RAIL FOR LONDON LIMITED v THE MAYOR & BURGESSES OF THE LONDON BOROUGH OF HACKNEY

The Claimant, Rail for London Limited (Rail for London), seeks to strike out the section of the defence of the Defendant, London Borough of Hackney (Hackney), relating to estoppel by convention. Rail for London contends that the relevant paragraphs of the Defence, namely paragraphs 52 to 57, disclose no reasonable grounds for defending the claim (the Application). Rail for London contends that Hackney is impermissibly seeking to use estoppel by convention as a sword rather than a shield, and thereby create an enforceable right to receive rent where none previously existed, which would have the effect of varying the relevant lease between them without Hackney having provided any consideration.

Given that the application is put in this way rather than seeking to knock out the use of estoppel by convention on any more fact-sensitive basis, I can set out the facts shortly.

The relevant aspect of the proceedings for present purposes concerns the meaning and operation of a 3 May 1996 underlease for 99 years and a day from 25 March 1994 (Lease C) of railway arches and buildings at Kingsland Viaduct (the Arches) made between Hackney as landlord and Rail for London’s predecessor, London Underground Limited (London Underground). The parties are in dispute about the effect of a surrender of the sub-lease to Lease C on the rent payable under Lease C.

The rent payable under Lease C, which is defined as “the Rent”, includes the “Basic Rent” (paragraph 1(a) of Schedule 4). The Basic Rent is in turn defined insofar as relevant for present purposes as “for each Relevant Year the Basic Rent (as defined in the Underlease) that is received by the Tenant pursuant to the provisions of the Underlease…” (clause 1.1).

The Underlease referred to (Lease D) was granted on the same day as Lease C by London Underground to London Industrial (Kingsland Viaduct) Limited (London Industrial), for 99 years from 25 March 1994. Under the definitions in clause 1.1 of Lease D, Basic Rent is defined in Lease D as “for each Relevant Year the Percentage of the Net Income”, Net Income is in turn defined as “the Gross Income minus the Expenses in any Relevant Year”, and Gross Income “means in relation to each Relevant Year the aggregate of the following actually received by the Tenant”, the Tenant being London Industrial and its successors in title.

Lease C and D were granted pursuant to a principal agreement, also dated 3 May 1996, between Hackney, London Underground, London Industrial, and London Industrial plc. Pursuant to that agreement, the freehold of Kingsland Viaduct and the Arches was transferred by Hackney to London Underground for £1, and a sequence of leases were entered into whereby the Arches were let by London Underground to Hackney for a nil or nominal rent under a Lease B for 99 years plus two days from 25 March 1994, Hackney sub-let to London Underground under Lease C, and London Underground entered into the sub-underlease, Lease D, with London Industrial. These arrangements were entered against the backdrop of the possible extension of the East London Line from Shoreditch to Dalston.

On 21 November 2003, Lease D was surrendered for £7,788,500 (the Surrender). Rail for London became tenant under Lease C in 2009. Until September 2019, London Underground and then Rail for London continued to pay rent under Lease C: the tenant under Lease C made quarterly payment towards its rental liability, before providing a statement of income and expenses in respect of the Arches and then making an annual balancing payment.

In December 2019, Rail for London asserted that as a result of the Surrender, no Basic Rent had been due after the Surrender, and sought repayment of around £6m paid during the intervening period. Hackney disagreed. Ultimately Rail for London issued proceedings on 9 March 2021 (the Proceedings) seeking various declarations, including that Basic Rent is not payable under Lease C.

The trial is listed for 4 days in October 2022.

In the Particulars of Claim, Rail for London asserted that from 22 November 2003, the day after the Surrender, the Basic Rent under Lease C was zero (paragraph 25(d)) and therefore from that date the Basic Rent ceased to be payable under Lease C (paragraph 26). Rail for London therefore sought in the prayer a declaration that “the Basic Rent is not payable under Lease C”. There was no claim pleaded for restitution of overpaid rent.

In its Defence, Hackney pleaded that (i) on the true construction of Lease C, Basic Rent continued to be payable after the Surrender, (ii) further or alternatively a term should be implied to this effect (paragraphs 37-51), and (iii) further or alternatively Rail for London was “estopped by convention from asserting that Basic Rent is not payable under Lease as a result of the Surrender” (paragraph 52). In aid of contentions (i) and (ii), Hackney pleaded that the admissible background included a number of matters, including that the commercial purpose of the Principal Agreement was (among other things) that Hackney would obtain by the mechanism of a long lease-back in the form of Lease C, the net rental income from those arches for a 99 year term (paragraph 35(d)(vii)). Rail for London denies that this was part of the commercial purpose.

