Security Interests—specifying “events of default”

Andrew Bridgeford

This note examines the 2003 Jersey case of EM TV & Merchandising AG v Bayerische Landesbank, in which it was held that the requirement under the Security Interests (Jersey) Law 1983 that a security agreement “specify” events of the default can be satisfied by those events being specified in the security agreement by a clear cross-reference to events listed in another document. The question remains relevant in Guernsey under the similar provisions of the Security Interests (Guernsey) Law 1993 and in relation to pre-2014 security interests in Jersey. It is argued that the decision in EM TV was correct as a matter of statutory interpretation of the word “specify”. However it is also argued that a preferable analysis, leading to the same result, is that the “security agreement” which is required by the Laws need not be contained within a single document. In this regard reference is made to English cases under the formerly applicable provisions of s 40 of the Law of Property Act 1925; and also under s 4 of the Statute of Frauds 1677, including the 2012 Court of Appeal decision in Golden Ocean Group Ltd v Salgaocar Mining Inds PVT Ltd.

1  The 2012 judgment of the English Court of Appeal in Golden Ocean Group Ltd v Salgaocar Mining Inds PVT Ltd^[1] throws an interesting alternative light on the issue that was considered in 2003 by the Royal Court of Jersey in EM TV & Merchandising AG v Bayerische Landesbank and others^[2].

2  Article 3(1)(f) of the Security Interests (Jersey) Law 1983 provides: “(1) For the purposes of this Law a security agreement shall— . . . (f) specify the events which are to constitute events of default”. The issue in EM TV was whether a security agreement may “specify” the events of default, as required by this provision, by an express cross-reference to events listed in another document, rather than having to set them out in extenso within the document referred to as the security agreement. The then Bailiff held that cross-referencing of this kind, if sufficiently clear, is an unexceptionable way by which the events of default can be “specified” by a security agreement in accordance with the Law. The very common and useful way of proceeding was thus vindicated. I argue that the same result can alternatively be reached by a different route, bearing in mind the decision in Golden Ocean.

3  The Security Interests (Jersey) Law 2012 repealed the Security Interests (Jersey) Law 1983. Jersey now has a very different legislative framework for security interests, with wholly new provisions governing inter alia creation, registration, priority and enforcement. Inevitably, the attention of Jersey practitioners is heavily focused on the new Law. With that in mind, it may seem academic to re-examine one of the issues that may have arisen under the 1983 Law.

4  That is not so for two reasons. First, the new Law provides that “continuing” security interests”^[3] created under 1983 Law will, after the coming into force of the new Law, remain governed by it and not by the new Law.^[4] These security interests have priority over new-Law security interests in the same collateral, and questions not only of their validity but also priority (between themselves) and enforcement continue to be governed by SIJL-1983. Secondly, the Security Interests (Guernsey) Law 1993 is very similar, albeit not in all respects identical, to Jersey’s 1983 Law. Guernsey is not, at present, proposing a wholesale replacement of its statute.

5  The following abbreviations are used in this note—

(a)   “1985 Amendment” means the Security Interests (Amendment) (Jersey) Law 1985;

(b)   “LPA-1925” means the Law of Property Act 1925 (of England and Wales);

(c)   “SIGL-1993” means the Security Interests (Guernsey) Law 1993;

(d)   “SIJL-1983” means the Security Interests (Jersey) Law 1983;

(e)   “SIJL-2012” means the Security Interests (Jersey) Law 2012; and

(f)    “SI Laws” means SIJL-1983 and SIGL 1993 (but not SIJL-2012).

6  References to numbered articles alone are to the provisions of SIJL-1983. The equivalent provision (section) in SIGL-1993, where materially identical, is placed in square brackets immediately afterwards. Minor differences in wording are generally indicated by putting the Guernsey word or phrase in square brackets immediately after the Jersey expression which it replaced.

EM TV & Merchandising AG v Bayerische Landesbank

7  In contrast to SIJL-2012, both SIJL-1983 and SIGL-1993 are highly prescriptive about the form and content of a valid security agreement. EM TV concerned one of those requirements: the requirement to “specify” the events of default. A parallel issue does not arise under SIJL-2012, since there is no statutory requirement to set out the events of default.

