Banca Intesa Sanpaolo SpA & Anor v Comune Di Venezia

These appeals from the orders of Foxton J dated 14 October 2022 and 8 February 2023 concern the validity of two interest rate swap (“IRS”) transactions (“the Transactions”) entered into in December 2007 between the claimant Banks (to which I will refer as “the Banks”) and the Comune di Venezia (to which I will refer as “Venice”) and Venice’s entitlement to restitution of sums paid under the Transactions. The judge found that the Transactions were void because Venice had no power to enter into them under Italian law (on the basis that the Transactions were speculative and involved recourse to indebtedness in breach of Article 119(6) of the Italian Constitution) which, in English law, means that Venice lacked capacity. He also found that Venice’s counterclaim for restitution of sums paid under the Transactions was not time-barred under the Limitation Act 1980, but that the Banks were in principle entitled to raise a defence of change of position to that counterclaim.

In broad terms, the Banks’ appeal contends that the judge erred in finding that (i) the Transactions (which were subject to English law and jurisdiction) were void because Venice lacked capacity under Italian law; and (ii) if the Transactions were void, Venice’s counterclaim for restitution was not time-barred by reason of section 32(1)(c) of the Limitation Act 1980. The Banks also seek permission to appeal on a ground which contends that the decision of the Italian Supreme Court in Banca Nazionale del Lavora SpA v Comune di Cattolica (2020) (“Cattolica”) upon which the judge placed considerable reliance in reaching his conclusions on Italian law should not be given retrospective effect. This was an argument which was not run by the Banks before the judge. By an order dated 27 April 2023, Males LJ ordered that the issue of whether permission to appeal should be granted should be considered on a rolled-up basis at the hearing of the appeals.

Again in broad terms, Venice’s appeal contends that the judge erred in finding that: (i) the counterclaim for restitution was governed by English, as opposed to Italian, law; and (ii) English law in principle affords a change of position defence to the Banks.

As the judge recorded in [17] of his judgment, the parties agreed a factual narrative on which he drew extensively for Section C of his judgment dealing with the facts and there was only a relatively limited area of factual dispute. In the circumstances I have been able to summarise the factual background in so far as it is relevant to the appeals in reliance on that Section C which contains a fuller analysis of the entire factual background.

On 23 December 2002, Venice issued a 20-year floating rate bond (the “Rialto Bond”), and around the same time entered an IRS with Bear Stearns (the “Bear Stearns swap”) for the same notional amount as the Rialto Bond in order to hedge its interest rate exposure. Under the Bear Stearns swap, Venice benefitted from both a cap on the variable rate payable under the Rialto Bond and a floor below which it would pay a fixed rate (a so-called “collar” transaction). The original maturity date for the Bear Stearns swap was 23 December 2005 subsequently extended in 2003 and 2004 to 23 December 2022, also with amendments to the rates payable. By the amendment effective on 23 June 2004, Venice was exposed for the period from 23 December 2006 to 23 December 2022 to the risk of paying a fixed rate of 5.45%, although it had the benefit of a cap of 7%.

In 2007, Venice proposed a restructuring of the Rialto Bond so as to free up resources in its municipal balance sheet. On 20 December 2007, it was restructured by extending its maturity date to 2037, with an amended coupon (the “Amended Rialto Bond”). As a consequence of that restructuring, the Bear Stearns swap was no longer aligned with Venice’s exposure to interest rate risk under the Amended Rialto Bond. However, Bear Stearns was not willing to amend the Bear Stearns swap and even threatened to terminate it. Accordingly, Venice agreed to a restructuring of the Bear Stearns swap with the Banks which took the following form. Bear Stearns, the Banks and Venice entered into ISDA novation confirmations under which 68% of the Notional Amount of the Bear Stearns swap was novated to Banca Intesa in return for a Novation Fee of €5,484,200 and 32% of the Notional Amount of the Bear Stearns swap was novated to Dexia in return for a Novation Fee of €2,580,800. The Novation Fees reflected the value to Bear Stearns (i.e. the negative mark to market (“MTM”) in its favour) of the Bear Stearns swap. The Banks thus stepped into the shoes of Bear Stearns, as is made clear by the terms of the ISDA Novation Agreement exhibited to the ISDA Novation Definitions. The ISDA Novation Definitions also make it clear that each party is acting on its own account, which negatives any suggestion that the Banks were acting as agents for Venice, a point to which I will return later in this judgment. The effect of the novations was to transfer to the Banks the future rights and obligations of Bear Stearns (by reference to which the negative MTM was calculated) in respect of their respective proportions of the novated amounts.

Since the terms of the Bear Stearns swap were not aligned with the Amended Rialto Bond, the existing swap as novated needed to be restructured. Therefore, Venice and the Banks agreed to new terms for the Transactions as set out in ISDA Master Agreements and trade confirmations. Under the Transactions with the Banks, the Banks and Venice agreed a new collar IRS with an extended maturity date matching the Amended Rialto Bond, under which Venice again benefitted from a cap on the variable rate that was payable under the Amended Rialto Bond and agreed to pay a fixed rate if the variable rate fell below the floor. Unlike the Bear Stearns swap, the terms of the Transactions matched the Amended Rialto Bond as to termination date, the interest rate paid by the Banks, the notional amount and the amortisation schedule. However, as would have been the case with any renegotiation between swap counterparties of an existing IRS transaction (for example if Bear Stearns had agreed the amendment Venice was seeking), the terms of the Transactions reflected the existing negative MTM (to Venice) of the novated amounts of the Bear Stearns IRS. Separately, the Banks also entered into “back-to back” swaps with other banks that hedged their exposure under the Transactions.

An important point which should be noted at this stage concerns the status and validity of the Bear Stearns swap. As Ms Sonia Tolaney KC for the Banks pointed out, Venice did seek to argue before the judge that the Bear Stearns swap was invalid so that the Transactions were void for common mistake. The basis for the alleged invalidity was that the Bear Stearns swap had not been approved by the City Council. That argument was roundly rejected by the judge at [324] to [330] of his judgment and is not subject to appeal by Venice. As Ms Tolaney KC also pointed out, Venice never sought to argue before the judge that the Bear Stearns swap was not a hedging transaction but was speculative. Although Mr Raymond Cox KC for Venice sought to suggest in his submissions that there was no basis for an assumption that the Bear Stearns swap was a hedge since the judge made no finding to that effect, given that the judge rejected the case of invalidity that was run by Venice and found that the Bear Stearns swap was valid, the matter is res judicata and this Court will proceed on the basis that the Bear Stearns swap was a valid hedging transaction.

On 31 October 2008, some six weeks after the collapse of Lehman Brothers, Venice received a letter from the President of the VIII Commission of the City Council of Venice querying whether the Transactions were binding and/or could be cancelled on the grounds that they were speculative.

In 2009 and 2010, disputes involving Italian local authorities which had entered into English law IRS transactions first began to be litigated in the Commercial Court. Many of these concerned jurisdictional disputes but the English Courts also engaged with the question whether these IRS transactions were precluded under Italian law. The most significant of these cases was Dexia Crediop SpA v Comune di Prato (“Prato”) in which Dexia commenced proceedings in December 2010 seeking sums due under swaps which had not been paid by Prato. The trial took place before Walker J in June and July 2014 and judgment was handed down on 25 June 2015 ([2015] EWHC 1746 (Comm)). In the light of the Italian law adduced before him in that case, Walker J rejected Prato’s arguments that the IRS transactions, which involved a number of restructurings rolling over negative MTMs from earlier transactions, were void for lack of capacity. He held that there was no general prohibition in Italian law on entering speculative swaps and the swaps did not constitute indebtedness contrary to Article 119(6). That latter finding was appealed and Walker J’s judgment was upheld by the Court of Appeal ([2017] EWCA Civ 428; [2017] 1 CLC 969).

Venice did not seek to dispute the validity of the Transactions at that time, but continued to pay sums which fell due under the Transactions. However, in June 2019, Venice commenced proceedings against the Banks in Italy claiming damages for breach of contractual and non-contractual advisory duties in relation to the Transactions. Those proceedings are ongoing. Then, in August 2019, the Banks commenced the present proceedings in the Commercial Court seeking declarations that the Transactions were valid and binding and alternative relief in contract and tort.

Then, on 12 May 2020, the Joint Sections of the Italian Supreme Court of Cassation, the highest civil court in Italy, issued Decision No 8770/20, Cattolica, holding, inter alia, that for the purposes of Italian law in effect until 2013, a local authority did not have capacity to enter into speculative derivatives and that certain types of swaps could constitute indebtedness for the purposes of Article 119(6). The judge found that this Decision represented a radical change from what Italian law was previously thought to be, albeit that there are decisions of lower courts, including the judgment of the Court of Appeal of Bologna in the Cattolica case itself, to the same effect. After the Decision of the Supreme Court, Venice served its Defence and Counterclaim in the present proceedings.

The Supreme Court Decision in Cattolica has spawned a fresh wave of litigation in the Commercial Court in which Italian local authorities seek to argue that IRS transactions entered by them are void for want of capacity on their part. The first such case to reach trial was Deutsche Bank AG v Comune di Busto Arstzio (“Busto”), heard by Cockerill J, in which she handed down judgment on 12 October 2021 ([2021] EWHC 2706 (Comm)). She reviewed the Decision in Cattolica and upheld the validity of certain collar IRS transactions, whilst expressing some doubts about the reasoning of the Supreme Court.

The trial in the present case then took place in June and July 2022.

Sections A to C of the judgment ([1]-[74]) are introductory and cover the factual and procedural background, including the judge’s views on the experts and witnesses who gave evidence. The judge made findings on three key factual areas of dispute at Section D: [75]-[100], not relevant for the purposes of this appeal. At Section E: [101]-[106], he reviewed the nature and components of IRS transactions. At Section F ([107]-[127]), he reviewed the analytical framework for the different legal issues. First, he concluded, following and applying the decision of the majority of the Court of Appeal in Haugesund Kommune v Depfa ACS Bank [2010] EWCA Civ 579; [2012] QB 549 (“Haugesund”), that the characterisation of whether an issue of Italian law raised an issue of “capacity” or some other kind of challenge to the validity of the Transactions was a matter of English law. In that context he also cited the judgment of Lord Leggatt in SR Properties v Rampersad (“Rampersad”) [2022] UKPC 24 at [23]-[24] distinguishing between issues of capacity, illegality and authority.

At [115] the judge considered the issue of the date on which the content of Italian law is to be ascertained, noting that issues of whether Venice had capacity to enter the Transactions or whether they were illegal under Italian law were to be determined by reference to the law in force when the Transactions were entered into. He cited the decision of the House of Lords in Adams v National Bank of Greece [1961] AC 255, where bonds issued by a Greek bank with an associated guarantee from another bank in 1927 were governed by English law. In 1953, a successor bank succeeded to the rights and obligations of the guarantor bank by virtue of Greek legislation. However in 1956, the Greek parliament passed further legislation retrospectively excluding the obligations under the bonds and the guarantee from the scope of the succession. The House of Lords concluded, as the judge said for differing reasons, that this legislation did not have the retrospective effect of discharging the successor bank’s liability under the bonds and the guarantee. As the judge said at [118] this approach is relatively easy to apply in the case of legislation (as in Adams) or a government decree such as in Lynch v Provisional Government of Paraguay (“Lynch”) (1871) LR 2 PD 268 but more difficult to apply to court decisions. He noted that in our own legal system case law is a source of law which adopts the “declaratory” theory as to the effect of judicial decisions on points of law. He also noted that the issue of whether there could ever be a post-contractual change in the interpretation or application of legislation in another jurisdiction, in force when an English law contract was concluded, of so significant an effect that an English court would refuse to give effect to that change on the basis of the principles recognised in Adams, was not raised before him. This is the issue which the Banks do now seek to raise in their Ground 4 on which permission to appeal is sought.

He went on to consider the approach to ascertaining the content of foreign law, saying that there was no challenge to the summary of the principles by Cockerill J in Busto at [105]-[108]. That summary included the statement by Walker J in Prato at [128] which was accepted by both parties before this Court as stating the applicable principle:

“The task for the Court is to evaluate the expert evidence of Italian law and to predict the likely decision of the highest court in the relevant Italian system of law if this case had been litigated there on each of the points in dispute. As explained below, these courts are the Council of State for administrative law matters and the Court of Cassation for civil law matters.”

Cockerill J went on to consider to what extent she could diverge from the decision in Cattolica concluding at [108] that she could do so if satisfied that, however high an authority, it did not represent the law, but would have to be astute to give it full weight before concluding that that was the correct course. In the present case, the judge engaged in further analysis at [121] to [127] of the approach the English court should adopt if asked to accept expert evidence that a decision of the Italian court was wrong. However, since the Banks did not argue on this appeal that Cattolica was wrongly decided and did not represent Italian law, it is not necessary to refer to that analysis.

At Section G the judge turned to consider the issue of whether Venice lacked capacity to enter the Transactions. He began by reviewing the key legislative and administrative instruments dealing with Venice’s capacity. He began with Article 119(6) of the Italian Constitution, originally adopted in 1947:

“Municipalities, Provinces, Metropolitan Cities and Regions have their own assets, allocated to them pursuant to general principles laid down in the State law. They may have recourse to indebtedness only for the purpose of financing investment expenditures [with the simultaneous definition of amortization plans and provided that the budget balance is complied with reference to all entities of each region]. Any State guarantee on loans taken out by them is excluded.”

The italics in that citation were added by the judge and he noted that the words in square brackets were added by Constitutional Law No 1/2012.

Article 2 of Regulation 420/1996 required local authorities to hedge against exchange rate risk. In 1999, the Italian financial regulator (CONSOB) issued a determination (the “CONSOB Determination”) clarifying that a transaction qualified as hedging if three conditions were met (set out by the judge at [133]), only the first and second being relevant for the purposes of the appeal:

“i)

The transaction is explicitly carried out to reduce the risks connected with an underlying debt instrument. [“limb (a)”]

ii)

There is a ‘high correlation’ between the characteristics of the underlying debt and those of the derivative transaction. [limb (b)]”

The judge referred at [136] to Decree 389 issued in December 2003 by the Italian Ministry of Economy and Finance (“MEF”), Article 3.2 of which allowed derivative transactions of various kinds, and at [137] to [139] he quoted Article 3.17 and 3.18 of Finance Law 350/2003 which addressed the meaning of the concepts of “indebtedness” and “investment” in Article 119 of the Constitution.

The judge went on to address the clarifications provided by the MEF in various circulars as to the proper interpretation of the legislation and as to the types of derivatives that could be entered into and whether they constituted indebtedness. He then outlined the evolution of legislation after the Transactions were entered into, concluding at [148] that, after an initial liberalising of the ability of local authorities to enter into derivative transactions, there was a progressive tightening of the position, culminating in a general prohibition (with limited exceptions) effected on 27 December 2013 by Law 147/2013.

The judge then dealt with the market backdrop to the evolution of that legislation, noting that, from the end of 2000 onwards, Italian local authorities took advantage of their new ability to enter IRS transactions. When interest rates began to rise in 2005, a number of these were restructured on revised terms. This was often done by adjusting the terms of the restructured IRS in such a way as to make the bank whole in respect of the negative MTM under the original IRS. The judge noted that the 2008 financial crisis led to a dramatic reduction in interest rates which meant many local authorities found themselves paying fixed rate interest on their borrowings but receiving much lower floating rate payments under their IRS. As he said at [152], increasingly disadvantageous IRS terms against prevailing market rates and the corresponding strain on local authority finances resulted in a significant volume of litigation. Where IRS transactions were entered into on the terms of the ISDA Master Agreement, they were subject to English law and jurisdiction, so that many disputes appeared in the Commercial Court.

The judge went on to summarise the cases which were heard, which it is not necessary to repeat here, simply to note that the first trial addressing arguments referring to Article 119 and Law 350/2003 was Prato, the decisions in which both before Walker J and the Court of Appeal I have summarised at [10] above. The judge also summarised the decision of Cockerill J in Busto, noting at [166] that that was the only previous English case to engage with the Supreme Court decision in Cattolica, which is a significant development in the response of Italian courts to swaps contracts entered by local authorities. This had to be borne in mind when considering findings of Italian law made by English courts before Cattolica.