The components of the estoppel by convention pleaded by Hackney were that:

Hackney went on to counterclaim that Rail For London had breached its obligations under Lease C by failing to pay rent since September 2019, and sought declarations (i) in the alternative that (a) its interpretation of Lease C was the correct one, (b) a term was to be implied into Lease C to the same, and (c) that Rail for London was estopped by convention from asserting that Basic Rent is not payable under Lease C as a result of the Surrender, and (ii) that Rail for London provide Hackney with the material necessary to calculate the unpaid Basic Rent.

In its Reply to Defence and Counterclaim, Rail for London pleaded in relation to estoppel by convention that (i) paragraphs 52 to 57 of the Defence disclosed no reasonable grounds for defending the claim because Hackney was seeking to use estoppel by convention as a sword, not a shield, (ii) it was denied that the parties shared a clearly communicated common assumption in the terms alleged, (iii) it was denied that Hackney relied on the alleged common assumptions to its substantial detriment or at all, and (iv) it was denied that it would be unconscionable to permit Rail for London to resile from any convention established by the parties (paragraphs 32 to 35). The contention in point (i) forms the basis of the strike-out application before me.

The Application relies on limb (a) of CPR r.3.4(2), namely that that the statement of case discloses no reasonable grounds for bringing or defending the claim.

Mr Bhose QC, appearing with Mr Datta for Hackney, submitted that an application to strike out in a context such as this should not be granted unless the Court is certain that the claim is bound to fail. He relied on [22] of Hughes v Colin Richards & Co [2004] EWCA Civ 266, a case in which the Court of Appeal dismissed an appeal against the rejection of a strike-out application. Peter Gibson LJ was dealing in that paragraph with the correct approach when the pleadings show significant disputes of fact between the parties going to the existence and scope of the alleged duty of care. Peter Gibson LJ held that the correct approach was that the Court must be certain that the claim is bound to fail in order to strike it out. In doing so, he relied on the House of Lords decision in Barrett v Enfield London Borough Council [2001] 2 AC 550 at 557 per Lord Browne-Wilkinson, and set out the passage that follows in Lord Browne-Wilkinson’s judgment, namely that “in an area of law which was uncertain and developing (such as the circumstances in which a person can be held liable in negligence for the exercise of a statutory duty or power) it is not normally appropriate to strike out. In my judgment it is of great importance that such development should be on the basis of actual facts found at trial not on hypothetical facts assumed (possibly wrongly) to be true for the purposes of the strike out”. Mr Bhose and Mr Datta submitted that in light of Hughes, the Court need only conclude that Hackney’s case was arguable.

Mr Hutchings QC, appearing for Rail for London, contended that the situation in Hughes should be distinguished from the present. He argued that here the question of whether estoppel was being used as a sword rather a shield did not depend on contested facts. Rather, the effect of striking out here would be to remove the need for the final hearing to be a witness action. He referred me in this regard to paragraph 85 of Nugee J’s judgment in Barness & Ors v Ingenious Media & Ors [2020] PNLR 10, where the Judge stated that he had been referred “to some of the authorities on the caution that should be exercised before striking out…a claim, particularly in an areas of developing jurisprudence or a novel factual situation” but that “[i]t is also well established however that the Court should use the powers it has to prevent a claim going forward to trial if it is confident that there is in truth nothing in it; that is particularly so if the effect is to remove a defendant, or a discrete area of factual enquiry, from the proceedings entirely” ([85]).

As I explain in more detail below, in my judgment the relevant legal points on estoppel by convention are open to serious argument, so there are reasonable grounds to defend the claim. In my judgment, this is not close to a situation where the Court can be confident that there is nothing in Hackney’s argument, or that it is bound to fail. I also consider that it would be appropriate in an area of law that is uncertain and contains a number of decisions that are not straightforward to reconcile (such as the circumstances in which estoppel is impermissibly used as a sword), for the Court to make its decision on the basis of the Court’s precise factual findings as to the alleged convention and other elements of the alleged estoppel by convention. Attractively though Mr Hutchings presented his oral submissions, for these two reasons I refuse the Application.