8  It is common practice for the events of default in Jersey security arrangements to be specified by cross-reference to events set out in one of the other transaction documents, rather than being set out in full within the Jersey security document. This saves unnecessary duplication and ensures consistency.

9  Widespread as the practice is, the plaintiff in EM TV argued that this way of proceeding did not satisfy the requirement in art 3(1)(f) of SIJL-1983 (which is reflected by art2(1)(f) of SIGL-1993) that a security agreement “specify the events which are to constitute events of default”.^[5] In order for the security agreement to “specify” the events of default, those events would, it was argued, need to be set out in the “security agreement” itself and could not be referred to by cross-reference to events listed in another document. The result, if correct, would mean that the security agreement under consideration in the case would have been invalid.

10  The plaintiff argued that support for this analysis could be drawn from the fact that the 1985 Amendment to SIJL-1983 had removed the requirement to “specify” various other matters required by the Law, replacing it with more convenient requirements.^[6] The requirement to specify prior encumbrances was discarded. The original requirement regarding what needed to be stated about the “collateral”—which was that the agreement had to “specify . . . particulars of the collateral sufficient to enable it to be identified”—was replaced by a stipulation that the agreement must “contain provisions regarding the collateral sufficient to enable it to be identified”. In the same vein, the requirement to “specify . . . the nature, duration and amount of the obligation payment or performance of which is secured under the security agreement” was replaced by “contain provisions regarding the obligation payment or performance of which is to be secured sufficient to enable it to be identified”. But the requirement to “specify” the events of default was retained. Counsel for the plaintiff in EM TV argued that to give the word “specify” a meaning which allowed the incorporation of the list of events of default by cross-reference to another document would be to give that requirement the same meaning as “contain provisions sufficient to enable it to be identified”. Given the nature of the 1985 Amendment, this, he argued, cannot have been intended.

11  Sir Philip Bailhache, B rejected these contentions. He held that events of default are adequately “specified” for the purpose of SIJL-1983 if they are definitely or explicitly identified in the security agreement, taken together with any documents incorporated by reference, provided that there is no ambiguity. No appeal was made, but counsel for the plaintiff reiterated his intriguing case in a subsequent article in the Jersey Law Review (as it then was), arguing that there was sufficient doubt about the correctness of the Royal Court’s decision such that practitioners should take a cautious approach to the matter.^[7] The Jersey Royal Court might depart from EM TV if convinced that the case was wrongly decided. The Guernsey courts would be free not to follow EM TV if called upon to decide the parallel point under SIGL-1993. Should practitioners in either Island be concerned about the matter?

“[S]pecify the events which are to constitute events of default”

12  Bailhache, B observed in EM TV that—

“To construe the statute [as preventing the incorporation of events of default by cross reference to another document] would not only run counter to the presumed intention of the legislature, but would also produce inconvenience . . .”

In his subsequent article Advocate Robertson criticised this conclusion on the basis that there was no reason to suppose that the legislative intention went against his case; on the contrary, he argued, the 1985 Amendment suggested that the legislature intended that the requirements regarding events of default to be “stricter” than were required in the case of the other prescribed matters and so as not to include the possibility of incorporation by cross-reference.

13  Was this criticism justified? In construing an unclear statutory provision, the judge is entitled to weigh various presumptions in the overall balance. It is of interest to consider the presumptions in somewhat more detail. One is that the legislature is presumed to intend the court to interpret legislation in a way that supresses the mischief against which it is directed (Bennion on Statutory Interpretation, s 289)^[8] and which does not result in a disproportionate counter mischief (Bennion, s 318). It is also presumed that the legislature intends that a court should assess the likely consequences of rival constructions, both for the parties themselves and for others in the future; and if on balance the consequences are likely to be more adverse than beneficent, this is a factor weighing against a particular construction (Bennion, s 286).^[9] The present author would argue that the then Bailiff was justified in the conclusion he reached as to the presumed intentions of the legislature in both respects.