At Section H of the judgment, the judge analysed the Cattolica decision, starting with the legal context, noting that the financial pressures which IRS transactions imposed on those who had entered into them had led to a renewed legal focus on them. He summarised the legal theories in circulation when Cattolica came to be determined, which in broad terms focused on arguments as to whether IRS transactions were gaming transactions and whether there were breaches of investor protection legislation. He then summarised the evolution of the Cattolica litigation and considered the practical status of court decisions in the Italian legal system, concluding at [184] that decisions of the Supreme Court have particular normative force as “the highest court of judicial…jurisdiction” and that even greater normative force attaches to decisions of the Supreme Court sitting in Joint Divisions as happened in Cattolica.

The judge then set out and adopted at [185] Cockerill J’s summary at [137] of Busto of the structure of the Supreme Court judgment. At [186] he summarised the issues which arose before him as to the effect of the Cattolica judgment, only two of which are relevant on this appeal:

“i)

Does Section 8 of Cattolica hold that Italian local authorities lack capacity to enter into speculative derivative transactions, and, if so, was that decision correct as a matter of Italian law?

ii)

Did Sections 8 and/or 10 of Cattolica hold that swaps were a form of indebtedness (whether for the purposes of Article 119 of the Italian Constitution or otherwise) and that local authorities did not have capacity to enter into them other than for the purpose of financing expenditure? …”

In the next Section of the judgment, the judge summarised the submissions on the speculation issue, and his analysis as to what Cattolica decided on this issue as a matter of Italian law. He concluded at [196] that although the reasoning may not be entirely satisfactory, what was described as the “legal and axiological framework” in Cattolica at [8.3] “is clear enough”, as follows:

“i)

A public authority had contractual capacity to conclude derivative contracts until the 2013 Finance Law came into effect.

ii)

However, only in the case of a hedging (and not a speculative) derivative “could ... a local authority be said to have capacity to enter into them”.”

At [197] he rejected the Banks’ argument that these paragraphs only involved a finding that, as a matter of Italian law before an Italian court, a local authority was prohibited from entering into such transactions, rather than lacking substantive power or legal ability. He was satisfied that the decision was that local authorities lacked the substantive power to enter into speculative derivative contracts rather than that they were acting illegally in doing so. He also rejected the Banks’ argument that, if that was what Cattolica decided, it was wrong as a matter of Italian law. Since that argument is not repeated on appeal, it is not necessary to summarise the judge’s detailed reasons for rejecting it.

The judge then turned to his analysis of when a derivative is “speculative” for Italian law purposes, noting at [202] that it was common ground that while the Italian legal or regulatory regime treats the question of whether a derivative is a hedge or speculative as significant for certain purposes, Italian law does not provide a definition of what constitutes a speculative derivative.

He said at [203] that the evidence as to what made a derivative a hedge or speculative comprised: (i) evidence of Italian law (both from the experts and from Italian case law); (ii) reference by Venice to English case law on this topic and (iii) evidence from the two market practitioners as to their understanding. He found the evidence in the first category of greatest assistance.

He noted the evidence of the Banks’ expert, Professor Gentili, that the concept of hedging had de facto achieved a specific and legal meaning under Italian law, namely that the derivative satisfied the CONSOB Determination test (set out at [20] above). The judge accepted that the CONSOB Determination was of assistance when determining whether a derivative was a hedge, noting that it had been relied on in a number of Italian court decisions, including the Supreme Court in Decision No. 19013/2017. However, the judge said it was clearly not exhaustive. It was not formulated with specific considerations regulating local authority finance in mind and did not deal with more nuanced questions. A number of Italian court decisions had addressed the issue without reference to the CONSOB Determination or the three-stage test there set out. He then set out at [212] an analysis of a number of Italian court decisions, mainly from courts of first instance and courts of appeal, although he noted the Supreme Court Decision No. 21830/2021, which concluded that a vanilla IRS swap transaction was a hedge and not speculative.

He then referred to a number of English authorities on which Mr Cox KC for Venice had relied. The judge placed only limited weight on these although he said that it was interesting that many of the factors highlighted in the Italian case law are echoed in the English cases. He was not assisted by the evidence of the market participants as to what does and what does not constitute speculation for the purposes of the Italian law question he had to decide. At [222] he concluded:

“I will not attempt to formulate a definitive test of what makes a derivative speculative as a matter of Italian law, when the Supreme Court in Cattolica did not itself do so, and when it remains possible to apply a restriction by reference to that criterion without doing so ([215]). The Italian case law identifies a number of indicia or features which, either individually or in combination, may have the effect that a derivative is a hedging transaction, or a speculative transaction (many of which, as Mr Cox KC submitted, are reflected in English case law on the same topic).”

The judge then turned to consider whether the Transactions in the present case were speculative. He considered at [224] that in closing Mr Cox KC had cut right to the heart of the matter, the fact that the terms of the swap had been structured so as to cover amounts which the Banks had to pay to Bear Stearns to close out the Bear Stearns swap. At [225] he set out what he considered the essential facts relating to Venice’s decision to enter the Transactions.

The judge’s analysis and reasoning at [226] to [232] as to why he concluded that the Transactions were predominantly speculative are of sufficient significance to the determination of this appeal to merit citation in full:

“226.

The impact on the terms of the Transactions of structuring the collar to cover the costs of winding up the Bear Stearns IRS was considerable:

i)

This was the principal reason why the Transactions had a very significant MTM in the Banks’ favour from the outset (a combined positive Day 1 MTM of c €10.5m in the Banks’ favour).

ii)

This was the principal reason why the value to the Banks of the interest rate floor (estimated by the experts at between €12.4m and €12.974m) was more than five times the value to Venice of the cap (estimated by the experts at between €1.7m and €2.4m).

iii)

This was the principal reason why, on the basis of a Day 1 statistical probabilistic calculation, the probability of Venice losing money on the Transactions was high:

a)

On Ms Bowie’s calculations, the probability of a negative pay-off for Venice under the Transactions was between 77.1 and 78.7% (depending on whether or not the calculation is performed on an “absolute” basis or on a basis which discounts future cashflows to present value), whereas if the amount paid to wind up the Bear Stearns IRS is removed from the calculation, the figure is 59.3%.

b)

Mr Malik did not put forward his own calculation of the probability of a negative pay-off for Venice under the Transactions or challenge Ms Bowie’s calculation of 78.7%, saying that in his experience banks did not produce such calculations and he did not consider that they had utility for customers. He did perform a calculation removing the amount paid to wind down the Bear Stearns IRS from Ms Bowie’s calculation and arrived at a figure of 57.3%. The Banks’ closing submissions did not challenge Ms Bowie’s figure, but noted that “stripping out” the Bear Stearns IRS wind up cost reduced that figure to 57.3% on Mr Malik’s figures.

iv)

On Mr Malik’s evidence, it led to the floor being between 80 and 100 basis points higher than it would otherwise have been.

v)

On Ms Bowie’s calculations, it meant that the Transactions involved a modelled “realistic worse case” outcome for Venice of the order of €70.6m (modelling to a 95% confidence level). Mr Malik gave evidence that the MTM distribution analysis which Ms Bowie had performed would be of limited use to customers, and that banks did not provide MTM distribution analyses to customers. He did accept, however, that the effect of including the large negative MTM from the Bear Stearns IRS was to lower the probability of the Transactions being positive in the future.

227.

In my view, Mr Cox KC was right to submit that, in addressing the cost of winding up the Bear Stearns IRS through the terms of the Transactions, Venice was obtaining the possibility that interest rate movements during the life of the Transactions would be such that Venice would not have to pay a sum equivalent to the wind-up cost, but in return, was running the risk that interest rate movements during the life of the Transactions would be such as to lead to it paying a great deal more.

228.

The decision to address the cost of winding-up the Bear Stearns IRS within the terms of the Transactions had other consequences:

i)

It meant that the terms of the Transactions were in material and financially significant respects (the level of floor and cap) not determined by the terms of the Rialto Bond (although I accept that important terms were so determined – the Notional Amounts, the amortisation rate, the maturity date and the interest rate received by Venice from the Banks).

ii)

It meant that the minimum interest rate which Venice was committing to pay was not aligned with the forward rate curve at the time of contracting.

iii)

It involved the assumption by Venice of a new and significant risk (viz of having to pay interest to the Banks at the floor level while receiving interest payments at a much lower rate) which did not arise under the Rialto Bond. While the character of the Transactions must be determined ex ante, some indication of the degree of risk run can be seen in the fact that by the end of the most recent payment period (24 June 2022), Venice had made total payments to the Banks of €70,995,695.95 (in part because EURIBOR 6m became negative in November 2015).

229.

Standing back, therefore, and considering the matters discussed in [226] to [228] above:

i)

The Transactions were explicitly carried out in the terms adopted both to reduce the risks connected with the Rialto Bond and to cover the winding-up costs of the Bear Stearns IRS (CONSOB Determination (a)).

ii)

While many of the terms of the Transactions matched the financial characteristics of the Rialto Bond, important and financially highly significant terms were arrived at for other reasons (CONSOB Determination (b); the Court of Cassation Decision No 19013/2017, the Court of Rome 8 January 2016, Decision No. 212, the Court of Novara, 24 July 2012, Decision No. 569, and, the Court of Turin, 21 October 2021, Decision No. 4685).

iii)

There was a very significant difference between the MTM of the cap and the floor, such that Venice was providing the Banks with a protection of a significantly greater value than the protection it was obtaining from the Banks (the Court of Appeal of Milan in Decision No 2393/2020 and cf Standard Chartered Bank v Ceylon Petroleum Corp [2012] EWCA Civ 1049, [9]-[12]).

iv)

The fact that the desire to cover the winding-up costs of the Bear Stearns IRS was a highly significant factor in setting the terms of the Transactions itself pointed to the speculative character of the Transactions (Decision of the Court of Orvieto of 12 April 2012 and of the Court of Turin of 21 October 2021 in Decision No 4685/2021). It meant that the Transactions were, to a significant extent, serving the purpose of seeking to address a past adverse event (cf Standard Chartered Bank v Ceylon Petroleum Corp, [9]-[12]).

v)

The significant non-alignment between the terms of the Transactions and the prevailing forward rate curve was also suggestive of speculation (Supreme Court Decision No 21830/2021).

vi)

The fact that Venice took on a significant new risk to which it was not exposed under the Rialto Bond was also suggestive of speculation (Professor Alibrandi’s evidence and cf. Credit Suisse International v Stichting Vestia Groep [2014] EWHC 3103 (Comm), [217], UBS AG v Kommunale Wasserwerke Leipzig GmbH [2014] EWHC 3615 (Comm), [159] and Standard Chartered Bank v Ceylon Petroleum Corp [2012] EWCA Civ 10494, [9]-[12])).

230.

As Venice submitted in its closing, the structuring of the Transactions to cover the costs of winding-up the Bear Stearns IRS with its substantial negative MTM was:

“akin to borrowing money but instead of repaying it on predictable terms, entering into a bet with a range of possible outcomes. Venice might never have had to repay the Bear Stearns money at all, if rates had suddenly risen to well above the cap and stayed there such that it was in the money throughout the life of the swap. Conversely, Venice might – as has in the event occurred – have had to pay it back many times over, because its impact on the floor level [has] resulted in Venice paying much more than would otherwise be the case. The only rational basis for proceeding in such a way is the possibility that the bet could have worked out better for Venice than if it had simply paid the Bear Stearns break cost itself …

Borrowing money on terms that one might never have to repay it, might have to repay a much greater sum, or might have to pay anything in between for it depending on where interest rates sit, is speculation.”

231.

Having regard to the cumulative effect of these factors, and regardless of whatever uncertainties might arise at the fringes of this debate, I am satisfied that an Italian court would clearly find that the Transactions were speculative for the purpose of the legal restriction under Italian law formulated in Cattolica. The most that can be said is that the Transactions served mixed motives (as I accept). However, the Court of Appeal of Milan in Decision No 2393/2020 noted that:

“An invalidity that - due to its speculative characteristics - must therefore affect the entire contract, and not only the part that concerned the imbalance between the MTM of the cap option and the MTM of the floor option, as also argued in the alternative by the appellant and, in the opinion of this Court, without foundation.”

232.

In any event the significance of the speculative elements of the Transactions (as outlined at [226]-[228] above) were such that the Transactions can fairly be characterised as predominantly speculative.”

The judge then dealt at Section J with Indebtedness. He rejected Venice’s primary case that all IRS transactions involved recourse to indebtedness for the purposes of Article 119(6) (which is not an issue on this appeal). He then considered its alternative argument that the Transactions themselves breached Article 119(6) because they involved recourse to indebtedness otherwise than for the purpose of financing expenditure. He considered the position under Italian law leaving Cattolica aside and then in the light of it. He concluded at [252], as Cockerill J had held in Busto:

“In short, Cattolica holds that Article 42(2)(i) (and by implication Article 119(6)) applies to the following swaps:

i)

Swaps “if they are of the type with an upfront loan”. This picks up the statements that a swap with an upfront provision constitutes indebtedness in [10.1.2] and [10.1.3].

ii)

If the negotiation of the swap “entails extinction of the previous underlying loan agreements”. This picks up the reference to the termination of indebtedness in [10.5].

iii)

If the negotiation of the swap entails significant modifications of the underlying loan agreements, even if they remain outstanding. This picks up the reference to amending the underlying contract including “by extending the length of the debt exposure” in [10.5].””

He had concerns about the Supreme Court decision, but did not feel able to say that it did not represent Italian law.

He then applied his findings to Venice’s so-called Third Argument that the Transactions involved an “upfront” in the form of the payment made to unwind the Bear Stearns swap which Venice contended was “priced in” to the terms of the Transactions. He concluded at [261]:

“I am satisfied that the amount paid by the Banks to Bear Stearns and then “embedded” into the terms of the Transactions constituted an “upfront” for the purposes of the Cattolica principles:

i)

The rationale for treating a swap with an upfront payment by the bank to the local authority as expenditure or indebtedness is because it involves taking a benefit at one point in time (and in one financial year) in return for structuring the transaction in a manner which, in Day 1 PV terms at the date of transacting, is adverse to the local authority, with the attendant enhanced risk of payments by the local authority in subsequent financial years. It may be that there will never in fact be a negative cash outflow by the local authority (because the market moves in a manner which ultimately reverses that adverse Day 1 PV from the local authority’s perspective). However, Cattolica decides that, as a matter of Italian law, that risk is sufficient to engage Articles 119(6) and 42(2)(ii).

ii)

That rationale is equally applicable in the present scenario. There was “jam today” (in that the funds were made available to meet the price of exiting a transaction which Venice wished to exit) in return for accepting a greater risk of bare bread tomorrow. The fact that the “upfront” here was not paid to neutralise an imbalance in the MTM of the respective obligations, but the respective obligations are structured in an unbalanced way to cover the cost of the “upfront” does not negate the issues of inter-budgetary equity which Article 42(2)(i) recognises nor the limits of Article 119(6) as established in Cattolica.

iii)

I accept Mr Cox KC’s argument that the fact that the payment in question moves from the Banks to Bear Stearns rather than through Venice does not change the analysis. Professor Gentili accepted in cross-examination that if A (sc Venice) had asked C (sc the Banks) to make the payment to B (sc Bear Stearns), it would still be treated as an upfront payment by C to A (Day 6/107). That is essentially what happened.

iv)

While in no way determinative, it is noteworthy that the Banks referred to the payment being made to Bear Stearns to unwind the Bear Stearns IRS as an upfront. A particularly telling internal communication, in the context of the case as a whole, was a note prepared by Intesa in 2009 when explaining the high negative MTM of the Transactions from Venice’s perspective by “the need to absorb the Upfront paid to the municipality for the early termination of the derivative with Bear Stearns”.”

At [268] the judge concluded in relation to the second element of Article 119(6), “for the purpose of financing investment expenditure”, that it followed from his conclusion that the Transactions, as a whole, were speculative that they were not undertaken for the purpose of financing investment expenditure. In any event, the upfront payment was the loan element which rendered the Transactions a recourse to indebtedness (per Cattolica), and that was entered into in order to meet the winding-up costs of the Bear Stearns IRS (even if the Rialto Bond was issued to finance expenditure). This was the background to the second swap transaction in Cattolica itself. The upfront paid to the benefit of Venice did not in any way reduce or replace the outstanding amount under the Rialto Bond, but created “new debt”.