Mr Hutchings’ argument is that estoppel by convention cannot be used in the present factual circumstances to enforce what is in effect a gratuitous variation of a contract that increases the allegedly estopped party’s contractual obligations without any corresponding increase in the other party’s obligations or other consideration. Mr Bhose QC and Mr Datta strongly resist the characterisation of Hackney’s use of estoppel as an attempt to bring about a variation of Lease C. They characterise the relevant convention as binding the parties to a shared interpretation of the pre-existing agreement that was entered into in 1996. In more detail, they contend that (i) estoppel can establish a necessary element of a cause of action, (ii) which in turn allows the effect of a contract to be extended by estoppel as long as the rules setting the requirements for a valid contract are not thereby unacceptably subverted, and (iii) Hackney’s case here does not rely on estoppel as a cause of action itself, because its putative cause of action is breach of Lease C, which is capable of arising independently of the estoppel by virtue of the parties’ pre-existing relationship through their entry into Lease C, and (iv) nor does it subvert the rules relating to consideration, because the necessary consideration was supplied at the point of concluding Lease C, which is the contract that Hackney seeks to hold Rail for London to a shared interpretation of.

Reflecting the controversy of the precise limits of estoppel by convention, I was referred to a considerable number of cases, together with commentaries on the area dealing with a significant further number of authorities. In order to analyse the parties’ competing contentions, I shall focus in chronological order on the main cases that I was referred to, and pick up a number of their submissions in the course of doing so. Mr Hutchings centred his oral submissions around the Court of Appeal decision in Riverside Housing Association Ltd v White and White [2006] HLR 15.

Both parties placed reliance on Amalgamated Investment and Property Co v Texas Commerce International Bank [1982] QB 84 (CA) and adduced arguments as to the basis on which Amalgamated Investment had been distinguished in Riverside. Amalgamated Investment itself concerned the guarantee of a loan in circumstances where the precise route for making the loan changed from the original plan and this caused a question to arise as to whether guarantee constituted a binding and effective guarantee that covered the loan. The chronology is distilled in Brandon LJ’s judgment at 127D-130D and in more detail in Goff J’s first instance judgment at 90F-101C.

It was originally agreed that a loan should be made by the bank to one of the claimants’ Bahamian subsidiaries, to be secured by a mortgage over a Bahamian property and a guarantee from the claimants. That guarantee was signed. It was then agreed that instead of the loan being made directly by the bank, the bank would loan money to its own Bahamian subsidiary and that subsidiary would then make the loan to the claimants’ Bahamian subsidiary (the “Nassau loan”). However, the guarantee, which had already been signed, was not revised to deal expressly with this change of plan. Those events all took place in 1970. The bank had made another loan to the claimants earlier that year (the “UK loan”), secured on UK properties. Little happened for a few years until 1974 to 1976, when the bank and claimants conducted further transactions between themselves that proceeded on the assumption that the guarantee covered the Nassau loan, including the execution of charges intended to secure not only the UK loan but the guarantee as well.

The bank applied the proceeds of the UK properties to meet what it thought was the remaining debt under the guarantee. The claimants contended that this should not have been done because the guarantee did not catch the Nassau loan and launched proceedings for a declaration that the claimants were under no liability to the bank under the guarantee. Robert Goff J held at first instance that although the guarantee did not as a matter of construction cover the Nassau loan, the claimants were estopped from contending that it did not cover the loan. On appeal, the Court of Appeal held that the guarantee did cover the Nassau loan, and that in any case the parties were estopped by convention from denying that the guarantee covered the loan.

Lord Denning MR decided the part of the case relating to estoppel on the basis of the broad principle that “[w]hen the parties to a transaction proceed on the basis of an underlying assumption- either of fact or of law- whether due to misrepresentation or mistake makes no difference- on which they have conducted the dealings between them- neither of them will be allowed to go back on that assumption when it would be unfair or unjust to allow him to do so” (122C).

Eveleigh LJ applied the narrower principle set out in the then current edition of Spencer Bower and Turner, Estoppel by Representation, 3rd ed (1977) that “[w]hen the parties have acted in their transaction upon the agreed assumption that a given statement of facts is to be accepted between them as true, then as regards that transaction each will be estopped against the other from questioning the truth of the statement of facts so assumed” (126A-G). He regarded the relevant transaction for these purposes as the realisation of the securities of the bank. Therefore, he concluded that the bank could rely on estoppel to meet the claim that it should not have used the UK properties to discharge the guarantee. He did not consider that the bank could have succeeded on a claim on the guarantee itself if that had been the question before the Court, because he considered that the operation of the estoppel was limited to the transaction in which it arose (126D-F).