14  As to the question of mischief, the purpose, inter partes, of a formal requirement of writing and signature can be said (in the authoritative words of Lord Hoffmann in the 2003 case of Actionstrength Ltd v Intl Glass Engineering In.GI.En. SpA^[10]) to be the protection of people “from being held liable on the basis of oral utterances which were ill-considered, ambiguous or completely fictitious”. SIJL-1983 and SIGl-1993 regulate not only the form of a security agreement but also particular aspects of its content, including the required specification of events of default. This was presumably because a further reason was felt to exist in the context of security for there to be a clear written record of the prescribed matters: the position of third parties in an insolvency of the grantor will be affected by the creation and enforcement of a security interest in the grantor’s property.^[11] There is nothing in any of these mischiefs that requires one to read the SI Laws as forbidding the specification of the relevant provisions by means of a reference to terms set out in another document, provided that that the cross-reference is unambiguous. In particular, cross-referencing, if clear, causes no prejudice to third parties. Nor does there seem to be any conceivable reason why a cross-reference should be allowed in order to record the relevant details of the “collateral” and the obligation secured but not the “events of default”. As was concluded by Bailhache, B, the essence of the requirement that events of default are “specified” in written form is to ensure that there is no ambiguity as to what counts as an event of default and this can be achieved by cross-reference to another document, provided that this is clearly done.

15  As to any counter-mischief and any adverse consequences that would arise on the plaintiff’s interpretation, there would, as the then Bailiff also noted, be a widespread inconvenience in requiring events of default to be set out in each case in extenso in a single document called the security agreement. Such a result would run against accepted commercial practice. Furthermore, the fact that the plaintiff’s argument, if accepted, would have invalidated innumerable existing security arrangements can hardly be ignored. Moreover, it is submitted that the plaintiff’s argument in EM TV would, if accepted, have caused an injustice to the particular defendant. The parties had freely entered into a security agreement which incorporated the events of default by cross-reference in what was a commonly accepted way. The effect of Bennion, s 286 must be that the court should tend to lean against any interpretation of a statutory provision that would invalidate an agreement entered into by parties who were acting in accordance with a reasonable and very commonly accepted interpretation of the formal requirements of the Law for an agreed purpose.

The 1985 amendment

16  The present author would also argue that the 1985 Amendment, upon which the plaintiff placed considerable reliance, cannot on balance be used to support its case. The Report of the Finance and Economic Committee^[12] presented to the States Assembly disclosed a general intention behind the 1985 changes. As noted by Advocate Robertson, the opening paragraph tells us that the 1985 Amendment was intended—

“to make absolutely certain that security agreements, which relate to a variety of complex and high value lending commitments, could not be set aside for obscure technical reasons.”

In relation to the requirements of art 3(1), the Report tells us more specifically that the intention was to utilise—

“terminology which is in more general terms so that it could not be argued that inadvertent non-compliance of a security agreement with certain specific terms in the Law had led to it not being a security agreement for the purposes of the Law and therefore the lender losing his security.”

17  It can therefore plausibly be said that the amendments to art 3(1) were intended to loosen some of the formal requirements of the Law in order to make it more practical and easier to comply with. That is a policy which, in itself, makes it rather less likely that the legislature intended the continuing requirement to “specify” the “events of default” to be interpreted in quite the narrow way sought by the plaintiff—an interpretation running counter to a practice which was common, convenient and hitherto regarded as unexceptionable, and the adoption of the plaintiff’s case by the courts would risk causing the very “inadvertent non-compliance”, fatal to a security agreement, that the 1985 Amendment was explicitly seeking to avoid. It is more likely that the legislature regarded the requirement to “specify” events of default as already sufficiently flexible, already consonant with the general policy of ensuring that security agreements were not at risk of invalidity for what the Report calls “obscure technical reasons”.