In the next Section of the judgment, the judge dealt with the consequences under English law of the Italian law position, specifically whether, applying English conflict of law principles, the effect of his findings of Italian law (that the Transactions were speculative or contravened Article 119(6)) was that Venice did not have capacity to enter into the Transactions and, accordingly, that they were not valid. He noted at [270] that he had earlier concluded that, on the basis of Cattolica, the restriction on local authorities entering speculative derivatives which that decision derived from Article 119(6) had the effect that local authorities had no substantive power or legal ability to enter into such transactions rather than the measure being one which prohibited a local authority from entering a transaction which it had power to enter into. He considered the language of Article 119(6) (local authorities “may have recourse to indebtedness only for the purpose of financing investment expenditures”) was itself suggestive of a limitation on the power of local authorities.

Applying the criteria of English law as to whether an issue is properly categorised as one of capacity and specifically the test adopted by the majority of the Court of Appeal in Haugesund, he concluded at [271] that, as a matter of English law classification, Article 119(6) was a restriction on capacity and the consequences for that conclusion on an English law contract were a matter for English law. He said at [274]:

“Applying English law, and on the basis of Haugesund, the inevitable consequence of my conclusion that, on the basis of the Speculation and/or Indebtedness Arguments, Venice lacked the substantive power or legal ability to enter into the Transactions, is that they are void”.

He said that he had reached this conclusion with some diffidence, given that it results from a decision of the Italian Supreme Court reached some 13 years after the Transactions were entered into which “completely altered the legal landscape”. He noted there may be room for a legitimate debate as to whether the security of obligations governed by English law: “should be capable of being subject to a continuing jurisprudential jeopardy of this kind arising from the courts of the domicile of one of the contracting parties”, referring back to what he had said at [116]-[119], which I summarised in [16] above.

Having considered various issues which are not relevant to this appeal, the judge turned in Section Q to Venice’s restitution claims. He identified at [385] the three issues which arose:

In relation to the first issue, the judge noted that, while Article 10(1)(e) of the Rome Convention (which was applicable at the relevant date of the Transactions) established that issues stemming from the consequences of a contract being void are governed by the law which would apply if the contract had been concluded, the United Kingdom had entered a reservation as to that provision when adopting the Rome Convention. Accordingly, the common law principles of conflicts of law continued to apply.

At [388] the judge noted that in Prato, Walker J had concluded that the putative choice of English law to govern the swap if valid was not a sufficient connection with England to outweigh the many connections with Italy. He also noted that in Busto Cockerill J, albeit obiter, had reached a different conclusion.

At [390], the judge concluded that the unjust enrichment claim had its closest and most real connection with English law by reason of the choice of law clause in the Venice Master Agreement for the following reasons:

“i)

Article 8(1) of the Rome Convention recognises that where the issue arises as to the existence of an agreement which, if concluded, would be subject to English law by virtue of a choice of law, English law will be applied in determining whether or not a contract has been concluded. That reflects the importance attached to a putative applicable law, even when there is a dispute as to whether or not a contract was concluded.

ii)

On that basis, English law has been applied in this case to issues relating to the Transactions such as the consequences of Venice’s lack of capacity and whether the lack of actual authority on Mr Dei Rossi’s part is sufficient to render the Transactions void.

iii)

Those matters are sufficient to show that, even when the validity of the contract is in dispute, or it has been determined that the contract is void, the parties’ putative choice of English law is still legally significant.

iv)

Further, there is, at least, a logical connection between the system of law which decides that a contract is void (English law in this case), and the law to be applied in determining what the consequences of it being void are on the parties to the extent they had purported to perform it. [The judge then cited an observation by Jenkins LJ] in Arab Bank Ltd v Barclays Bank (Dominions, Colonial and Overseas) [1953] 2 QB 527, 572]

v)

While the issue does not appear to have been raised directly in Haugesund, Aikens LJ appears to have assumed that English law governed the restitutionary claim in respect of a void contract which would have been governed by English law, and addressed the significance of the foreign statute to such a claim not on the basis that the case was being argued by reference to English law only as a matter of convenience but on the basis that English law was indeed applicable (see [97]- [100]).

vi)

Finally, the payments in question were made by Venice on the understanding that they were required by English law obligations, in discharge of English law debts, with the Banks having the same understanding in receiving them. The fact that the payments were made and received on the basis of assumed English law obligations is, to my mind, highly significant, it being the natural expectation in those circumstances that English law would apply to issues relating to security of receipt and rights of recovery.”

At [393] the judge noted that it was common ground that there was no change of position defence as a matter of Italian law. It was accepted that the defence was available in English law but its availability on the facts was in dispute. The judge noted that he had held that a change of position defence was available in the case of an ultra vires contract in a non-swaps context in School Facility Management Ltd v Governing Body of Christ the King College [2020] EWHC 118 (Comm) which was upheld in the Court of Appeal ([2021] EWCA Civ 1053). He also noted criticisms and concerns about that decision raised by academic commentaries. Accordingly he sought to arrive at the answer through first principles.

At [399] the judge considered various “fixed positions”, the first of which was where a recipient acquires a benefit on what appears to be an unconditional basis and spends money as a result, where it is generally accepted that a change of position defence will apply. He contrasted the position in Haugesund, where a local authority used the upfront receipt from an ultra vires swap transaction to make investments which ultimately failed. When the counterparty claimed for restitution, the local authority attempted to rely on the failed investments to support a change of position defence. It was clear that no such defence was available, given the local authority’s awareness that it would always have been required to pay the amount back. He also noted that the defence will not be available where money is paid to a recipient subject to a condition which the recipient knows has yet to be satisfied and the recipient spends the money for its own purposes.

He considered the present case raised a different problem, as he put it at [400]:

“the parties shared a common understanding that they owed each other binding obligations, on the faith of which one of them changed his position, and then faced an unjust enrichment claim from the other when the true position became apparent”.

He said that the idea that a change of position defence can never be available in a void contract case was not attractive. It paid little regard to protecting security of receipt and those who have conducted themselves on the basis of appearances which underlie the defence, long standing concerns of English commercial law. The judge considered the present case might be said to be even stronger than his first category in [399] because here both the payer and the recipient: “were acting on the basis of an apparent state of affairs that the condition for Venice’s payments (the existence of legally enforceable rights to counter-payments) had been satisfied”.

The judge said that it was routine and objectively foreseeable that banks entering into transactions of this type will hedge them. He was satisfied that a change of position defence was in principle available in this case notwithstanding that Venice’s right to restitution arises from the fact that a condition of the payments (a legally enforceable right to the counter-payments) was not satisfied.

The judge then dealt with Venice’s second objection to the change of position defence, that the relevant change of position, the entry into the hedging swaps, occurred before receipt of the payments of which Venice sought restitution and that the payments under those hedging swaps were made because of the legal liability to do so under the swaps. The judge noted two first instance decisions which had held that those objections are fatal to any attempt by a bank to rely on its liabilities under a hedging swap as giving rise to a defence of change of position to a claim by its original swap counterparty to recover payments made under a void swap, the decision of Hobhouse J in Westdeutsche Landesbank Girozentrale v Islington LBC (“Westdeutsche”) [1994] 4 All ER 890 and that of Clarke J in South Tyneside MBC v Svenska International plc (“Svenska”) [1995] 1 All ER 545.

He cited passages from those judgments and at [405] identified a number of “threads” or, effectively, objections to the change of position defence running through them. The first was that the banks had committed themselves to the terms of the hedging swap in advance of the receipt. The judge said that the law had moved on considerably since the decisions in those cases, noting that it was now clear from the decisions of the Privy Council in Dextra Bank and Trust Co Ltd v Bank of Jamaica [2002] 1 All ER 193 and the Court of Appeal in Jones v Commerzbank AG [2003] EWCA Civ 1663, ([38] and [47]) that the defence of change of position can be established by action taken before, but in anticipation of, the receipt of the amounts of which repayment is sought. He cited passages from the Privy Council judgment which, inter alia, referred to the exclusion of anticipatory reliance in Svenska having depended on the “exceptional facts” of that case, whilst recording that the decision of Clarke J had been the subject of criticism.

The second thread was that the banks would be founding the defence on the apparent validity of a void transaction with a public authority. The judge noted that he had rejected this argument in Schools Facility Management, stating that “there is no principled basis for the distinction which the College invites me to draw in its submissions between anticipatory and consequential change of position in public authority cases.” Venice did not dispute that general proposition. It placed particular emphasis on the third thread, that the banks’ decision to enter the hedging swaps was a wholly independent decision taken for their own purposes. The judge rejected that argument, concluding at [411]: “…the decision to incur the obligation (in certain market conditions) to make payments under the hedge was undoubtedly taken in anticipation of the fact that, in those same market conditions, the bank would receive a largely equivalent payment from its counterparty under the impugned swap.” Having cited Lord Burrows’ summary of the law in A Restatement of the English Law of Unjust Enrichment (2012), the judge said at [413]: “I can find nothing in that summary which would deny the Banks a change of position case where they had entered into back-to-back transactions by which they assumed (conditional) payment obligations in anticipatory reliance of receiving essentially the same payments from Venice.”

The final thread was that it was not known whether entry into the hedging transactions would, or would not, ultimately prove to be economically disadvantageous for the banks. Having considered various sub-issues, the judge rejected that argument and concluded that a change of position defence was available in principle whilst noting that issues of quantification would be for a subsequent trial.

The judge then considered the third issue set out at [41] above, whether Venice’s restitution claims were time-barred under English law. He noted that proceedings were commenced on 15 August 2019, so that prima facie payments made by Venice prior to 15 August 2013 are not recoverable. However, Venice relied on section 32(1)(c) of the Limitation Act 1980, contending that the payments were made under the mistaken belief that the Transactions were valid and binding and that it could not with reasonable diligence have discovered its mistake until the decision of the Supreme Court in Cattolica.

The judge noted at [429] that it is clear as a matter of law that local authorities who paid amounts believed to be due under swap contracts could rely on section 32(1)(c) when seeking to recover those payments on the basis that their belief involved a mistake (Kleinwort Benson Limited v Lincoln City Council [1999] 2 AC 349). That conclusion was restated but with an important qualification, in Test Claimants in the FII Group Litigation v Revenue and Customs Commissioners [2020] UKSC 47; [2022] AC 1 (“FII”), where the Supreme Court held that time would run not from the date when the “truth” of the position has been established by an authoritative determination, but from the date when the paying party ought through due diligence to have appreciated that it had a worthwhile restitutionary claim. The judge set out the summary by the majority of the Supreme Court of the effect of their conclusions at [213] of their judgment.

The judge set out his conclusion applying the FII test to the facts at [430]:

“i)

England and Wales is the contractual forum in which Venice’s claims had to be brought.

ii)

A noticeable feature of this case, therefore, is that the commencement of proceedings in the contractual forum could only challenge the position under Italian law as a matter of fact, rather than by (for example) seeking to take the point to the Supreme Court to determine the position under English law.

iii)

So far as the likely position in English proceedings is concerned as noted at [381] above, as late as 2015, Walker J rejected very similar arguments in the Prato case, and an appeal against the Article 119 aspects of that decision failed in 2017, both decisions not treating the Bologna Court of Appeal decision in Cattolica as sufficient. Those findings could have been given the status of prima facie evidence in any English proceedings (see [165(ii)), and in any event are likely to have strongly influenced any English judge. Indeed, given the terms of the Court of Appeal judgment quoted at [162], it is difficult to see how the claims would have been viable in the absence of a decision at a higher level.

iv)

I have already explained at [276] my reasons for concluding that the decision of the Supreme Court in Cattolica represented a fundamental change in the interpretation of the relevant legislative and regulatory provisions.

v)

For those reasons I am satisfied that, exercising reasonable diligence, Venice could not have discovered that it had a “worthwhile claim” prior to the Cattolica decision in the Supreme Court.

vi)

It follows that none of Venice’s claims for restitution are time-barred.”

The judge then dealt with various counterclaims by Venice that are not relevant to this appeal.

The Grounds of Appeal

57.

The Banks pursue five Grounds of Appeal:

(1)

The judge wrongly held that the Transactions were ‘speculative’ under Italian Law because the pricing reflected the negative MTM of the Bear Stearns IRS;

(2)

The judge wrongly held that the Transactions involved the payment of an “upfront”, and thus “recourse to indebtedness” otherwise than “for the purpose of financing investment expenditure” within the meaning of Article 119(6);

(3)

Only if Grounds 1 and 2 fail, the judge wrongly held that the Italian law rules as to speculation and indebtedness (as explained in Cattolica) were properly characterised under English law as limits on Venice’s capacity;

(4)

Only if Grounds 1 to 3 fail, the judge's decision to apply the law as stated in Cattolica on 12 May 2020 to the Transactions which were entered into in 2007 was wrong because the development of Italian law in Cattolica was not reasonably foreseeable as at the date of the Transactions, and thus constituted a retrospective change of law;

(5)

Only if, contrary to the preceding Grounds, this Court upholds the judge’s conclusion that the Transactions were void, the judge wrongly held that, by reason of section 32(1)(c) of the Limitation Act 1980, Venice’s claims in restitution for payments made more than six years prior to issue of the claim form were not time barred.

58.

The judge granted permission to appeal on Grounds 3 and Ground 5 (in part). By order of Males LJ dated 27 April 2023, the Banks were granted permission to appeal on all Grounds except that permission to appeal on Ground 4 has, as noted at [2] above, not yet been granted.

59.

Venice’s appeal raises two Grounds of Appeal for both of which the judge granted permission:

(1)

The judge was wrong to find that Venice’s counterclaim for restitution of the net sums paid by it under the Transactions was governed by English law rather than Italian law;

(2)

The judge was wrong to find that there is a principled case for recognising a defence of change of position to the extent of any swap payments made by the Claimants under their back-to-back swaps with other banks. It is contended that the judge erred in two respects:

(a)

he allowed a defence of change of position in principle even though such a defence could have the effect of substantially enforcing contracts which were void pursuant to the ultra vires doctrine; and

(b)

he failed to give any weight to the fact that the back-to-back swaps were independent transactions over which Venice had no control and of which it had no knowledge.

Summary of parties’ submissions: Grounds 1 and 2 of the Banks’ appeal

Since I have concluded, for reasons set out below, that the Banks’ appeal succeeds on Grounds 1 and 2, their other Grounds do not arise and anything said in this judgment about those other Grounds will be strictly obiter. Likewise since the consequence of allowing the appeal on Grounds 1 and 2 is that the Transactions are valid and binding on Venice, its restitutionary claims and thus its Grounds of Appeal do not arise, so that anything said about those claims and Grounds in this judgments will be strictly obiter. In the circumstances, I propose to concentrate on the parties’ submissions on the Banks’ Grounds 1 and 2 and to deal relatively briefly with the submissions on all the other Grounds. Where counsel referred the Court to authorities, I will set out the relevant citations in this section of the judgment and avoid repetition in the Discussion section.

The principal submission of Ms Tolaney KC on Grounds 1 and 2 was that the judge wrongly concluded that the Transactions infringed Italian law because he misanalysed the Transactions and wrongly concluded that simply rolling over the negative MTM under the Bear Stearns swap transformed the swap with the Banks into a speculative transaction. She noted, as I have already recorded at [8] above, that Venice advanced no case that the Bear Stearns swap was speculative so that the appeal must proceed on the basis that, under Italian law, it was a valid and binding hedging transaction.

Ms Tolaney KC took the Court through the Cattolica decision. She asked the Court to note that the Supreme Court had not suggested that there was anything wrong with a negative MTM at the inception of a transaction or that the existence of such a negative MTM would render the transaction speculative. It expressly recognised at [4.7] of the judgment that the negative MTM represented the cost at which a party could unwind a derivative early or at which another bank would be willing to take over the derivative, which was, of course, the present case.

The Banks accepted for the purposes of this appeal that in section 8 of the judgment, the Supreme Court decided that local authorities could not enter into speculative derivatives, and although they were critical of the Court’s reasoning, they did not suggest that Cattolica was wrongly decided as a matter of Italian law. However, what Ms Tolaney KC described as a crucial point on the appeal is that the Supreme Court did not purport to state a test for what constitutes speculation or hedging.

Ms Tolaney KC drew attention to [9.8] where the Supreme Court said that a hedging derivative could only be entered by a local authority if the MTM criteria could be measured effectively, from which she submitted that it was clear that the Supreme Court was recognising that a hedging derivative can have a negative MTM at inception without becoming speculative.