Brandon LJ also set out with approval the passage from Spencer Bower but with the addition of the previous sentence that “[t]his form of estoppel is founded, not on a representation of fact made by a representor and believed by a representee, but on an agreed statement of facts the truth of which has been assumed, by the convention of the parties, as the basis of the transaction into which they are about to enter” (130G-H). He considered that the relevant transactions entered into were the making of the new arrangements (in 1974-6) with regards to the overall security held by the bank in relation to both the UK and Nassau loans, which took place on the shared assumption that the guarantee covered the Nassau loan, and held that this was a classic example of estoppel by convention (131A-D). Importantly, he went on to deal with the objection raised by the claimants that the bank was seeking to use estoppel as a sword rather than a shield and that this was something which the law of estoppel did not permit, or put another way that a party could not found a cause of action on an estoppel. He considered that if the bank had brought an action against the claimants on the guarantee, the claimants would have contested the construction of the guarantee, and the bank would have been able through their reply to invoke estoppel to preclude the claimants “from questioning the interpretation of the guarantee which both parties had, for the purpose of the transactions between them, assumed them to be true” (131G). He continued:

The important point for the purpose of the Application is how the subsequent Court of Appeal in Riverside treated Amalgamated Investments. Foreshadowing his submission on Riverside, Mr Hutchings focused on the reasoning of Brandon LJ and Eveleigh LJ. He contended that the critical feature common to both of their judgments was that the common assumption pre-dated the relevant transaction, unlike in the present case, where the assumption post-dated the entry into Lease C. I make two observations on that:

Before dealing with Riverside itself, it is necessary to deal with three decisions in the intervening period after Amalgamated Investments.

The first is the House of Lords decision in Republic of India Steamship Co Ltd (No.2) [1998] AC 878. Mr Datta, who made the bulk of Hackney’s oral submissions after Mr Bhose QC fell ill for part of the hearing, relied on the statement of the law of Lord Steyn, delivering the only reasoned judgment, that “[i]t is settled that an estoppel by convention may arise where parties to a transaction act on an assumed state of facts or law, the assumption being either shared by them both nor made by one and acquiesced in by the other” (913D-E). As Mr Hutchings pointed out, Republic of India was not a case in which the question of the shield / sword distinction arose and was not a contract case. However, in my judgment it is notable that the test set out by Lord Steyn does not contain any requirement that the estoppel arise on the entering into the transaction.

I can deal briefly with the House of Lord decisions in Johnson v Gore Wood & Co [2002] 2 AC 1. Mr Hutchings relied on the judgment of Lord Goff in his skeleton and quoted from it. However, Lord Goff’s comments on estoppel by convention do not form part of the ratio of the case. Lords Cooke and Hutton agreed with the judgment of Lord Bingham on abuse of process, which contains his discussion of estoppel by convention, including quoting with approval (at 33C-G) Lord Denning MR’s judgment in Amalgamated Investments. Lord Millett dismissed the estoppel by convention argument on the facts (61A-C).

Mr Hutchings also quoted at length in his skeleton from the Court of Appeal decision in Baird Textile Holdings Ltd v Marks & Spencer plc [2002] 1 All ER (Comm) 737. However, in my judgment, the decision does not assist him. Mance LJ, whose judgment was cited with approval in Riverside, explained at [88] that the question of how far estoppel might assist in bringing about a cause of action had remained a matter of dispute over the years since Amalgamated Investments, but that it “may enlarge the effect of an agreement, by binding parties to an interpretation which would not otherwise be correct” ([88]), that it could be used to bind a person to accept that they were party to a contract ([89]) and estop a person who held out another as his agent from denying that such person was his agent in making the actual contract ([90]). What he rejected in [91], the passage cited from his judgment by Mr Hutchings, was that the “law ought to attach legal consequences to a bare assurance or conventional understanding (falling short of contract) between two parties, without any actual contract or third party being involved or affected” (underlining added). He considered that this would bypass the ordinary requirements of estoppel, such as “the objective intention to affect some actual or apparent pre-existing legal relationship” and “the requirements of contract (consideration, certainty and an intention to create legal relations)”. In other words, he did not consider that it could be used to create a legally binding relationship from scratch through a bare assurance or conventional understanding falling short of contract. That is not the present case, where Lease C forms a legally binding relationship between Rail for London and Hackney.