18  Moreover, it is plausible to regard the change as to how the “collateral” had to be recorded as driven by a particular concern which arose in cases where it was intended to create security in securities held in a portfolio. The original wording required a security agreement to “specify . . . particulars of the collateral sufficient to enable it to be identified”. One concern was that the strict requirement to specify “particulars” had the impractical effect of requiring the security agreement to be amended, or even entered into afresh, every time that the composition of an investment portfolio changed. The draftsman certainly had this sort of problem in mind; the Report specifically highlights a general concern that SIJL-1983 might not allow security to continue over collateral that had been altered. But the Report refers to this issue only in relation to a change proposed to the definition of “collateral” in art 1, by virtue of which the words “and includes initial, substituted and additional property which is so subject from time to time” were added. The relevant part of the Report comments—

“The alteration to the definition of “collateral” in Article 1 is intended to ensure that there can in future be no doubt that the property which is subject to a security agreement may be altered from time to time and therefore that, for instance, a security agreement can operate over a changing portfolio of shares.”

19  The definition of “collateral” was thus altered. It could be said, however, that the principal impediment to the continuation of such security was not the definition of “collateral” but rather the requirement in art 3(1) that “particulars” of the collateral had to be specified in the security agreement. It is difficult to see how a change to the definition of “collateral”, by itself, could have addressed that issue; indeed the change actually effected, without more, might even have made it worse. The revised art 3(1), on the other hand, avoids both the word “specify” and the rather stringent expression “particulars of the collateral” and opts instead for a requirement that a security agreement “contain provisions regarding the collateral sufficient to enable it to be identified”. The result is that collateral comprised by securities in an investment portfolio can be identified at any given time by reference to the particular securities account or investment agreement in or under which they are held. Had the wording of art 3(1) not been altered, but only the definition of “collateral”, it would have remained more than arguable that the requirement for a security agreement to “specify . . . particulars” of the “collateral” meant that the agreement had to be amended every time that any of the investments in a portfolio were sold and replaced by others.

20  Similarly, the change to the requirements of art 3(1) regarding the “secured obligation” could also be seen as motivated by the need to make SIJL-1983 better suited to common situations where the details, previously constrained to be set in stone, were in fact going to be subject to variation during the intended life of the security. The focus referred above was on variable collateral; here it is on secured obligations which might, in some respect, also be variable or unpredictable. A loan facility might be repayable on demand or during its life could be subject to amendment in amount or duration. The original wording was extremely prescriptive. It required a security agreement to “specify . . . the nature, duration and amount of the obligation payment or performance of which is secured under the security agreement”. This was replaced by a looser, more practical provision which requires a security agreement to “contain provisions regarding the obligation payment or performance of which is to be secured sufficient to enable it to be identified”.

21  It is true that, in all this, the word “specify” was left above the high water mark of the 1985 Amendment in so far as the “events of default” were concerned. It is also true that the revised requirements in respect of the “collateral” and the “secured obligation” are in terms which rather clearly allow incorporation of the relevant details to be made by reference to other documents or to other facts. But that is irrelevant. It does not show that the requirement to “specify” the events of default did not already allow incorporation by cross-reference.

22  The plaintiffs’ argument assumes (a) that the relevant amendments were directed at the issue of incorporation by cross-reference and (b) that the Jersey States Assembly thought it appropriate to draw a distinction, allowing cross-reference in order to identify the collateral and the secured obligations but not the events of default. In the first place, there is no hint, either in the Report or in the 1985 Amendment itself, that the draftsman was concerned at all with the issue of incorporation by cross-reference; and in the second, even if this was an issue, there is surely no plausible reason of policy why such a distinction should have been drawn.

23  Only if the changes to the other requirements were specifically intended to introduce the possibility of incorporation by cross-reference would they re-enforce the plaintiff’s case about the residual use of the word “specify”. For the reasons mentioned above, the changes to art 3(1) are otherwise plausibly explicable. The word “specify” should therefore be considered aside from the 1985 Amendment and the plaintiff’s argument is not re-enforced by it.

24  The balance of these factors favours, it is submitted, an interpretation of the word “specify”, as it was used in this particular legislative context, which allows events of default to be “specified” by cross-reference to another agreement or document, as was held by the Royal Court. But the case need not turn on the meaning of “specify”. There is an alternative and, I would argue, preferable basis for reaching the same result.

“[A] security agreement shall . . .”