She accepted that, to succeed on the appeal, the Banks needed to win on both Grounds 1 and 2, but the two were closely related. She dealt first in relation to those Grounds with the standard of review in relation to findings of foreign law. There was a considerable body of case law on this, but it has been considered in detail recently by the Privy Council in Perry v Lopag Trust Reg [2023] UKPC 16; [2023] 1 WLR 3494 (“Perry”) in the context of the Board’s practice regarding concurrent findings of fact. Of particular relevance was the passage from the judgment of Lord Hodge on behalf of the Board at [10] to [15]:

“…Findings of fact as to foreign law are in a special category in part because, in certain circumstances, in particular when the foreign law is a common law system analogous to the judges’ domestic law, the judge at first instance and the judges in the appellate courts can use their legal skills and experience in the analysis of domestic law to analyse the foreign law. In such circumstances the appellate judges are not at any significant disadvantage in carrying out that analysis compared with the trial judge. While the circumstances of cases may vary widely, the Board derives some propositions from the case law.

11.

First, the task of the trial judge when there are disputed questions of foreign law is to determine what the highest relevant court in the foreign legal system would decide if the point were to come to it: Dexia Crediop SpA v Comune di Prato [2017] EWCA Civ 428; [2017] 1 CLC 969 (“Dexia”), para 34; Morgan Grenfell & Co Ltd v SACE Istituto per I Servizi Assicurativi del Commercio [2001] EWCA Civ 1932 (“Morgan Grenfell”), para 50. It is not sufficient for a party to identify a judgment of a foreign court of first instance which may be on point and assert that the task of the appellate court is simply to analyse that judgment.

12.

Secondly, if the foreign legal system is a common law system which adopts a similar approach to legal reasoning and statutory interpretation to that of English law, the English judge at first instance is entitled and required to bring to bear his or her knowledge of the common law and the rules of statutory construction in analysing the foreign law. So too is the appellate court. In MCC Proceeds Inc v Bishopsgate Investment Trust plc [1999] CLC 417 (“MCC Proceeds Inc”), a case concerned with the construction of the Uniform Commercial Code which was part of the law of New York, a common law system, Evans LJ giving the judgment of the Court of Appeal stated (para 13):

“When and to the extent that the issue calls for the exercise of legal judgment, by reference to principles and legal concepts which are familiar to an English lawyer, then the [appellate] court is as well placed as the trial judge to form its own independent view.”

The important words in that statement are “to the extent” and the reference to familiar principles and legal concepts. The court went on to state that it was not entitled to substitute its own view for the view of the trial judge when there was acceptable evidence to support the judge’s finding unless the English court interprets the statute in accordance with English rules of construction and there is no evidence that different rules would govern the foreign court’s construction or evidence that the words would have a special meaning in a foreign context (para 20). The Court of Appeal in Dexia cited these passages in MCC Proceeds Inc with approval (paras 38 and 39). The Court of Appeal in Dexia also quoted from the judgment of the Court of Appeal in Morgan Grenfell paras 50 and 51, in which the court observed that where the court was faced with differing views as to Italian law, which was not based in any relevant aspect on the common law, there was less room for the judge to apply his or her own legal training and experience to help to resolve the relevant question.

13.

Thirdly, where the foreign law is in a foreign language the trial judge will often be dependent on translations of the relevant texts, which may or may not be precise and which may or may not be disputed, and on the evidence of the foreign law experts to understand the meaning and nuances of the foreign language in the relevant text. Thus, in Byers v Saudi National Bank [2022] EWCA Civ 43; [2022] 4 WLR 22 (“Byers”), the trial judge had to address questions of Islamic law, of which the only authorised texts were in Arabic, and he had to work with translations and with the assistance of foreign law experts. The Court of Appeal concluded that it should be slow to interfere with the judge’s findings of fact on Saudi Arabian law and should do so “only … in accordance with the principles applicable generally to findings of fact made by a trial judge who has based his findings on evidence from witnesses” (para 105). In reaching that view the Court had regard to the foreign language of the authorised texts, the fact that the concepts and principles of Saudi Arabian law were far removed from the common law, the lack of any familiarity of the English courts with the practice and culture in the capital markets of Saudi Arabia, and the fact that the judge at first instance had depended on the assistance of extensive expert evidence to explore and explain the many Saudi Arabian court decisions to which the experts referred in support of their contentions.

14.

Fourthly and more widely, where the first instance judge is dependent upon the evidence of foreign law experts, who disagree as to the interpretation and application of a foreign law, and has to decide issue by issue whose evidence to prefer, the judge will have regard to all the evidence presented to him. The judge will reach a view based on an assessment of each expert having regard to each expert’s evidence as a whole, and the way in which each expert answered the questions posed in chief and on cross-examination to justify his or her opinions. The judge will thus evaluate the experts’ reasoning. Not all the matters which have influenced the judge in forming a view on which evidence to prefer will always be recorded in any detail in a judgment or can be ascertained from reading a transcript of the proceedings. The judge will have regard to “the whole of the sea of evidence presented to him whereas an appellate court will only be island hopping”. Those words of Lewison LJ in FAGE UK Ltd v Chobani UK Ltd [2014] EWCA Civ 5; [2014] FSR 29, para 114, are in such circumstances as applicable to a case involving expert evidence on foreign law as they are to cases involving the evidence of witnesses of fact more generally. See the judgment of Longmore LJ in Dexia at para 42.

15.

There is thus a spectrum of circumstances in which the principal variable is the degree to which the judge can use his or her skill and experience of domestic law and of the domestic rules of statutory interpretation to ascertain the foreign law and apply it to the case in question. For example where a judge is an English lawyer, at one end of the spectrum there are cases in which the foreign law is a common law system which applies the same or analogous principles and means of legal analysis as English law. In such cases there will be considerable scope for the trial judge to bring to bear his or her legal skills and experience in domestic law in determining and applying the foreign law. The judges of a court hearing the first appeal will also be able to bring to bear their own skill and experience. In such a circumstance the members of the Board also would be able to do so and would be unlikely to invoke the practice relating to concurrent findings of fact. At the other end of the spectrum are cases of disputed foreign law in which the skill and experience of the judge in domestic law has a minimal role to play in the ascertainment and application of foreign law, as in Byers. In such cases the court at each level of the hierarchy is dependent on the written and oral evidence of expert witnesses, tested by cross-examination. The trial judge’s findings on the content and application of foreign law have a close kinship to other findings of fact. In that circumstance the first appellate court will be slow to intervene in the judge’s assessment and the Board’s practice in relation to concurrent findings of fact should be adopted.”

Ms Tolaney KC made two particular points. The first was that where a case lies on the spectrum which Lord Hodge refers to does not depend on a bright line distinction between common law and civil law. What matters is the content of the foreign law and whether the judge can bring his or her own legal judgment to bear on the question that has to be resolved. She pointed to Byers as a case at one extreme of the spectrum where the concepts and principles of Saudi law were far removed from English law. However, if a civil law jurisdiction had the same rule as under English law, then the judge would be bound to bring his or her own legal judgment to bear and the appellate court would be in the same position.

The other point, which also emerged from [15] of Perry, was that there may be more room for the exercise of legal judgment when it comes to applying the foreign law to the facts. Ms Tolaney KC referred in that context to [51] and [52] of the judgment of the Court of Appeal in Morgan Grenfell (cited by Lord Hodge at [11]:

“51…However, in approaching the expert evidence of Italian law, it was in our view appropriate for the judge to have at least some regard to his own experience and training in so far as it was relevant to the particular issues which he was considering.

52…the judge would have to decide, in the light of the principles of Italian law which he found to exist, what conclusions the Corte di Cassazione [the Supreme Court] would have reached on the key questions in the case. In carrying out that exercise the judge would apply the principles of Italian law to the facts as he found them, which would involve essentially the same exercise as is performed by the judge in every case. To that extent at least he would have to apply his own legal training and experience.”

Ms Tolaney KC submitted that there was no doubt that in the present case, the judge brought his own legal judgment to bear, not least because Cattolica did not itself lay down any test as to what constitutes speculation and he was invited by Venice to approach the matter as a question of both Italian and English law, which he did.

In relation to Ground 1, Ms Tolaney KC took the Court through how the Transaction worked. She referred to the Bear Stearns swap and the amendments to it, which I have summarised at [5] above. She noted that, in 2007, Venice wanted to restructure the Rialto Bond by extending its maturity to 2037 and to free up some short-term cash. The rate payable under the Amended Rialto Bond was six-month Euribor plus 0.21% so Venice continued to be exposed to a variable rate risk, albeit over a longer period of time. By this time the rates had moved, with the consequence that there was a substantial negative MTM under the Bear Stearns swap. Since the appeal must proceed on the basis that the Bear Stearns swap was hedging, not speculative, it was a valid and binding obligation to which Venice was already exposed which needed to be addressed during the restructuring. Since Bear Stearns was unwilling to restructure its swap, Venice needed to find another bank or banks to take it over and then restructure it, given that, as the judge found at [83] of the judgment, Venice was not willing to part with cash to cancel the Bear Stearns swap, because of budgetary constraints.

Ms Tolaney KC emphasised that the Banks took over the Bear Stearns swap by way of novations, so they stood in the shoes of Bear Stearns. The payments they made as the price for the novations were made to Bear Stearns not to Venice, nor could they properly be described as payments on behalf of Venice. The Bear Stearns swap was not closed out but continued to exist, albeit with the notional amount reduced to zero. Ms Tolaney KC pointed out that the judgment does not grapple with the fact that if the Transactions are null and void as the judge found, this is an extraordinary windfall to Venice when it has never challenged the Bear Stearns swap itself.

She also emphasised that, under the terms of the Transactions which restructured the Bear Stearns swap, Venice would receive exactly the same amount of interest as payable by it under the Amended Rialto Bond. The notional amount and the amortisation schedules were also the same. As she pointed out each Confirmation expressly stated:

“The Municipality of Venice and [Banca Intesa/Dexia] mutually acknowledge that this Interest Rate Swap transaction, executed under the abovementioned financial conditions, takes into account the Mark to Market of the Interest Rate Swap agreement originally entered into the Municipality and Bear Stearns on 19/12/2005 and transferred on this day to [the Bank] for a quota equal to [68%/32%] (the Old Confirmation is attached to the Novation Agreement entered into on this date). This Mark to Market, equal to [€ 5,484,200/€2,580,800] at the moment of the transaction, will be reflected in this swap entered into in respect of the restructured bond.”

Ms Tolaney KC submitted that this meant that the Transactions had been priced on the basis of current expectations that the Banks would receive the same amount over the lifetime of the swaps as Bear Stearns was expected to receive under its swap.

She made six points on the Transactions. First, that in legal and commercial terms, Venice did not borrow any money under the swaps but the Banks paid the novation fees to Bear Stearns as the price of stepping into its shoes. Second, that the pricing of the Transactions by reference to the negative MTM under the Bear Stearns swap did not bring into being any new risk for Venice but matched the risk it was already exposed to. Third, that whilst it was true that if rates moved in the future, Venice’s contingent liability might go up or down, that did not make the Transactions speculative, since that is true of every swap. In Cattolica the Supreme Court recognised that risk but did not suggest that in itself made the swap speculative. Fourth, the negative MTM under the Bear Stearns swap meant that the Transactions were less advantageous than if Venice had started with a clean slate, but it did not start with a clean slate because it was already exposed to paying the floor rate under the Bear Stearns swap and could not wish away that risk. Fifth, it is instructive to consider what would have happened if Bear Stearns had agreed to restructure its swap. It would have said we will restructure the swap and match the Rialto Bond maturity date of 2037 and roll over the MTM, so precisely what happened would have happened, but that would not make it speculative or borrowing. Sixth, any change to the swap would require the negative MTM to be addressed in some way, because no sensible commercial counterparty is simply going to give up that which is in its favour. The Supreme Court in Cattolica recognised at [4.6]-[4.7] that the negative MTM could be addressed in that way, which it appeared to consider acceptable, without rendering the swap speculative.

Ms Tolaney KC submitted that in dealing with this issue, [230] of the judgment highlights two fundamental errors of principle. First it mischaracterises the legal and commercial effects of the Transactions, because the payment of the novation fees by the Banks to Bear Stearns cannot be described as borrowing by Venice. Second, this paragraph of the judgment overlooks that the Bear Stearns swap was not speculative so the negative MTM was a non-speculative risk to which Venice was already exposed. Novation of the swap from one bank to another under the ISDA terms of novation could not transform it into a speculative transaction.

She submitted that the judge had begun this section of his judgment at [222] by recognising that the Supreme Court in Cattolica had not attempted a definition of what makes a derivative speculative as a matter of Italian law, but saying that there were a number of indicia in the Italian case law which may have the effect of a derivative being speculative or hedging, which Mr Cox KC submitted reflected the English case law on the same topic. This was an error because the question for the Court was what the Italian Supreme Court would have decided, not a whole series of lower courts, let alone English decisions. The judge had fallen into this error because this is what Venice had suggested should be his approach in its written closing submissions at [24].

Ms Tolaney KC submitted that (i) to (v) of [225] setting out the essential facts relating to Venice’s decision to enter the Transactions demonstrated that Venice had to enter into a swap with another bank or banks that rolled up the cost of the Bear Stearns swap and that the restructuring was beneficial to Venice because it obtained protection against a significant rise in the rate under the Rialto Bond over its extended term, all of which is not suggestive of speculation but a classic description of hedging. She submitted that there was no suggestion that renegotiating the Bear Stearns swap would have made that swap speculative, so how could renegotiating with other banks make speculative what was previously not.

She submitted that [225(vi)] is where the judge misanalysed the swap. He analysed the rolling over of the negative MTM as if it were a discrete and separate part of the transaction having a free-standing function, whereas the swap was all one transaction. At [226] the judge sets out the impact of structuring the swap to cover the costs of winding up the Bear Stearns swap, but this was not new borrowing or a new risk. Ms Tolaney KC pointed out that it was the unchallenged evidence of the Banks’ expert Mr Malik that the risk of Venice losing money under the existing Bear Stearns swap after the restructuring of the Rialto Bond was greater than under the Transactions.

She noted that at [228] the judge accepted a high degree of correlation between the terms of the Transactions and the Amended Rialto Bond as regards the Notional Amounts, the amortisation rate, the maturity date and the interest rate received by Venice from the Banks. However he was troubled by financially significant differences, the level of the floor and cap, which overlooks that that was always going to be the case, even if this was hedging, because it was essentially the price for the Banks providing the restructuring. If the absence of correlation in relation to the floor and the cap was significant, this would be the case with all swaps, even plain vanilla ones.

Ms Tolaney KC was particularly critical of [229] of the judgment, where the judge determined that the Transactions were speculative by referring to a series of Italian and English court decisions, not just those of the highest Court and demonstrated that he was exercising his own legal judgment. She pointed out that, at [229(i)] he concluded correctly that limb (a) of the CONSOB Determination was satisfied but at (ii) he erroneously concluded that limb (b) was not despite the high degree of correlation between the Transactions and the Rialto Bond. She also pointed out that, on the basis of the expert evidence of the Banks’ expert Professor Gentili, compliance with the CONSOB Determination was determinative of the Transactions being hedging, referring to the acceptance of the CONSOB test in Supreme Court decision No. 19013/2017. Venice’s expert Professor Alibrandi did not agree that the CONSOB Determination was exhaustive, but accepted in cross-examination that, if the Transactions satisfied the first two limbs of the CONSOB Determination, they were likely to be hedging. Ms Tolaney KC submitted that the judge should have concluded that the CONSOB Determination was an exhaustive test and that, since both limbs were satisfied, the Transactions were hedging, not speculative. On that narrow ground alone the judgment could not stand.

In [229(iii)] the judge emphasised the very significant difference between the MTM of the cap and the floor but, as Ms Tolaney KC submitted, they will never be equal since the Banks will be looking to cover their costs and make a profit. It is clear that, as a matter of Italian law, there is nothing objectionable about that, because, if there were, the Banks could never enter a swap, since they could only do so by pricing it at a loss. This emerged from Cattolica but also from a decision of the Council of State, the highest Italian administrative court, in case 5962/2012, from which it is clear that the cap and the floor do not need to match. The judge did not cite either, but only a decision of the Court of Appeal of Milan and the English case of Standard Chartered Bank.