Given that it was not necessary on the facts, what Mance LJ does not do is explain the precise circumstances in which estoppel may enlarge the effect of an agreement, by binding parties to an interpretation which would not otherwise be correct. In particular it was not necessary for him to deal with whether an estoppel by convention generated after a contract is entered into must decrease the obligations of the party relying on the estoppel, rather than being used after the transaction is entered into in order to increase the obligations of one of the parties. Those are the issues that the present case concerns.

Turning to Tesco Stores Ltd v Costain Construction Ltd [2003] EWHC 1487 (TCC), which is quoted extensively in Mr Hutchings’ skeleton, Mr Hutchings points out that HHJ Richard Seymour QC held that “[t]he decision in Baird Textile Holdings Ltd. v Marks & Spencer Plc is, in my judgment, plain in holding that a cause of action cannot be founded upon an estoppel by convention….In seeking to rely upon an estoppel by convention that a contract which has not been made was made, or that a contract which contained set of terms A in fact incorporated set of terms B so as to be able, in the one case to sue on the contract and in the other to sue on set of terms B, it is obvious that Tesco is seeking to found a cause of action on estoppel” ([196]). Mr Hutchings suggests that this means that where a contract contains set of terms A, a party cannot use post-contract conduct to generate an estoppel by convention that allows it to sue on set of terms B.

In my judgment, Tesco Stores does not provide decisive support for the Application. The bulk of the reasoning in the relevant section of the judgment, [191]-[198], is making the point that if the parties are not in agreement that set of terms B should apply, no estoppel by convention will logically be able to arise that sets of terms B should apply. That is consistent with the way that the case is analysed at [8.59] of the latest edition of Spencer Bower, which is now titled Spencer Bower: Reliance Based Estoppel (5th ed, 2017). I consider that the point being made at the end of [196] does appear to offer some support for Mr Hutchings’ arguments. However, to the extent that the judgment is going further than that in the passages seeking to apply Baird Textiles, whether in the extract from [196] above or in [198], I have some difficulties for the reasons explained above in seeing that Baird Textiles goes further than dealing with an attempt to generate a legally binding relationship from an estoppel by convention. Further, Baird Textiles explains that it can be possible to use estoppel to “enlarge” a contract, as do later authorities. The question is in what circumstances and to what extent it can be used in this way.

The Court of Appeal decision in Riverside Housing Association Ltd v White [2006] HLR 15 formed the centrepiece of Mr Hutchings’ oral submissions. He contended that Riverside was decisively in favour of the Application and that the decision was right. I cannot determine as a first instance Judge the correctness of the decision. Similar points apply to the reliance by Mr Bhose and Mr Datta in their written submissions on criticism of Riverside and a number of other recent cases in the recent book of Michael Barnes QC, The Law of Estoppel (2020) at [2.32] and surrounding paragraphs. However, to the extent that these respective submissions bear on what Riverside decides and the grounds on which it distinguishes Amalgamated Investments, I have taken them into account.

In Riverside, the landlord housing association sought possession of a dwelling let under an assured tenancy on the basis that rent that was lawfully due had not been paid. The relevant tenancy allowed the landlord to increase the rent by giving four weeks’ notice provided that the rent payable was increased annually with effect from the first Monday in June each year. As part of resisting possession, a dispute arose as to whether this part of the rent increase provision was valid and therefore whether the rent increase notices were valid. The Court of Appeal held that the rent increase notices did not comply with the terms of the tenancy, that the terms of the tenancy had not been varied, and- relevantly for present purposes- that the landlord could not give effect to the rent increase notices through estoppel by convention.

The tenant successfully submitted that the landlord’s estoppel argument was seeking to use estoppel as a sword. Peter Gibson LJ, giving the judgment of the Court, started this section of the judgment by referring to the Court’s previous decision in Aristocrat Property Investments v Harounoff (1982) 2 HLR 102, which concerned similar facts in relation to a statutory tenancy ([62]). In Aristocrat, the Court held in rejecting the landlord’s case that the tenant could not waive the statutory requirements for a valid rent increase notice and that in any event the landlord would have to make his estoppel the foundation of the positive case for recovery of rent. Having set out Aristocrat,Peter Gibson LJthen described the facts of Baird and extracted [88] of the judgment of Mance LJ, to which I have referred above, and commented that “save in somewhat special circumstances, such as were those discussed in the Amalgamated case, an estoppel by convention cannot be used as a sword, as Mance L.J. accepted by agreeing with the other members of this Court” ([63]).