25  My suggestion is that it is unnecessary to make the word “specify” do all the work. Rather, this sort of case can be seen as turning on what is meant in the SI Laws by the expression “security agreement”. It is in this respect that the more recent English Court of Appeal decision in Golden Ocean Group Ltd v Salgaocar Mining Inds PVT Ltd^[13] throws an interesting light on the issue.

26  The unspoken assumption in the plaintiff’s argument in EM TV is that the words “security agreement” as used in art 3(1)(f) [2(1)(f)] necessarily refer to what would conventionally be described as a single written document. But there is nothing in art 3(1) [2(1)], or indeed anything else in the insular SI Laws, that requires this to be so. Nor do the mischiefs referred to above, against which these provisions were presumably directed, require the agreement to consist numerically in only one written document. Indeed, the words “security agreement”^[14] when used in art 3(1)(a) [2(1)(a)] cannot refer to a written agreement at all: in that particular provision, “security agreement” must be used in a conceptual “meeting of the minds” sense, since otherwise the requirement there that “a security agreement shall . . . (a) be in writing” would be tautological. But the real point is that the expression “security agreement”, when used for the purposes of the other provisions of art 3(1) [2(1)], does not necessarily refer to a single written document. One and the same “security agreement” could be spread across what are different but linked documents. This conclusion is not undermined by considerations of customary law. The expression “agreement” is not defined, either in SIJL-1983 or SIGL-1993, and should therefore be interpreted against the background of customary law. There is no requirement of customary law that the consensus of subjective wills between parties which gives rise to a contract should, if recorded in writing, necessarily be recorded in a single document. It is therefore hard to see why the position should be any different with regard to an “agreement” for the purposes of the SI Laws. Furthermore, and perhaps decisively, art 3(2) [2(2)] can legitimately be brought into play as well. It expressly says that, subject to art 3(1) [2(1)]—

“a security agreement may be in such form[,] and [may] contain or refer to such matters[,] as shall be agreed between the parties to such agreement [as may be agreed between the parties].”^[15]

27  The scenario under consideration is one where there is a primary document which sets out the main terms of the security agreement and a secondary document, part of which is expressly and unambiguously referred to by the primary document in order to specify the events of default. If it is accepted that the SI Laws do not require a “security agreement” to consist in a single document, then why should the combination of primary and secondary document not properly amount to a “security agreement” between the parties, which is “in writing” and which “specifies” the events of default? Provided that the other formal requirements of art 3 [2] are met, there seems no reason such a “security agreement” should not be entirely valid. The only issue might be whether the agreement is “signed by the debtor”.^[16] In practice, both documents are likely to have been executed by the debtor—it usually being a party to both—but even if only the primary document is executed by the debtor, the “security agreement” as set out across the two documents, is still one which is “in writing” and which has been “signed by the debtor”.^[17] Furthermore, the “cross-reference” need logically only be in one direction, from the primary, signed document to the other; it is not necessary for the second to refer back the first.

28  Support for this way of looking at the problem comes from a line of cases in England under the old s 40 of LPA-1925 and s 4 of the Statute of Frauds 1677.

29  Until its requirements were replaced in 1989, s 40(1) of LPA-1925 provided that an action could not be brought on a contract for the sale or other disposition of land—

“unless the agreement upon which such action is brought, or some memorandum or note thereof, is in writing, and signed by the party to be charged or by some other person thereunto by him lawfully authorised.”

This wording derived from s 4 of the Statute of Frauds 1677. Section 4 of the Statute of Frauds itself continues to apply in England and Wales in relation to guarantees and actions enforcing them: it provides that—

“the Agreement upon which such Action shall be brought or some Memorandum or Note thereof shall be in Writing and signed by the party to be charged therewith or some other person thereunto by him lawfully authorised.”