She submitted that in (iv), the judge made the same error of principle as previously of treating the novation of the Bear Stearns swap as a separate event distinct from the other aspects of the restructuring. He also suggested that the rolling over of the negative MTM under the Bear Stearns swap was itself a reason why the Transactions were speculative, relying on two Italian first instance decisions, one of which was an interlocutory decision. At (vi), the judge says that Venice took on a significant new risk to which it was not exposed under the Rialto Bond which was suggestive of speculation. Ms Tolaney KC again submitted that there was no significant new risk but only the rolling over of the risk to which Venice was already exposed.

She submitted that the judge’s conclusion on the basis of lower court decisions, let alone English cases, was asking the wrong question. The right question was how the highest Italian court would apply Italian law to the question whether these Transactions were speculative or hedging. Instead the judge went through and relied upon decisions of the lower courts, some of which were inconsistent with the decisions of the highest Court. In doing so, he did not rely upon expert evidence of Italian law about those decisions. The English cases to which he referred were wholly irrelevant. She submitted that, even if she was wrong that the satisfaction of the two limb CONSOB Determination test was conclusive, the decisions of the Supreme Court and the Council of State do not consider the elements on which the judge relied as definitively speculative.

Ms Tolaney KC submitted that the decision of the Council of State in case 5962/2012 proceeded on the basis of the economic convenience of the swap and the underlying debt looked at together. She sought to draw three points from it: (i) that the highest Italian courts look at the underlying debt and the swap together to assess the overall effectiveness of a restructuring; (ii) as in Cattolica, the decision makes it clear that there is nothing objectionable under Italian law about a negative MTM from inception; and (iii), linked to (ii), there is no reason why the value of the cap should be equal to the value of the floor and an imbalance between the two did not render the swap speculative or invalid. In relation to the second point, she also relied upon the Supreme Court Decision No 21830/2021 which referred at [2.61] to a negative MTM as a completely lawful profit margin.

She submitted that there was no rational basis for concluding that the highest Italian court would hold that merely rolling forward a negative MTM amounts to speculation, so [229(iii)] is not sustainable and, to the extent that the judge attributed weight to inconsistent first instance authorities, he lost sight of the right question. In any event in many of those authorities, the swap was found to be speculative because of a significant disparity between the underlying debt and the swap, which was not the case here.

Ms Tolaney KC submitted that Ground 2 of the Banks’ appeal gave rise to two questions: were the novation fees properly to be regarded as “upfront” payments and, if so, is an “upfront” payment made in connection with a hedging swap impermissible. It was agreed between the Italian law experts that an upfront payment in the context of derivatives is an amount of money paid by one party to another to rebalance the financial situation of the parties in “non-par swaps” i.e. swaps whose value at inception is not equal to zero. Ms Tolaney KC’s short point was that there was no payment by the Banks to Venice to rebalance the transaction. The novation fees were paid to Bear Stearns and, in any event, the negative MTM was rolled over, not balanced out.

In concluding at [261] that the fact that the money moved from the Banks to Bear Stearns rather than through Venice made no difference to his conclusion that it was an upfront payment, the judge fell into error. At (iii) the judge concluded that Professor Gentili had accepted this in cross-examination, but that misunderstood his evidence. What Professor Gentili accepted was that if Venice was entitled to receive a payment but requested the Banks to pay it to Bear Stearns, so there was a delegated payment, that would be an upfront payment. However that was not this case, because Venice was never entitled to the novation fees and never requested the Banks to pay them to Bear Stearns.

Even if the novation fees were upfront payments, because the Transactions were a hedge and not speculative, their payment did not infringe Article 119(6) of the Constitution because the purpose of the Transactions was to manage the risk arising from the underlying debt and did not increase that risk.

Mr Cox KC began his submissions on Grounds 1 and 2 with five general points. First, that what was wrong with the Transactions as a matter of Italian law was that they exposed Venice to paying interest at the floor level, potentially and in fact, far above the interest due under the Rialto Bond. Second, for the purpose of Italian law, one looks at the Bond and the swap, not at any previous swap. If the previous swap were closed out and were speculative and the new swap were not speculative, it would not be controlled under Article 119(6). Third, whether a negative MTM made a swap speculative depended on the circumstances. In this case the negative MTM with roll-over was very significant. Fourth, although Venice thought it had capacity to enter the Transactions, the effect of Cattolica is that it did not, all along. Fifth, he submitted that it was wrong to treat the Bear Stearns swap as hedging or compliant. The judge had made no finding about that. For reasons I have already foreshadowed at [8] above, that fifth point is wrong and demonstrated a fundamental fallacy in Venice’s argument. This appeal has proceeded on the basis that the Bear Stearns swap was valid and was hedging.

He then made submissions about the standard of review by an appellate court of a judge’s findings of fact, including of foreign law, starting with the well-known statement of principle by Lewison LJ in FAGE UK Ltd v Chobani UK Ltd [2014] EWCA Civ 5; [2014] FSR 29 at [114]-[116]. He also cited Lewison LJ’s judgment in Volpi v Volpi [2022] EWCA Civ 464; [2022] 4 WLR 48 where at [2] he set out what he said were well-settled principles as to the restraint which an appellate court should exercise before it interferes with a judge’s findings of fact. At [4] Lewison LJ said, citing Byers, that similar caution applies to appeals against a trial judge’s evaluation of expert evidence. Mr Cox KC went on to cite the judgment of the Privy Council in the recent case of Perry at [12] cited at [65] above. He submitted that in almost every respect, the present case was at the Byers end of the spectrum.

He submitted, in relation to Italian law, that one looked at the swap itself to consider whether it is speculative and did not ask the question whether the swap buyer was better or worse off from having closed out one swap and entered another. In relation to correlation between the Transactions and the Rialto Bond, there was no correlation on one key point, the floor in the Transactions. Venice would potentially have to pay the floor rate for a very long time if interest rates fell below the floor, a crucial reason why the Transactions were speculative. The forward rate curve was below the floor when it was sold so that it was known from the outset to be loss-making.

Mr Cox KC took the Court to the Supreme Court decision in Cattolica emphasising that at [8.3] the Court had determined that: “the distinction between hedging and speculative derivatives, based on the criterion of the different degree of risk of each of them, meant that only in the first case could a local entity be said to have capacity to enter into them.” He also referred to [4.6] on which Ms Tolaney KC had relied and submitted that if the negative MTM being postulated means that a swap is not hedging an underlying risk, it is speculative.

He also took the Court to the Council of State decision 5962/2012. In that case loans were restructured into a new bond issue which was then protected by a collar derivative, so it was not a case of a closed-out swap. The conversion of the loans into a bond was found to be economically advantageous and the cost of the swap was well within the amount of that advantage so under Article 41 the whole deal was economically advantageous within the parameters allowed. The case was simply not about whether the swap was speculative or how you would assess whether it was or whether the negative MTM rendered it speculative.

When Mr Cox KC turned to address how the judge had dealt with the circumstances in which a derivative was speculative as a matter of Italian law, he accepted, in answer to questions from the Court, that the Italian law experts had confined themselves to the question of the test for speculation under Italian law and had not addressed the question whether, as a matter of Italian law, the Transactions themselves were speculative.

Like Ms Tolaney KC, Mr Cox KC focused on some of the Italian cases to which the judge referred in [212] of his judgment. He took the Court to the decision of the Supreme Court No. 10913/2017. As the Court pointed out, and Mr Cox KC accepted, that case refers to the CONSOB Determination as the test for whether or not a derivative is hedging. As we also pointed out, that was a case of renegotiation of a swap, but both the original swap and the replacement swap were found by the Court at [8] not to have been hedging, in contrast to the present case, where the original Bear Stearns swap was hedging and is not being impugned.

Mr Cox KC placed reliance on the decision of the Court of Appeal of Milan No. 2393/2020, referred to by the judge in [212(iv)], which post-dated the Supreme Court decision in Cattolica. He submitted that was a case where there was a large mismatch between the value of the cap and the floor and it was held that the swap was not hedging but speculative, in accordance with the general rationale in Cattolica. It was no answer to say, as Ms Tolaney KC had in relation to that case, that it was not a case where the imbalance between the cap and the floor arose from the exit cost of a previous non-speculative swap. Mr Cox KC submitted that it did not matter to the analysis that the cause of the imbalance was a close-out or something else. In reply, Ms Tolaney KC submitted that the Milan decision was inconsistent with the two decisions of the Supreme Court and the Council of State decision and should not have been relied upon by the judge.

Mr Cox KC also relied upon the decision of the Court of Orvieto in 2012, referred to in [212(v)] of the judgment as another case of rolling over existing swaps into new swaps which absorbed the negative MTMs from the existing swaps. He referred to a passage in the judgment which said that the majority of the swaps under review stemmed from a renegotiation which represented a deviation from the ordinary course of the operations with derivatives that a local authority is allowed to carry out. However, as we pointed out in argument, that was not a renegotiation of swaps in the context of debt restructuring, but the local authority doing it for its own purposes.

He also relied upon the decision of the Court of Turin No. 4685/2021, referred to in [212(vi)], as another case where a rollover from an earlier swap involved a negative MTM, which was an indication of the renegotiated swap being speculative. Finally he referred to the Supreme Court Decision No 21830/2021 at [212(vii)]. He submitted that this case of a vanilla swap gave assistance on the indicia of hedging: correlation between the swap and the underlying debt and alignment between the terms of the swap and the forward curve for interest rates.

Mr Cox KC then made submissions about the section of the judgment from [226] to [232] (which I have cited at [34] above) dealing with whether the Transactions were speculative. He submitted that the judge’s approach at [227] was the right one in line with Cattolica. At [228(iii)] the judge made the clear comparison between the underlying Rialto Bond and the swap. As he said at [229], the disparity between the cap and the floor was not matched in the Rialto Bond and meant that there was a big imbalance, a new and significant risk to which Venice was not exposed under the Rialto Bond. Overall on these paragraphs, Mr Cox KC submitted that this Court should accept that the judge was correct that the right approach to whether the Transactions were speculative or not depended on the comparison with the underlying Rialto Bond and his treatment of the question could not be criticised.

In relation to [230] of the judgment, which was much criticised by the Banks, Mr Cox KC submitted that the effect of rolling over the negative MTM was that Venice did not have to pay Bear Stearns for the close-out of its swap, but it did have to pay for it through the terms of the Transactions. That situation was aptly described as akin to borrowing. In any event, by this point in the judgment, it is obvious that this paragraph is not decisive of the issue whether the Transactions were speculative. The judge is simply identifying another way of characterising the transaction.

In relation to Ground 2, Mr Cox KC submitted that the issue was a narrow one, whether, as a matter of Italian law and on the facts, there were upfront payments because of the way the swaps were structured. Again this Court could only interfere if the judge was plainly wrong. He referred to [10.1.3] of Cattolica which found upfront payments by swap providers to swap buyers were indebtedness for the purposes of Article 119(6). Because the payment is expected to be repaid through the terms of the swap, it is to be treated as a loan. He referred to [261] of the judgment and the judge’s point that, merely because the upfront payment might not have to be paid back if the market moved in a particular manner, that did not stop it being indebtedness because of the risk that Venice might end up having to pay it back. He submitted that the judge had been right that it made no difference to the analysis of it being an upfront payment, that the payment was made direct by the Banks to the previous swap provider, Bear Stearns, since it had the same economic effect. Venice benefited from the payment since the Bear Stearns swap was discharged, but was expected to pay it back through the terms of the swap.

In her reply submissions on Ground 1, Ms Tolaney KC emphasised that in determining whether the Transactions were speculative, the judge had sought to impose his own test, applying an English law approach to the analysis of various Italian cases without analysing critically their inconsistencies. Worse still, he had then sought to bolster his conclusions by reference to English cases, some of which were not on point and none of which was relevant to the exercise he should have been undertaking. She adopted what I had put to Mr Cox KC in argument, that the judge made his own evaluation on the issue of speculation.

In answer to the question which Males LJ had put to Mr Cox KC during his submissions, as to whether the Italian law experts had confronted the critical question on this appeal, whether the comparison was only between the risks under the new swap and the underlying Rialto Bond, with the pre-existing liabilities under the Bear Stearns swap being irrelevant, Ms Tolaney KC said they had not. That had not been Venice’s case, nor did the experts say that, as a matter of Italian law, you do not bring into account the position under the previous closed-out swap. In any event, the function of the expert evidence was limited to expressing an opinion on the test for speculation under Italian law, whether CONSOB imposed the relevant test or whether it was a broader test. The experts were not asked to opine on how that test applied to the facts, which was for the judge, as Venice accepted in its written closing submissions at trial. Ms Tolaney KC referred the Court to [186] of those submissions where it is said that: “It is therefore understood that there is now no serious dispute that the Speculation Question [how one should determine whether a derivative is speculative] is a question of pure fact rather than a question of Italian law”.

In reply, Ms Tolaney KC also made the point about the Supreme Court Decision No 21830/2021, that, although the Court had said that alignment with the forward curve was evidence of an effective hedging strategy, it had not said that the absence of alignment meant that the swap was not a hedge. The Court expressly acknowledged that things can change, so that a swap falls out of alignment.

In relation to Ground 2, Ms Tolaney KC submitted in reply that Mr Cox KC had not engaged at all with her point that Grounds 1 and 2 stand or fall together. The judge’s reasoning in [268] as to why the Transactions were not undertaken for the purpose of financing investment expenditure within Article 119(6) was infected by the same errors as he had made in relation to Ground 1. Furthermore, Venice’s own expert, Professor Domenichelli, had accepted in cross-examination that a hedging swap did not infringe Article 119(6). In relation to the judge’s assessment that the payment by the Banks of the novation fees had the same economic effect as an upfront payment, Ms Tolaney KC submitted that this was another example of the judge forming his own view, exercising his own legal judgment, not relying on expert evidence. In fact, she submitted that, in cross-examination, Venice’s expert Professor Alibrandi had admitted that “from a formal perspective” this was not an upfront payment.

As I have said, given my conclusion that the appeal succeeds on Grounds 1 and 2, I can deal with Ms Tolaney KC’s submissions on the other Grounds relatively shortly. On Ground 3 she submitted that the judge should have recognised that the rules identified in Cattolica were mandatory rules of Italian law, not restrictions on Venice’s capacity. As such, he should have concluded that since the rules were ones which prohibited Venice from exercising its capacity in a particular way, rather than restricting its capacity, as a matter of English law they were inapplicable and he should have concluded that Venice had capacity to enter the Transactions. Ms Tolaney KC referred to the distinction between the concepts of ultra vires and illegality drawn by Lord Leggatt in Rampersad, to which I referred at [15] above.

She submitted that the judge fell into error in concluding at [271] that Article 119(6) conferred the power to have recourse to indebtedness and then restricted it, as the Norwegian Act had in Haugesund. She asked the Court to note that at [273], the judge recorded that Venice had not taken issue with the statements of the Banks’ expert that Article 119(6) and Cattolica did not restrict the capacity of local authorities to contract as a matter of Italian law. She submitted that Article 119(6) was not a rule conferring the power of local authorities to borrow or, specifically, to enter swaps. It is a rule prohibiting local authorities from entering certain types of swap for an illegitimate purpose. On that basis, the judge should have concluded that Venice did have capacity to enter the Transactions and that Haugesund was distinguishable.

In relation to Ground 4, the Banks needed permission to appeal. It had not been argued before the judge, but it was a pure point of law and the suggestion by Venice that further evidence would be needed to deal with it was wrong. On the judge’s findings, Cattolica was unforeseeable (which was Venice’s own case) and the decision had retrospective effect. She submitted that, as a matter of English conflicts of laws, the capacity of a foreign legal entity is to be determined by reference to the law when the entity purported to exercise its powers and the English Court would not apply an unforeseeable judicial decision purporting to deprive the entity of capacity retrospectively. In the alternative, to give effect to Cattolica would be contrary to English public policy because it is inconsistent with A1P1 of the ECHR.

On the first point, she submitted that Venice clearly had capacity to enter the Transactions in 2007. To allow the Transactions to be declared null and void 13 years after the event by an unforeseeable Italian judicial decision would be both remarkable and monstrous for this jurisdiction, since it would undermine the sanctity of English law in important ISDA international agreements and since it would leave English law contracts subject to the shifting sands of changing Italian decisions, for example if Cattolica was not the last word and Italian courts were later to change their minds (Italian law does not have a system of binding precedent). Ms Tolaney KC submitted that the conclusion that English law would not recognise a retrospective foreign court decision in these circumstances was supported by a number of authorities.