Peter Gibson LJ turned to Amalgamated Investments itself, focusing on the judgment of Brandon LJ. He explained that Brandon LJ had considered that estoppel could be used in aid of a claim by the bank on the guarantee, and set out the concluding paragraph of reasoning that I have quoted above and which the landlord relied on in Riverside ([64]). Having done so, he distinguished Amalgamated Investments in the following way:

Mr Hutchings contended that (i) this was clear authority that Hackney was wrong to say that estoppel could be used to generate a cause of action by supplying the missing piece in the jigsaw, because the landlord submitted in Riverside that the fact that there was rent due under the tenancy independently of the rent increase notice ([60] of the judgment) was decisive but this argument did not succeed, (ii) both here and in Riverside the landlord was seeking to allege that estoppel had the effect of varying the contract without consideration,and (iii) the present case was in fact a stronger one than Riverside for rejecting the use of estoppel by convention, because here no rent was due independent of the estoppel, and also because in Riverside the estoppel was only being used to seek to validate a past rent demand whereas here it was being used to seek to allow future rent demands.

I asked Mr Hutchings what he submitted was the precise basis on which the Court of Appeal in Riverside was distinguishing Amalgamated Investments. His answer was that it was simple: the Court of Appeal in Riverside was stating that Amalgamated Investments was different because the common assumption in Amalgamated Investments predated the entry into the relevant transaction. I have difficulty with that explanation, because it does not feature expressly in the discussion of Amalgamated Investments in the immediately preceding [64] or in the discussion of Amalgamated Investments in the extract from Baird Textiles set out in [63]. Moreover, Mr Hutchings accepted that estoppel by convention can apply, subject to certain conditions, where the common assumption does not predate the entry into the relevant transaction or indeed any transaction.

Mr Bhose submitted, right at the end of the hearing, that the ground of distinction was that the in Riverside the landlord was seeking to rely on a subsequent contractual notice to increase rent. As Peter Gibson LJ put it at [65], the position depended on the validity of the rent increase notices. In other words, the landlord in Riverside was seeking to give legal effect to a notice that had been issued after the contract had been entered into, which notice expressly sought to vary upwards the amount of rent due, rather than relying on an estoppel as to the meaning of the pre-existing agreement, here Lease C.

This chimes with Hackney’s broader submission that in the present case, it is not using estoppel by convention to try to give effect to a failed attempt at varying the contract. The convention relied on is, Hackney argues, not that the parties considered that the meaning of Lease C had been changed on the Surrender by some promise or assurance from Rail for London. Rather, following the Surrender, the parties proceeded on the assumption that Lease C in its original form required rent to continue to be paid notwithstanding the surrender.

It seems to me that on this argument, the present case is not- as Baird was- an attempt to use estoppel by convention to generate a legal obligation out of nothing. Putting Hackney’s argument a slightly different way, take Brandon LJ’s example in Amalgamated Investments set out in [26] above of the bank suing on (i) the original terms of the guarantee, and (ii) as a fallback to meet the response that the original terms of the guarantee did not impose the relevant legal obligation, (iii) relying on the parties’ post-contractual conduct as to what the guarantee had always meant in order to generate an estoppel. On Hackney’s argument, by analogy here (i) Hackney is contending as its primary case that on its true construction Lease C means and has always meant that rent should continue to be paid whether or not Lease D is surrendered, and (ii) to meet the response that Lease C does not mean this, (iii) Hackney argues that parties have placed such a construction on Lease C by their post-contractual dealings so as to generate an estoppel by convention as to the meaning of Lease C. This reasoning would place Riverside into the Baird category of cases on the basis that the parties understood in Riverside that the notices were intending to increase the rent that was due, the higher level of rent was not due under the original terms of the lease, and therefore the landlord’s argument was seeking to use estoppel to give effect to the landlord’s attempt through the notice to increase the rent from the level originally due.

An important limit of estoppel by convention is that it cannot be used in the contractual context to subvert the doctrine of consideration. Both parties before me accepted this.

Spencer Bower usefullyputs the matter in the following way: while estoppel by convention

Mr Hutchings’ argument is that when one looks at the effect of the estoppel, it has the effect of increasing Rail for London’s obligations after the contract has been entered into, so it is a variation by the backdoor and therefore would subvert the doctrine of consideration.