30  In one line of cases, the question arose whether the required “memorandum or note” of an oral agreement for the sale of land might consist in all the salient details required by the common law (parties, property and consideration^[18]) being recorded across two or more documents that could be joined together in order to comprise the “memorandum or note” required by statute. By the early nineteenth century, the courts had accepted joinder of documents for this purpose (under the similar wording of the original s 4 of the Statue of Frauds) so long as the documents were expressly or by implication linked. It was later held that extrinsic evidence may be led, if necessary, in order to identify the second document and to establish the link between them: see Timmins v Moreland Street Property Co Ltd.^[19] As for signing, it was only necessary that one of these documents be signed: Timmins; Elias v George Sahely & Co (Barbados) Ltd.^[20]

31  Timmins and Elias both concerned whether the statutory requirement for “a memorandum or note” in writing evidencing an agreement (not itself amounting to the agreement) might consist in more than one document. A more directly analogous question is whether, for the purposes of LPA-1925 or the Statute of Frauds 1677, the written agreement itself could consist in more than one document.

32  That was a question which arose under the Statute of Frauds 1677 in the more recent decision of the English Court of Appeal in Golden Ocean Group Ltd v Salgaocar Mining Inds PVT Ltd. As noted above, the English law^[21] position is that s 4 of the Statute of Frauds still requires, in relation to guarantees and actions enforcing them, that—

“the Agreement upon which such Action shall be brought or some Memorandum or Note thereof shall be in Writing and signed by the party to be charged therewith or some other person thereunto by him lawfully authorised.”

33  The Golden Ocean case pitted these well-known words, scratched with quills by bewigged Restoration lawyers, against the commercial praxis of the modern computer age. The question was whether a chain of email correspondence could constitute the required “Agreement”. In the course of considering this question, the English Court of Appeal rejected any contention that the written “Agreement” required by the Statute of Frauds must necessarily consist of an agreement contained in a single document. Lord Justice Tomlinson (with whom the other judges agreed) observed at para 21 that the Statute of Frauds—

“contains no express indication that the agreement in writing required to satisfy its terms must be in one or even a limited number of documents. It is no doubt true that in 1677 a signed written agreement would often and perhaps always be contained in a single document, but Mr Kendrick very sensibly did not suggest that that provides a pointer to how the Statute should today be construed. However his argument did, as it seems to me, in all its elegant iterations, always come back to the point that an instrument of that sort would accord with most people’s understanding of what is meant by a contract in writing. That of course may be so, but we are immediately concerned with the understanding of professionals in the shipping market. Moreover the purpose of the requirement that the agreement must be both in writing and signed by the guarantor is not so much to ensure that the documentation is economical but rather to ensure that a person is not held liable as guarantor on the basis of an oral utterance which is ill-considered, ambiguous or even completely fictitious—see per Lord Hoffmann in Actionstrength at page 549 E. A combination of writing and an acknowledgement by signature of the solemnity of the undertaking has been chosen to eliminate that mischief. I see nothing in either the mischief sought to be eliminated or the means adopted to achieve that end which requires a limitation upon the number of documents in which the writing is to be found.”^[22]

34  It is submitted that counsel for the plaintiffs in EM TV was in the same position as counsel for the appellants in Golden Ocean: for all the elegance of his argument, it rests on the assumption that the written “security agreement” required by SIJL-1983 must necessarily consist within the four corners of one conventionally bound or stapled document. A single security agreement in the legal sense may consist across two linked documents, one of which may set out the events of default and, provided that the cross-reference is clearly made and there is no ambiguity, there should be no doubt that those events are “specified” in a conventional sense by a “security agreement” “in writing” that the parties have entered into, and thus without the need to make the word “specify” do any extra work.

35  Indeed, if it is accepted that the concept of a “security agreement” for the purposes of the SIJL-1983 and SIGL-1993 is not the same as the physical document, or documents, in which the agreement happens to be recorded, then the preferable analysis of a sufficiently clear cross-reference to events of default set out in another physical document is that such events are inevitably “specified” as part of one and the same “security agreement”. The question whether the word “specify” allows cross-referencing to another agreement cannot arise; but exactly as in the case of the analysis adopted by the court in EM TV, the link must be clear, for it must be established that the terms being referred to are indeed part of the parties’ “security agreement”. In practice this is unlikely to be a problem.

Andrew Bridgeford is a lawyer and writer and a consultant at Mourant Ozannes.