She referred first to the decision of the Court of Probate (Lord Penzance) in Lynch. The president of Paraguay had died leaving property in England and the provisional government then passed a retrospective decree invalidating his will. The Court held that the law of Paraguay, as the law of his domicile, had to be applied as it existed at the time of his death and the Court would not give effect to retrospective changes in the law made thereafter. Although that was a case about succession, Ms Tolaney KC submitted that there was an important similarity with capacity as both concepts concern the exercise of a power at a specific point in time.

She then referred to the decision of the House of Lords in Adams, the facts of which I summarised at [16] above. She went through the speeches. Viscount Simonds considered that the bondholders’ rights were to be determined under the 1953 law, under which the bank had assumed obligations to them, and not the subsequent 1956 law. He was unimpressed by the reliance on the 1956 law as being one affecting the status of the bank. He relied upon Lynch which he regarded as correctly decided. Lord Reid considered that the 1956 law was one which sought to discharge liabilities on the bank to pay the bondholders in England and English law would not regard it as effective. Nor would English courts recognise a piece of foreign legislation which purported to have retrospective effect (see pp 282-3). He considered that between the creation of the new bank in 1953 and the 1956 law, there were accrued contractual rights and obligations which the foreign legislature could not remove retrospectively.

Lord Radcliffe also considered that once the new bank was liable under the guarantee provided by its predecessors, that liability could not be removed by the subsequent decree of the Greek legislature. He also approved Lord Penzance’s judgment in Lynch. Lord Tucker also thought the principle in Lynch should be applied and that an English court should not give effect to foreign legislation which purported to discharge the bank’s liability under negotiable instruments governed by English law. Only Lord Denning took a different approach, concluding that it was contrary to English public policy for the 1956 law to transfer assets to the new entity but exclude liabilities, so that the Court should not apply the 1956 law as it was contrary to English public policy. He cited Lynch without disapproval.

Ms Tolaney KC submitted that the principles derived from Lynch and Adams should apply in this case. She submitted that, if the Italian legislature had passed a law in 2020 which provided that Italian local authorities did not have capacity to enter into swaps and purported to have retrospective effect back to 2001, then, on the basis of those cases, the English Court would not recognise that piece of legislation but would conclude that Venice had capacity to enter the Transactions as the law stood in 2007. If, as the judge found, English law would not apply an Italian law which provided that the Transactions were void, why should the position be any different if the Italian legislature retrospectively removed the local authorities’ capacity, which would have precisely the same effect? She submitted that if that was the position with regard to retrospective foreign legislation, the position with regard to a foreign judicial decision should be a fortiori because such a decision could be changed and would be less certain than legislation, especially since Cattolica was a decision in a civil law system which did not have the doctrine of precedent.

She submitted that it would be contrary to section 6 of the Human Rights Act to apply a foreign law which contravened the Banks’ Convention rights. Those rights under A1P1 were engaged here. She relied on Solaria Nergy v DBEIS [2020] EWCA Civ 1625; [2021] 1 WLR 2349 at [34] per Coulson LJ. She submitted that Venice’s argument that the Banks never had any possessions because the swaps were void under Cattolica cannot be right. The correct analysis must be to ask whether they acquired possessions under the law as it stood when they entered the Transactions. To deprive the Banks of those rights would be contrary to A1P1.

The final ground of the Bank’s appeal which concerned limitation would only arise if the Court were against the Banks on its first three grounds and dismissed the first ground of Venice’s appeal as regards the applicable law of its restitution claim. On the basis that that claim was governed by English law, Ms Tolaney KC submitted that the judge should have concluded that the claim was time barred, if he had applied the principles laid down by the Supreme Court in FII correctly. In that case, the Supreme Court departed from the earlier decision of the House of Lords in Deutsche Morgan Grenfell Group plc v Inland Revenue Comrs [2007] 1 AC 558 and held that, for the purposes of section 32(1)(c) of the Limitation Act 1980, time runs from the point in time when the claimant discovered, or could with reasonable diligence have discovered, his mistake, in the sense of recognising that a worthwhile claim had arisen or that he had been mistaken with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking advice and collecting evidence. The Supreme Court rejected the approach of the majority of the House of Lords in Deutsche Morgan Grenfell that the limitation period does not begin until the truth has been established by a final judicial decision. Lords Reed PSC and Hodge DPSC commented at [177] of their judgment that that approach; “does not merely extend the limitation period to the extent necessary to overcome the disadvantage arising from the mistake, but has the remarkable consequence of excusing the claimant from the necessity of bringing a claim until he can be certain that it will succeed: indeed, until it has in fact succeeded.”

On the facts, Ms Tolaney KC pointed out that, as set out at [9] above, in October 2008, Venice was consulted by the President of the VIII Commission of the City Council on whether certain swaps could be cancelled on the basis that they were speculative. In March 2014, the Court of Appeal of Bologna allowed the local authority’s appeal in Cattolica and held the swaps in that case to be null and void. She also pointed out that in Prato, where the local authority raised arguments about both indebtedness and speculation, as a consequence of which the swaps there were null and void for lack of capacity, the proceedings in the Commercial Court were commenced in 2010. Although in his judgment in 2015, Walker J held that the decision of the Court of Appeal of Bologna in Cattolica did not represent Italian law, Ms Tolaney KC submitted that what mattered for the purposes of section 32(1)(c) is that the issue of capacity was in play and local authorities were issuing claims on that basis. If Prato could raise the point, so could Venice have done. It did not need to know that its claim would succeed, only that there was an ability to plead a proper case, which there clearly was long before the decision of the Supreme Court in Cattolica. Since Venice had hitched its wagon on limitation entirely to the date of the Supreme Court decision in Cattolica, that ought to have been enough to dispose of the defence under section 32(1)(c) as fundamentally flawed following FII.

She submitted that there were three errors in the judge’s reasoning at [430] (cited at [55] above) that the claim was not time barred. First, as is apparent from (iii) of that paragraph, he focused wrongly on the likely outcome of the claim in England when he should have been asking whether Venice could plead a proper case. Second, in that sub-paragraph, he focused on the likely position in England after Prato but that was the wrong time at which to assess discoverability. Time began to run at the latest from the moment that a person would have known that a serious legal challenge had been made, which was long before the Court of Appeal decision in Prato. Third, at (i) and (ii) the judge attached significance to the fact that England was the contractual forum for the resolution of the dispute, but that made no difference to the application of section 32(1)(c). Venice’s case was always going to be based on Italian law. Ms Tolaney KC submitted that the judge’s reasoning did not do justice to the decision in FII and went badly wrong.

In relation to Ground 3 of the Banks’ appeal, whether the judge was wrong to conclude that the principle established by Cattolica that local authorities cannot enter speculative swaps went to capacity rather than being a prohibition, Mr Cox KC submitted that the judge had been correct to interpret Cattolica as deciding that Article 119 conferred a power on local authorities to enter swaps provided that they were not speculative. The judge had correctly rejected the Banks’ argument that the effect of the Supreme Court decision was that there was merely a prohibition on entering speculative swaps as opposed to Venice lacking the substantive power to do so.

Mr Cox KC referred to the conclusion of Tomlinson J at first instance in Haugesund, upheld by the majority of the Court of Appeal, that the relevant Norwegian law being considered there restricted the capacity of the local authority to enter into the swap. It gave a power, but limited it against the backdrop of a more general power to enter into contracts. That was very similar to the position here. It was common ground that Venice had general authority to enter contracts, but Cattolica decided that Article 119 gave a power to enter swaps provided that they were not speculative. Having decided that that was the position as a matter of Italian law, the question for the judge was whether, as a matter of English law, the Italian law went to capacity and he correctly decided it did.

In relation to Ground 4, Mr Cox KC submitted that the Court should not give permission to appeal because it would necessitate new evidence, applying Times Travel v Pakistan International Airlines [2022] EWCA Civ 415. When the Court pressed him on what new evidence Venice would wish to call on this issue, he identified evidence of the effect of Cattolica within the Italian legal and judicial system, but, as I pointed out, how Italian law treated the decision was not really relevant to the issue raised by Ground 4, which is whether an English court should treat the Supreme Court decision as having retrospective effect.

On the merits of the Ground, Mr Cox KC referred the Court to the principles set out in Dicey, Morris & Collins on the Conflict of Laws (16^th edition) on Changes in the Lex Causae from [2-076] onwards, running through the authorities and culminating in the statement at [2-090]:

“The prevailing practice of the English courts thus seems to be to apply the lex causae as it exists from time to time and to give effect if need be to retrospective changes therein.”

In relation to the cases on which the Banks relied, he submitted that neither Lynch nor Adams involved judicial decisions. They involved legislative interference by a foreign legislature. There was no qualification in them based on reasonable foresight. They did not involve capacity to enter a contract, although as I pointed out, Adams did involve discharge of a contract by subsequent changes in the foreign law, so that there was an obvious analogy with the present case.

On A1P1, Mr Cox KC submitted that the European Court of Human Rights had recognised that there was nothing wrong in principle with retrospective law within certain parameters. He referred to Zielinski v France (2001) 31 EHHR 19 and National & Provincial Building Society v United Kingdom 25 EHRR 127.

In relation to Ground 5, Mr Cox KC referred to various passages in the majority judgment in FII and subsequent cases and emphasised that the crucial question was whether Venice could with reasonable diligence have discovered its mistake. The standard to be applied is how a person in the business of Venice would act, on the assumption that it desired to know if it had made a mistake. When could Venice have discovered with reasonable diligence that the basis on which it had paid under the Transactions was questionable so as to give rise to a worthwhile claim?

Looking at [430] of the judgment, Mr Cox KC submitted that the judge made no error in having regard to the fact that Venice’s claim to restitution would have had to be brought in the UK. In relation to (iii) and Prato he made the point that the claim failed before Walker J and the Court of Appeal so that it was a bad claim. That did not make a claim by Venice at that time a worthwhile claim. In relation to indebtedness, Prato had decided that the legal definition of indebtedness in Article 3 of Law 350/2003 was exhaustive and did not include swaps, which is what Venice would have found if it had investigated. On speculation, that also was not accepted in Prato and, as the judge found in this case, Cattolica represented a significant discontinuity with what had been understood previously. Without that Supreme Court decision, there would have been no claim for lack of capacity.

Mr Cox KC accepted that the letter of 31 October 2008 from the President of the VIII Commission of the City Council was a “trigger” which asked how Venice was dealing with speculative swaps. However, it provided no grounds which would sustain a reasonable belief that there was a claim. It simply asked the question.

As with the Banks’ appeal on Grounds 3 to 5, since I have concluded that the Banks’ appeal succeeds on Grounds 1 and 2, Venice’s appeal becomes academic and I can deal with the submissions shortly. The essence of Mr Cox KC’s submission on his first ground as to the applicable law for the restitution claim was that, applying the correct test of with which system of law does the claim have its closest and most real connection, the judge should have concluded that it was Italian law.

In relation to the common law (which is applicable given the date when the Transactions were entered), Mr Cox KC referred to Rule 230(2) in the relevant (14^th edition) of Dicey, Morris & Collins which provided:

“(2)

The proper law of the obligation is (semble) determined as follows:

a)

If the obligation arises in connection with a contract, its proper law is the law applicable to the contract.

b)

If it arises in connection with a transaction concerning an immovable (land), its proper law is the law of the country where the immovable is situated (lex situs);

c)

If it arises in any other circumstances, its proper law is the law of the country where the enrichment occurs.”

He referred to a number of authorities dealing with the principle, including Yugraneft v Abramovich [2008] EWHC 2613 (Comm) where Christopher Clarke J said that a) to c) of Rule 230(2) were indicators but did not replace the rule. He also referred to the Scottish case in the Outer House of the Court of Session in Baring Brothers v Cunninghame DC [1997] CLC 108 where, having gone through the authorities and academic writings, Lord Penrose concluded in relation to the same passage in the earlier 12^th edition of Dicey, Morris & Collins that: “so far as concerns the authorities referred to, the sub-rule [a)] is wholly without judicial support”. He went on to say that the rule was to be applied flexibly and that where the restitutionary remedy was required because of the inapplicability or failure of a contract: “the existence of the contract or the attempt by parties to enter into a contract, will be relevant and material factors in resolving any issue which arises, but will not per se be determinative of that connection.”

Mr Cox KC also referred to Prato and the passage from the judgment of Walker J at [164] which the judge in the present case cited at [388]. He relied on the conclusion of Walker J at [168], although he accepted that it was at the extreme end of the scale:

“My conclusion is thus that I should have no regard to provisions in the master agreement and schedule identifying English law as the governing law and recording an irrevocable submission to the jurisdiction of the English courts. In these circumstances, subject to a proviso advanced by Dexia, there can be no contest that the obligation to give restitution to Prato has its closest and most real connection to Italy.”

In Busto Cockerill J had distinguished Prato in part because of the weight to be given to English law as the law governing the question of capacity and in part because (unlike in the present case) the enrichment had taken place in England.

Mr Cox KC submitted that the Rule in Dicey was to be applied flexibly as the editors themselves recognised in the 14^th edition at [34-028] where they say:

“clause (2)(a) seeks only to assist in the identification of the proper law of the restitutionary obligation in circumstances of contractual failure; it does not state an inflexible rule which must be applied without exception to every case connected to a contract…”

He submitted therefore that the test was a flexible one whereby a restitutionary claim is governed by the proper law of the obligation which involves identifying the country with which the obligation, claim or “critical events” (in the words of Lord Penrose) have their closest and most real connection.

Mr Cox KC criticised the judge’s analysis at [390] of the judgment on the basis that he submitted that the judge had not applied that flexible test, but had concluded that, by reason of the choice of law clause in the void contract, the restitutionary claim was governed by English law. The judge did not apply any of the other factors connecting this claim with Italy: that the parties were Italian, performance was in Italy, the unjust enrichment was in Italy, everything was in Italy other than the English choice of law. Mr Cox KC also relied upon the fact that the mandate agreement between the Banks and Venice was governed by Italian law and the Banks were regulated by the Italian financial regulatory authorities, although he did not articulate how the mandate agreement remained relevant or even if it was still in force.

The oral submissions on (a) of Venice’s Ground 2, that the effect of allowing a change of position defence in principle was to substantially enforce contracts which were void pursuant to the ultra vires doctrine, were made by Mr Simon Paul. The primary case he advanced was that a change of position defence should not be available where it would prevent the reversal of benefits conferred under an ultra vires contract, because to allow it would stultify the ultra vires principle. His alternative case was that the defence was only precluded in an ultra vires context in cases, such as the present, of anticipatory reliance, because it was only in those cases that reliance was impermissibly placed on the assumed validity of a void contract as the basis for a change of position.

On either case, he submitted that English law denied a change of position defence where allowing it would stultify the policy reason for ordering restitution in the first place (the stultification principle). Mr Paul placed considerable reliance on the judgment of Clarke J in Svenska at 565b to e where the judge says that neither an estoppel nor a change of position defence is available in an ultra vires context, because otherwise the bank would in effect be relying upon the supposed validity of a void transaction. Mr Paul also referred to the decisions of Henderson J in the FII litigation ([2008] EWHC 2893 (Ch); [2009] STC 254 and [2014] EWHC 4302 (Ch); [2015] STC 1471) that the Revenue should not be entitled to rely on change of position as a defence to a Woolwich claim, because such a claim is founded on the unlawful levying of tax and therefore on the commission of a legal wrong. As I pointed out in argument, the rationale at play there was that the Revenue could not rely upon the change of position defence in circumstances where it had wrongfully levied the tax in the first place. That was the context in which the stultification principle arose. However Mr Paul submitted that the same principle should apply in the context of contracts which are void because they are ultra vires.

Mr Paul also referred to the decision of the Privy Council in Skandinaviska Enskilda Banken AB v Conway [2019] UKPC 36; [2020] AC 1111 which was a case where a defence of change of position was held to be unavailable in respect of a claim under a Cayman statute for reversal of an unlawful preference, following the position at English common law, referring in particular to [104] of the judgment of the Board. He submitted that this was an example of the same principle at play, albeit in a statutory context.