I agree that Hackney’s use of estoppel by convention to (among other things) seek to establish that Rail for London is liable for rent for the remainder of the term is at least towards the more controversial end of the spectrum. However, be that as it may, this is not on the face of it a case where Hackney is seeking to hold Rail for London to a promise it has made after entering into Lease to change the terms of Lease C. Here, Hackney’s argument is effectively that a post-contractual assumption as to the original meaning of the contract does not subvert the doctrine of consideration.

I can take Hackney’s other two arguments about Riverside more quickly. First, it was suggested in oral submission that Riverside could be regarded as per incuriam because Republic of India does not appear to have been cited to the Court of Appeal. I reject that submission, because I do not regard it as realistic that the Court of Appeal would or might have reached a different conclusion with the benefit of that authority given the reasoning in [65] and the reliance within that paragraph on the earlier Aristocrat decision. Second, it was suggested that Riverside, if it meant what Mr Hutchings contended, was inconsistent with later Court of Appeal decisions and those later decisions should be followed or it was at least arguable that they should. I shall deal with those below, together with one later case relied on by Mr Hutchings.

Mr Hutchings relied on the Court of Appeal decision in SmithKline Beecham plc v Apotex Europe Ltd [2007] Ch 271. In that case, the claimant gave a cross-undertaking in damages in respect of losses suffered by the defendant. Following the failure of the claim at trial, two companies affiliated to the defendant sought to enforce the cross-undertaking. One of the issues that arose was whether the claimant was estopped, whether through estoppel by convention or estoppel by representation, from denying that these affiliated companies were entitled to claim their own losses under the cross-undertaking. Jacob LJ, delivering the only reasoned judgment, held that Lewison J should have struck out the claim that the claimant was so estopped, for three reasons, one of which was that “an estoppel cannot be used as a key element of a claim (sword not shield) and particularly it cannot operate to create a legal relationship when there was none at the outset” ([103]).

As Jacob LJ concluded in [112], the case was analogous to Baird Textiles, because “it is a naked attempt to create a legally binding agreement when there never was one and never any intention to create one”. That is not the present case and therefore I do not consider that it provides decisive support for Mr Hutchings’ submission. It is true that, having stated that it would be astonishing if the estoppel could create a legal relationship between parties who had never even communicated with each other ([109]), he stated that in procedural terms an estoppel must inherently be raised as a riposte with the defendant raising an estoppel contending that the claimant could not assert particular facts ([110]), and that the affiliated companies’ case did not fulfil that model ([111]). However, there was no suggestion that he was intending to depart from the previous case-law in the area, including the summary of principle by Brandon LJ in Amalgamated Investments that was set out in Baird, which included that “while a party cannot in terms found a cause of action on an estoppel, he may, as a result of being able to rely on an estoppel, succeed on a cause of action on which, without being able to rely on that estoppel, he would necessarily have failed”. Therefore, it is difficult to read Jacob LJ’s judgment as going as far as ruling out entirely using estoppel as an element in a positive cause of action, rather than suggesting that there are limits to the ability to do so, such as where it is used to create a legal relationship from scratch. Moreover, while Hackney states in its submissions that the fact that it raises estoppel as a defendant to Rail for London’s claim for declaratory relief is not the decisive point, if SmithKline is read as stating that the procedural manner that estoppel is raised is decisive, then I note that Hackney is raising it as a riposte to Rail for London’s declaration.

In ING Bank NV v Ros Roca SA [2012] 1 WLR 472, which was relied on by Mr Datta, the defendant bank engaged the claimant bank to act as its financial adviser in its search for an investor on terms that provided for the latter’s fee to be calculated in a particular delay. In 2007, after the contract had been entered into and after a delay had occurred in bringing the investment transaction to a conclusion, the parties proceeded on a shared basis that the fee would be calculated on the basis of 2007 EBITDA. On the true construction of the contract the fee would have been based on 2006 EBITDA, which was higher. Carnwath LJ regarded the principles set out by Lord Steyn in Republic of India as a “succinct statement of the modern law” ([55]) and held that an estoppel by convention could arise. As such, an estoppel by convention was recognised in relation to the meaning of the contract in circumstances where the convention arose out of post-contractual events. Therefore, this provides some support for Hackney’s contentions. However, I would not go further than that, because the case was an example of the use of estoppel by convention to prevent a party to a contract relying on its full rights under it, as Rix LJ highlighted at [117] of his judgment.

