligation rather than the entering into of a completely new and different obligation, even if the clause purports to be no more than a variation. With regard to the “purview” principle, Atkin J. in Trade Indemnity Co. Ltd. v Workington Harbour and Dock Board11 said that, when looking at the scope of variations, one should consider whether the changes have the effect of “substituting a cathedral for a dock, or a construction of a dock elsewhere.” Whether or not the “purview” principle is engaged is highly fact specific and, where it is engaged, it will be necessary to construe the entire shipbuilding contract and not merely the words of the anti-avoidance provision.12
What to do in practice
From a buyer’s point of view, obtaining the refund guarantor’s consent by way of a carefully drafted supplemental letter is undoubtedly the best course. Trying to ensure that the refund guarantee is properly classified as a demand guarantee will also help exclude the application of the law on material change. Further and somewhat paradoxically, although the absence of an anti-avoidance clause constitutes the fourth factor in the Paget Presumption that a guarantee is a demand guarantee, a buyer’s best fall-back position is to include an anti-avoidance clause which casts its terms sufficiently wide to ensure that changes to the underlying contract would fall within the purview of the original refund guarantee.
 3 Q.B.D. 495.
See: The Hon. Mrs Justice G. Andrews and R. Millett QC, Law of Guarantees (, 7th ed., Sweet & Maxwell, 2015), at §9-024 and the cases cited therein.
Wuhan Guoyu Logistics v Emporiki Bank  ECWA Civ 1629, per Longmore L.J. at .
Gold Coast Ltd v Caja de Ahorros  1 Lloyd’s Rep 617, per Tuckey L.J. at .
M. Hapgood and Sir J. Paget, Paget’s Law of Banking (11th ed., Butterworths, 1996).
per Tuckey L.J. at .
 2 Lloyd’s Rep 261.
See: Chitty on Contracts (31st ed., Sweet & Maxwell, 2012), at section 45-112.
See: Polaris Steamship Co SA v A Tarricone Inc (The Nefeli)  1 Lloyd's Rep. 339; and Melvin International S.A. v Poseidon Schiffahrt G.m.b.H. (the “Kalma”)  2 Lloyd's Rep. 374.
 2 Lloyd’s Rep 588 (CA).
 AC 1.
CIMC Raffles Offshore (Singapore) Ltd v Schahin Holding SA  EWCA Civ 644.