In relation to the ultra vires policy itself, Mr Paul pointed out, which was uncontroversial, that it exists for the protection of the public. He also noted that a party cannot raise a defence of any species of estoppel which involves reliance on an ultra vires act. He also referred to the decision of the Court of Appeal in Kleinwort Benson Ltd v Birmingham City Council [1997] QB 381 in which a defence of passing on was held not to be available to the council in answer to a claim in restitution by the bank. The council sought to argue that the bank had suffered no loss because it had received payments under hedging swaps equivalent to those it had paid the council. Mr Paul relied upon a passage at 400H in the judgment of Morritt LJ where he regarded the impact of the ultra vires doctrine as one of the reasons why the defence should not be available.

He submitted that the judge erred in law in concluding at [424] that a change of position defence was available to the Banks for two reasons. First, as he had submitted, if he was right that the ultra vires doctrine precluded a change of position defence on grounds of stultification, it necessarily followed that the judge had erred. Second, the judge erred in not following Svenska. Erroneously at [406]-[408] he had treated that case as superseded by Dextra and Commerzbank but neither of those was a case of the contract being void as ultra vires.

Mr Cox KC dealt with (b) of Venice’s Ground 2, that the judge should have followed Hobhouse J in Westdeutsche and decided the defence of change of position was not available in principle because the taking out of the hedging swaps was an independent choice by the Banks, which took the risk of non-payment by Venice under the swaps. He referred to the decision of the Privy Council in Goss v Chilcott [1996] AC 788, where a finance company advanced monies on mortgage to the defendant husband and wife who lent the monies to the wife’s brother, a director of the finance company, who undertook to repay the monies but did not. On a claim by the liquidator of the finance company in restitution, the defendants’ defence of change of position failed before the Privy Council, on the grounds that, when they paid the money over to the wife’s brother, they knew they were under an obligation to repay it to the finance company and so they took the risk that he would not repay the money, in which case they would still have to repay. The Court asked Mr Cox KC how this case helped Venice, since here it could be said that the Banks were doing the precise opposite to taking a deliberate risk. Rather they were seeking to reduce the risk by entering hedging swaps. He sought to answer by saying the risk the Banks took was of not receiving payments from Venice, although he accepted that this was very different from the risk in Goss v Chilcott. Essentially, his submission was that in taking out the hedging swaps, the Banks made an independent choice. However, when the Banks took out these outwards swaps, they knew that if they did not for whatever reason obtain payment from Venice under the inwards swaps, they would still have to pay under the outwards swaps, just like the defendants in Goss v Chilcott.

Mr Cox KC referred to Westdeutsche where at 948h-j, Hobhouse J rejected the change of position defence inter alia on the basis that the swap entered by the bank was wholly independent of the transaction between the bank and the council. The latter provided the motive for entering the swap, that was all.

In her submissions in opposition to Venice’s first ground of appeal, Ms Tolaney KC submitted that at [390] the judge had applied the correct legal test, of with which legal system the unjust enrichment claim has its closest and most real connection, as he stated in the opening words of that paragraph. His conclusion that this was English law was correct for three related reasons. First, the parties chose English law, not Italian law, as the governing law of the Transactions. Second, under Article 8(1) of the Rome Convention, English law was applied to determine whether the Transactions were void and, in particular, applying Haugesund it was English law, not Italian law, that determined the correct characterisation of the effect of Cattolica as being that the Transactions were beyond Venice’s capacity. Third, there is a very close and real connection between the law which determines whether the Transactions are void and the law that determines what claims can be brought as a result of finding they are void. The subsequent restitution claim arises directly from the invalidity of the Transactions, a point made by Joanna Bird of the University of Sydney in her chapter on Choice of Law in Restitution and the Conflict of Laws (1995). Ms Tolaney KC submitted that applying the same law produces a logically and legally coherent result, is consistent with the expectations of the parties and has significant practical advantages. Switching from English law to Italian law is more likely to produce a legally incoherent result. It would remove from the equation, at the restitution stage, what the judge referred to in [390(vi)] as the “security of receipt” which English law provides.

She pointed out that, at [390] to [392], the judge was not purporting to lay down a general rule. He gave most weight to the English choice of law clause on the facts and gave detailed reasons for reaching that conclusion. The suggestion that he somehow overlooked the connections with Italy on which Venice had relied before him in their pleadings and submissions was unreal. In only the previous paragraph [389] he had quoted Cockerill J in Busto engaging in the same exercise. She submitted that the determination of the question with which system the claim had its closest and most real connection was quintessentially a matter for a Commercial Court judge, with which this Court should not interfere unless the judge made an error of principle, which he did not.

Ms Tolaney KC submitted that, if necessary, she would argue that, although the Transactions might be void, the governing law clause, clause 13(a) of the ISDA Master Agreement was severable and remained valid, relying on [53(v)] of the judgment of Lords Hamblen and Leggatt JJSC in Enka v Chubb [2020] UKSC 38; [2020] 1 WLR 4117.

In relation to the authorities on Dicey Rule 230(2)(a), she submitted that the statement by Lord Penrose in Baring Brothers that it was without judicial support which Mr Cox KC had adopted (see [126] above] was not correct. She referred to the judgment of Evans J in Arab Monetary Fund v Hashim [1993] 1 Lloyd’s Rep 543 at 565 rhc where the judge said:

“An obvious case where sub-Rule (2)(a) may apply is where the contract proves ineffective and claims are made for recovery of sums paid under it”.

Ms Tolaney KC made the point that, even in Baring Brothers, Lord Penrose accepted that in a restitution claim arising out of a failure of the contract, the existence of the contract or the attempt to enter it was a relevant and material factor in determining with which system of law the claim had its closest and most real connection. Once it was accepted as a relevant and material factor, the question of what weight to give to it was a matter for the trial judge.

In relation to the Italian factors identified by Mr Cox KC, she submitted that none of them should be given much weight. The mandate agreement preceded the Transactions and had nothing to do with the benefits conferred under the Transactions. The Banks’ regulatory duties under Italian law were irrelevant, as they also concerned the period before the Transactions were entered. Italy as the place of enrichment and domicile has little to do with the substance of the obligation to make restitution.

In relation to Venice’s stultification and ultra vires argument, Ms Tolaney KC made the point that this was a private law claim by Venice relying on a mistake of law. The House of Lords in Kleinwort Benson had held that a claimant could bring a claim in restitution for mistake of law, but expressly on the basis that a change of position defence would provide a measure of protection. As I put to her in argument and she agreed, none of the authorities on mistake of law, that is a mistaken belief that a contract was valid when it was not, suggests that whilst the change of position defence would be available where the invalidity of the contract was for another reason, it would not be available where the invalidity was because the contract was ultra vires.

On the stultification principle, she referred to the recent decision of the Privy Council in Attorney-General of Trinidad and Tobago v Trinsalvage (“Trinsalvage”) [2023] UKPC 26 where the government employed Trinsalvage on a harbour development project and paid for some of the work but not for the rest. The lower courts allowed Trinsalvage to recover the value of the work via a quantum meruit claim and rejected the argument that that claim was inconsistent with the statutory tendering regime. The Privy Council dismissed the government’s appeal. Ms Tolaney KC noted that, in the judgment of the Board given by Lord Burrows, it was recognised that a claim in unjust enrichment will be defeated if it would stultify the relevant statutory policy. However, at [44] the Board had explained why allowing the unjust enrichment claim would not be enforcing the void contract by another means. Ms Tolaney KC submitted that that analysis was equally applicable here: there was no question of the Banks seeking to enforce the Transactions. The only question was whether historic sums should be paid back and allowing the Banks their defence of change of position did not give effect to the contract. It simply recognised that it would be unjust to require them to pay back equivalent sums, so that the loss should lie where it falls.

She submitted, by reference to the passage from Lord Burrows’ Restatement quoted in the judge’s judgment at [412] and the Privy Council decision in Conway, that the question was whether the policy justification for the right to restitution precludes a change of position defence. In Conway, the policy justification was that the unlawful preference rules are intended to protect the creditors as a class, so that creditors’ interests as a whole override the interests of one of them. This point was made at [111] and a contrast drawn with what was described “as a conventional claim to restitution of a benefit transferred from one individual to another by a defective transaction” where it is the private interests of the two parties in issue and the interests of the one in restitution can be cancelled out because of a change of position of the other. Ms Tolaney KC submitted that this restitution claim by Venice was in that latter category: it was a private law claim in unjust enrichment relying on mistake of law and it was well-established that a change of position defence was available in respect of such a claim.

Ms Tolaney KC referred to Haugesund, where the question arose whether the English court should give effect to a claim in restitution governed by English law if it was contrary to the policy of a foreign statute. The argument was rejected on the facts because, as Aikens LJ said at [102], there were no findings of fact by Tomlinson J that the effect of the Norwegian statute was that recovery of money paid under an invalid contract was barred or would be contrary to the statutory intent. Ms Tolaney KC submitted that the position was the same in this case, since Venice had not laid the groundwork for a stultification case at trial, had not pleaded the point and therefore had called no expert evidence on it. There were no factual findings in the judgment which could support the argument and there had been no evidence that Italian public policy would preclude a defence of change of position. She submitted that Venice had conceded the point in its closing submissions at trial by accepting that, in general, an anticipatory change of position defence could be raised against a local authority bringing a restitutionary claim in respect of payments made under a void contract. Although they argued that such a defence was not available where the change of position involved entering hedging contracts because they were wholly independent and not linked to the receipt of any payment by the Banks from Venice, that was a different point.

Ms Tolaney KC submitted that the authorities showed clearly that one could have a change of position defence to a restitutionary claim in respect of payments under an ultra vires contract. She referred to the speech of Lord Templeman in Hazell v Hammersmith & Fulham LBC [1992] 2 AC 1 at 36D-E expressly recognising that where a council sought restitution of payments made under an ultra vires contract, it did not follow that they could recover them, a recognition of the availability of a defence of change of position. Likewise in Kleinwort Benson at 382F-H, Lord Goff of Chieveley expressly recognised the availability of a change of position defence to a restitutionary claim by a local authority in respect of payments made under an ultra vires swap contract. At 412D Lord Hope of Craighead also referred to the availability of a change of position defence as recognised in Lipkin Gorman v. Karpnale Ltd. [1991] 2 A.C. 548. She also referred to the passage at 417G-H where Lord Hope said that, whilst the objection may be made that time may run on for a very long time before a mistake of law could have been discovered with reasonable diligence, there was no sound reason for declining to give effect to section 32(l)(c) of the Limitation Act and identified the availability of a defence of change of position as avoiding the risk of injustice.

So far as Westdeutsche was concerned, Ms Tolaney KC submitted that Hobhouse J was not saying that a change of position defence was automatically ruled out because the contract was ultra vires. It depended on the facts and that was a case decided on its own facts. Likewise in Svenska, Clarke J was not saying that a change of position defence could never be available if the contract was ultra vires. She also referred to [471]-[472] of Foxton J’s judgment in Schools Facility Management (repeated at [405] of his judgment in this case) citing the Privy Council statement in Dextra that the exclusion of anticipatory reliance in Svenska depended on the “exceptional facts” of that case, whilst recording that the decision of Clarke J had been the subject of criticism in Goff & Jones. As I pointed out in argument, that may have been a polite way of the Privy Council saying Svenska was wrongly decided. Ms Tolaney KC noted that at [473] to [478] in Schools Facility Management, Foxton J had said that there can be no objection in principle to allowing a party who receives an ultra vires payment from a public body to advance a change of position defence and had gone on to explain in a comprehensive fashion the flaw in Svenska. In Busto, Cockerill J said at [414]-[415] that, in the light of Foxton J’s judgment in Schools Facility Management, she would not have been minded to follow the judgments in Westdeutsche and Svenska.

In relation to Kleinwort Benson v Birmingham on which Venice had relied, Ms Tolaney KC pointed out that it was not a case about change of position but passing on and that at 394A-B Evans LJ had expressly reserved his position on whether a change of position defence would be available, saying:

“…the existing authorities of Kleinwort Benson Ltd. v. South Tyneside Metropolitan Borough Council [1994] 4 All E.R. 972 and South Tyneside Metropolitan Borough Council v. Svenska International Plc. [1995] 1 All E.R. 545 do not establish a clear rule that no such defence can be raised. The Svenska case in particular turned on special facts, and I prefer to express no view as to whether that defence, which was recognised in Lipkin Gorman v. Karpnale Ltd. [1991] 2 A.C. 548, could ever be established by reference to market transactions.”

A similar reservation was made by Saville LJ at 394F.

So far as Venice’s alternative case that a defence of anticipatory change of position was not available in the case of ultra vires contracts, she submitted that such a defence was well established by Dextra which post-dated both Westdeutsche and Svenska. It was impossible to see why public policy concerns were engaged where the change of position came first but not where it came second.

Finally, on Mr Cox KC’s risk-taking point, Ms Tolaney KC pointed out that he relied on a principle that a change of position defence will be ruled out where the claimant knows that he will have to repay the money anyway. However, that was not this case, because the Banks thought that the money they received under the Transactions was theirs to keep. Furthermore, it was irrelevant that Venice was not a party to the hedging swaps. Most changes of position will involve a transaction between the defendant and a third party. Mr Cox KC’s argument was really an attempt to rerun the causation case he had run below on which he had lost and not appealed the judge’s finding of a causal link at [413].

The important initial question in relation to Grounds 1 and 2 is where on the spectrum identified by Lord Hodge in Perry the present case lies. Clearly, to the extent that the judge made findings as to what Italian law was, for example as to what the Italian Supreme Court in Cattolica decided, he was dependent upon the evidence from the Italian law experts about a civil law system with concepts unfamiliar to English lawyers and with a variety of sources of law. It is well-established that, in relation to those findings of Italian law, the approach in this Court should be that advocated by Lewison LJ in FAGE v Chobani and Volpi v Volpi in relation to findings of fact, including findings of foreign law, namely that this Court will not interfere with those findings unless they were plainly wrong: see the judgments of the Court of Appeal in Dexia at [34]-[42] and Byers at [98]-[105], approved by the Privy Council in Perry.

However, to the extent that the judge’s analysis and conclusions are based on his application of Italian law to the facts and, in particular, where they are not based on the Italian law expert evidence that, as a matter of Italian law, the facts required a particular answer, then there is more scope for this Court to interfere. This is an evaluative exercise to which the judge has brought his own legal skill and understanding, in the same way as he would in the case of a foreign legal system which was based on English law, and this Court is as well able as the judge to form a view as to the correct analysis. This is similar to the point made by the judgment of the Court of Appeal (Evans, Morritt and Chadwick LJJ) in MCC Proceeds Inc v Bishopsgate Investment Trust plc [1999] CLC 417 at [13]:

“When and to the extent that the issue calls for the exercise of legal judgment, by reference to principles and legal concepts which are familiar to an English lawyer, then the court is as well placed as the trial judge to form its independent view.”

Nevertheless, it is important to note that, although there is more scope for this court to interfere, if the judge has engaged in an evaluative exercise in relation to the facts, this Court will still exercise caution and only interfere if satisfied that the judge has erred in principle, for example by failing to take into account a material consideration.

Despite Mr Cox KC’s argument to the contrary, I consider that the section of the judgment headed: “Were the Transactions Speculative?” running from [224] to [232], most of which I have cited at [34] above, does not involve conclusions of Italian law based on the expert evidence, but the judge’s own evaluation of whether an Italian court would conclude that the Transactions were speculative, in effect his own application of Italian law to the facts. It is striking that nowhere in that analysis does the judge say that he finds the Transactions to be speculative on the basis of the expert opinion of Venice’s expert. This is because the Italian law expert evidence was confined to the question what the test was for speculation as a matter of Italian law, namely whether it was confined to the CONSOB Determination or was broader. Furthermore, as the judge recorded at [202], Italian law does not provide a definition of what constitutes a speculative derivative, nor does Cattolica (as the judge noted at [222]).

The experts were not asked to express an opinion as to how that test for speculation applied to the facts. Venice submitted in its closing submissions at trial, referred to at [100] above, that the determination of whether the Transactions were speculative was a question of fact for the judge. It follows that the Italian law experts did not address the critical question on the appeal whether, for the purpose of deciding whether the Transactions were speculative, a comparison should only be made between the Transactions and the Amended Rialto Bond, with the pre-existing liabilities under the Bear Stearns swap being irrelevant.