Rivertrade Ltd v EMG Finance Ltd [2015] EWCA Civ 1295 concerned the attempt of the claimant (Rivertrade) enforce security that one of the defendants (Holdings) had purported to contract to provide Rivertrade with and purported to assign to the claimant. That security was actually held at the time by another company (Forburg) in the defendant’s group, having initially been held by a yet further company (Finance) in the defendant’s group. The Court of Appeal held that an estoppel by convention, asserted by Rivertrade through its reply, could be used by Rivertrade to debar Finance from asserting its title, such as it was, against Holdings, and in holding that Forburg was estopped from asserting that it had a title which was better than that of Holdings or that it had a right to the security that was higher than that of Rivertrade. Kitchen LJ, giving the reasoned judgment, considered that Forburg was, with Holdings, a party to the relevant contract with Rivertrade, and that it was the common intention of the parties that the benefit of the security should be transferred to Rivertrade. Therefore, “this was not a case in which an estoppel is relied upon to create an enforcement right where none previously existed. It is instead one of those cases in which the estoppel is relied upon to bind the parties to an agreement to an interpretation which would not otherwise be correct.” ([50])

While the Court of Appeal did not focus on this point, I agree with Mr Hutchings that this was a case where the estoppel arose at the point that the contractual arrangement in question was entered into, rather than afterwards. Therefore, while the estoppel by convention had the result of allowing Rivertrade to bring a claim that would otherwise have failed, it did not deal with the use of an estoppel by convention that is generated after the contract was entered into.

I was also referred to a number of other cases by the parties, but I did not find them of significant assistance in resolving the Application. Of the other estoppel cases I was referred to, (i) Nippon Menkwa Kabushi Kaisha v Dawson’s Bank (1935) 51 Ll L Rep 147 is a Privy Council decision that significantly pre-dates any of the modern case-law and the passage relied on does not seem to me to go further than Brandon LJ in Amalgamated Investments; and (ii) Combe v Combe [1951] 2 KB 215 is a promissory estoppel case. While the langauge of using the doctrine as a shield not a sword appears in the judgment of Birkett LJ in Combe, the casedoes not answer the question that arises in the estoppel by convention context of when, to use the language of Spencer Bower at [8.58], the estoppel by convention alleged is sufficiently promissory in nature that it should be subject to the same limitations as in the promissory estoppel case law. In particular it does not answer the question of whether a shared assumption that an agreement has a particular meaning rather than that its meaning has been changed by a gratuitous promise falls is regarded as engaging those limitations.

Drawing the above threads together:

Finally, I should deal with one question linked to point (4) above that arose in oral submissions. Mr Bhose submitted that Rail for London’s Particulars, which pleaded that the Basic Rent dropped to zero from the moment of the Surrender (paragraphs 25(d) and 26), left it open to Rail of London to seek to recover the rent paid since the Surrender, whether by bringing a restitutionary claim following success in the present claim or by later amending the pleading. He pointed out that in pre-action correspondence Rail for London had asserted a claim for the rent paid since the Surrender. He therefore submitted that insofar as the estoppel by convention pleaded by Hackney was meeting the claim for a declaration that no rent had been due between 2003 and 2019, it was purely defensive step that was a classic use of estoppel as a shield, and that it would be for the Court at trial to determine whether, if the estoppel was made out, it was also unconscionable to allow Rail for London to resile in respect of unpaid rent from 2019 from the assumption that rent would continue to be due after the Surrender. It would, he submitted, be open to the Court to find that the estoppel did not extend to the unpaid rent. Mr Hutchings stated that there was no intention to leave open the possibility of recovering the rent paid, and this was the intention behind his drafting of the prayer as seeking a declaration that no Basic Rent is due, so he offered to make any necessary amendment to make clear that the declarations sought only relate to the unpaid rent. Therefore, he contended, there was no defensive element of the use of estoppel here.

For the reasons given above, I do not need to rest my decision to refuse the Application on whether estoppel by convention is being used defensively here in respect of the period from 2003 to 2019. However, I would note that it appears to me to be a necessary step in the logic of the claim that, as pleaded in paragraphs 25(d) and 26, that no Basic Rent was due from 2003, and therefore to that extent the estoppel by convention argument is being used to cut off that argument at root, rather than the argument just applying to rent from 2019 onwards. Indeed, the prayer asks for a declaration that Basic Rent is not due without specifying that this is for the period from 2019 onwards.