On any view, the judge’s judgment is impressive and well-reasoned. However, despite Mr Cox KC’s submissions seeking to justify the judge’s analysis of the issue whether the Transactions are speculative, I consider that the judge made a number of errors of principle. What might be described as the root error is his failure to factor into his analysis that the Bear Stearns swap was a valid contract which amounted to hedging. As set out at [8] above, Venice did seek to argue before the judge that the Bear Stearns swap was invalid, not on the ground that it was speculative, but on the ground that it had not been approved by the City Council. That argument was dismissed by the judge at [324] to [330] of the judgment. The judge should therefore have concluded that the Bear Stearns swap was hedging and was valid, binding Venice at the time of the restructuring of the Rialto Bond. However, in his analysis at [224] to [232], he does not recognise or give effect to that important consideration.

In my judgment, his failure to recognise that the Bear Stearns swap was a valid hedge led to a misanalysis on his part of the Transactions, as Ms Tolaney KC submitted. At the time when Venice restructured the Rialto Bond, because of the new terms and extended maturity date, it needed to renegotiate and restructure the Bear Stearns swap to bring it into alignment with the restructured Rialto Bond or (if Bear Stearns refused such a restructuring, as proved to be the case) enter into a new swap with another bank or banks which was aligned with the restructured Rialto Bond. Had Bear Stearns agreed to restructure its swap, it would be absurd to suggest that Italian law would have prohibited the restructuring of the Rialto Bond and the hedging swap. A new Bear Stearns swap aligned to the restructured Rialto Bond would have been just as much a hedge as the swap it replaced and yet, any restructuring by Bear Stearns would have had to take account of the existing negative MTM, which could not be wished away. Bear Stearns would have restructured its swap to correlate with the Amended Rialto Bond and rolled over the negative MTM so that, in effect, exactly what happened with the Banks would have happened with Bear Stearns, but it would still have been hedging, not speculative.

Because Bear Stearns would not renegotiate its swap, Venice had to approach the Banks, since as the judge found at [83], Venice was not willing to pay cash to cancel out the Bear Stearns swap, given its budgetary constraints. The Transactions had two elements which proceeded simultaneously. First, the Banks entered into the novations with Bear Stearns. They did so on their own account, not as agents for Bear Stearns. The price paid for the Banks standing in the shoes of Bear Stearns was the novation fees, which corresponded with the negative MTM at that time. Because the Bear Stearns swap, as novated to the Banks was not aligned with the Amended Rialto Bond, it needed to be restructured, hence the second element, which was the entering of the new swaps with the Banks, with an extended maturity date matching that of the Amended Rialto Bond and a cap on the variable rate of interest payable under the Amended Rialto Bond. The new swaps had an interest rate floor setting a fixed rate payable by Venice to the Banks, but the disparity between the cap and the floor, which so troubled the judge that it led him to conclude that the Transactions were speculative, corresponded with the negative MTM as it had been under the Bear Stearns swap, which the Banks had closed out by paying the novation fees. Contrary to the judge’s conclusion, that negative MTM was an existing exposure of Venice originally to Bear Stearns, now to the Banks. It was not some new exposure or risk, contrary to [229(vi)] of the judgment. If that existing exposure had remained to Bear Stearns, it would not have somehow converted a valid hedging swap into something speculative when the swap was restructured. It is difficult to see how, merely because the same exposure or risk was now to the Banks, what would otherwise have been hedging became speculative.

The judge correctly recognised at [229(i)] that the Transactions were explicitly carried out in the terms adopted both to reduce the risks connected with the Rialto Bond and to cover the winding-up costs of the Bear Stearns IRS, so that limb (a) of the CONSOB Determination was satisfied. He also accepted at [228(i)] that important terms of the Transactions corresponded with the Amended Rialto Bond, namely the Notional Amounts, the amortisation rate, the maturity date and the variable interest rate received from the Banks which corresponded with the rate payable under the Amended Rialto Bond. In those circumstances, he should have concluded that limb (b) of the CONSOB Determination was satisfied, because there was a high correlation between the characteristics of the underlying debt and those of the derivative transaction. On the basis of the CONSOB Determination, that should in turn have led to the conclusion that, since both limbs were satisfied, the Transactions amounted to hedging and therefore were valid and binding contracts.

The opinion of the Banks’ expert, Professor Gentili, was that the test for speculation was to be found in limbs (a) and (b) of the CONSOB Determination, as approved by the Supreme Court Decision No 19013/2017. Venice’s expert, Professor Alibrandi, considered that the CONSOB Determination was not necessarily exhaustive, although as set out at [77] above, she accepted in cross-examination that, if the Transactions satisfied limbs (a) and (b) of the CONSOB Determination, they were likely to be hedging. In his judgment at [208] and [209] the judge said that the CONSOB Determination was of assistance in determining whether or not a derivative was a hedge but not exhaustive. He reached that conclusion in part because, as he said at the end of [209]:

“Finally, there are many Italian court decisions which have addressed the issue of whether a derivative was hedging or speculative in nature without referring to the CONSOB Determination or applying the three-stage test…”

In my judgment, there are two errors in that approach. First, on the basis of the Italian law expert evidence as it stood at the end of the trial, the judge should have concluded that the CONSOB Determination was, at the least, likely to be determinative such that, if limbs (a) and (b) were satisfied, the derivative was hedging. Second, as set out in more detail below, reference to a whole series of Italian lower court decisions (in which the judge then engaged at [212]) involved asking the wrong question. The right question for the judge was what the highest Italian court, the Supreme Court, would have decided was the status of the CONSOB Determination, to which the answer was to be found in Decision No 19013/2017.

The judge did not reach the conclusion that limb (b) of the CONSOB Determination was satisfied, because he considered that the very significant difference between the negative MTM of the cap and the floor had been arrived at for different reasons than correlation between the Amended Rialto Bond and the Transactions ([229(ii)]) and rendered the Transactions speculative ([229(iii) to (vi)]). In my judgment, there a number of errors in that analysis. First, as Ms Tolaney KC submitted (see [79] above) the judge seems to have erroneously treated the novation of the Bear Stearns swap as somehow a separate event distinct from the other aspects of the restructuring. I agree with her that there is a similar error in [225(vi)] where the judge analyses the rolling over of the negative MTM as if it were a discrete part of the transaction having a separate function, whereas the swap was all one transaction.

The second and more fundamental error in [229] is that the suggestion in (vi) that, in entering the Transactions with the significant difference between the cap and the floor, Venice was taking on a significant new risk, is just wrong. The difference constituted the rolling over of the negative MTM to which Venice was already exposed under the Bear Stearns swap. Since that exposure did not render what was a valid hedging swap speculative, it is difficult to see how rolling over of the exposure into the Transactions could render them speculative. In any event, the existence of an initial negative MTM in the Transactions does not in itself amount to speculation, as Cattolica recognised.

The error in the judge’s analysis is reflected in his adoption of Mr Cox KC’s closing submissions at [230] of the judgment. Quite apart from the fact that, as Ms Tolaney KC submitted, there is no question of the rolling over of the negative MTM being akin to borrowing, as she also submitted the analysis overlooks that the Bear Stearns swap was not speculative, so that the negative MTM was a non-speculative risk to which Venice was already exposed. The novation and entering of the Transactions could not turn what had previously been hedging into speculation. The error in the judge’s analysis is also reflected in [224] where the judge referred to Mr Cox KC cutting to the heart of the matter, the fact that the terms of the swap had been structured so as to cover the amounts which the Banks had to pay Bear Stearns to close out its swap. Once again, that overlooks that the Bear Stearns swap was not speculative and the negative MTM was a pre-existing non-speculative risk to which Venice was already exposed.

The judge sought to support his conclusion that the Transactions were speculative by a fairly detailed analysis at [212] of what were (with the exception of the Decision of the Supreme Court No. 21830/2021, in which the Supreme Court concluded that the vanilla swap in that case was a hedge) a whole series of Italian first instance and regional court of appeal judgments in disputes concerned with swaps, which then fed into his conclusions at [229]. This was essentially the judge’s own English lawyer’s analysis of the Italian cases. He did not rely upon expert evidence of Italian law about those cases. Although at the end of his submissions, Mr Cox KC produced a schedule which purported to show references for each of the cases to expert evidence, as Ms Tolaney KC pointed out, this seemed to just be a word search through the transcripts and trial documents to all references to the cases, not focused at all on the issue of what constituted speculation. It is striking that nowhere in [212] or [229] does the judge refer to expert evidence and in my judgment Ms Tolaney KC is right that the judge did not rely in his analysis of the Italian cases on expert evidence of Italian law.

In any event, even if the judge had been relying on expert evidence, by analysing lower court decisions, the judge was asking the wrong question. As the Court of Appeal put it succinctly in Prato at [34]:

“In the case of disputed questions of foreign law, the task for the trial judge is to determine what the highest relevant court in the foreign legal system would decide if the point had come before it.”

Some of the lower court decisions were inconsistent with decisions of the Supreme Court. In many of those decisions, the swap was found to be speculative because of a significant disparity between the swap and the underlying debt, which, as Ms Tolaney KC said, was not the case here where there was high correlation between the Transactions and the underlying Amended Rialto Bond.

Although decisions of a lower court may in some circumstances be evidence of what the highest foreign court would decide if the question were to arise, those decisions do not assist in the present case for two reasons. The first is that there was no evidence from the Italian law experts that, by reason of those decisions, the highest court would hold that the Transactions were speculative. In the case of a system of law such as the Italian system, which is not based on case law or precedent, that is a material omission. The second is that the judge did have before him relevant decisions of the highest courts, namely Cattolica, which did not lay down a test for what amounted to speculation in the context of a restructured IRS, and Decision No. 19013/2017, which found that the applicable test was as set out in the CONSOB Determination.

The judge compounded his error of relying on lower court decisions to determine whether the Transactions were speculative by also relying on various English cases at [214] to [218], which also fed into his overall analysis at [229]. The English cases were wholly irrelevant to the question the judge had to determine, whether the Italian Supreme Court would have decided that the Transactions were speculative. To be fair to the judge, he seems to have gone down the route of using the Italian lower court cases and the English cases to determine whether the Transactions were speculative because that was the approach urged on him by Venice in their submissions. However, the approach was erroneous for the reasons I have given.

In my judgment, if the judge had focused on the right question and taken account of the fact that the Bear Stearns swap was a hedge which was a valid contract, he would have concluded that the Italian Supreme Court would have concluded that the Transactions were also hedging. They gave Venice the benefit of an extended maturity period and other terms to correlate with the Amended Rialto Bond, without altering the economic effect of the Bear Stearns swaps. On the basis that the Bear Stearns swaps were hedging, there was no logical or legal reason to conclude that the Transactions were anything other than hedging. Accordingly, Ground 1 of the appeal succeeds.

Ms Tolaney KC accepted that, to succeed on the appeal overall, the Banks need also to succeed on Ground 2, although she submitted that the two Grounds 1 and 2 stood or fell together. In relation to Ground 2, the first question which arises is whether, as the judge concluded, the payments by the Banks to Bear Stearns by way of novation fees constituted “upfront” payments. As with his analysis and conclusions as to whether the Transactions were speculative as a matter of Italian law, the judge’s analysis at [261] as to why the novation fees constituted upfront payments was really his evaluation and application of the relevant definition of an upfront payment in [4.6] of Cattolica to the facts of this case, so that it was at the end of the Perry spectrum which makes his conclusion more amenable to review in this Court, provided always that some error of principle on his part is identified.

It was agreed between the Italian law experts, as set out at [83] above, that an upfront payment in the context of derivatives is an amount of money paid by one party to another to rebalance the financial situation of the parties in “non-par swaps” i.e. swaps whose value at inception is not equal to zero. On the face of it, the payments of novation fees by the Banks to Bear Stearns do not fall within that definition. Notwithstanding that, Professor Alibrandi’s evidence was that the payments would be construed as upfronts on the basis that Venice had requested the Banks to make the payments, although, as Ms Tolaney KC pointed out, in cross-examination, Professor Alibrandi accepted that “from a formal point of view” it was correct that the Cattolica definition did not apply in the present case. There was no exploration of what exactly she meant by that or what its implications were, but in any event, in his analysis of the upfront point at [261], the judge does not seem to have relied on her analysis.

Ms Tolaney KC submitted that the judge’s analysis at [261] was wrong: there was no payment by the Banks to Venice to rebalance the transaction. The novation fees were paid to Bear Stearns and, in any event, the negative MTM was rolled over, not balanced out. I agree with her submission that, at (iii), the judge misinterpreted Professor Gentili’s evidence. He did not accept in cross-examination that, if Venice had asked the Banks to make the payment to Bear Stearns, it would still be treated as an upfront payment by the Banks to Venice. What he accepted was much narrower: that if Venice was entitled to receive a payment, but had requested the Banks to pay it to Bear Stearns, in effect as Venice’s agent, then it would be an upfront payment. As Ms Tolaney KC said, that was not this case given that any suggestion that the Banks were acting as agents for Venice in entering the novations is flatly contrary to the express terms of the ISDA Novation Agreement and Definitions.

The essence of Mr Cox KC’s argument as to why the novation fee payments were upfront payments was that the judge was correct that it made no difference to the analysis that the payment was made direct to Bear Stearns, rather than to Venice, which then paid Bear Stearns to close out its swap, because it had the same economic effect as an upfront payment by the Banks to Venice. As he put it, Venice benefitted from the payment by the Banks because the Bear Stearns swap was discharged, but was expected to pay it back through the terms of the swap. In my judgment, like the judge’s analysis on Ground 1, this argument overlooks that the Bear Stearns swap was valid hedging, under which the negative MTM was an existing exposure which Venice faced. If Bear Stearns had agreed to renegotiate its swap, I do not possibly see how it could be said that, in rolling over the negative MTM into a restructured swap, Bear Stearns was making an upfront payment. I do not consider that, in circumstances where the Banks took over the Bear Stearns swap and paid the novation fees effectively to stand in the shoes of Bear Stearns, it can be said that the novation fees somehow became an upfront payment.

In any event, even if the judge were right in his analysis at [261] and the novation fees were upfront payments, the further question still remains whether the judge’s analysis at [268], that the payments were not “for the purposes of financing investment expenditure” within Article 119(6) is correct. In my judgment, it is not since, as Ms Tolaney KC submitted, the judge’s reasoning is infected by the same errors as he made in relation to Ground 1. If, as I have found, the judge was in error in concluding that the Transactions were speculative, then it must follow that his conclusion that what he found were upfront payments fell foul of Article 119(6) must be equally flawed. It is also the case that, although Professor Domenichelli’s evidence was not entirely clear, he did appear to accept that a hedging swap would not infringe Article 119(6).

The judge’s reasoning at [268(ii)] seems to be his own assessment of what the position would be under Italian law. It was not suggested that any of the Italian law experts addressed this question. In my judgment, his reasoning is in error. There was no suggestion that the restructuring of the Rialto Bond fell foul of Article 119(6). It was clearly for the purposes of financing investment expenditure and, on that basis, the entering of the Transactions which were hedging and not speculative, as an integral and necessary part of that restructuring, was likewise for those purposes and not contrary to Article 119(6).

It follows that, for all those reasons, the Banks’ appeal must be allowed on Grounds 1 and 2 and the judge’s conclusion, reflected in the declaration at [1] of the Order of 8 February 2023, that Venice lacked capacity to enter the Transactions because they were speculative and contravened Article 119(6) so that the Transactions are void and unenforceable as a matter of English law, must be set aside. Venice did have capacity to enter the Transactions which were hedging, not speculative, and the Transactions did not fall foul of Article 119(6). The Transactions were and are valid and are binding on Venice.

That conclusion that the Banks’ appeal succeeds on Grounds 1 and 2 means that the other grounds of their appeal and Venice’s appeal, all of which are predicated on Grounds 1 and 2 of the Banks’ appeal failing, are academic. However, I will deal with them (at least those for which permission to appeal has been given) albeit more briefly than if they were determinative of the issues.

Ground 3 of the Banks’ appeal involves two steps. The first, which arises on the basis (contrary to my conclusion on Grounds 1 and 2) that the Transactions were speculative and/or contrary to Article 119(6) of the Italian Constitution, is whether, as a matter of Italian law Venice had power to enter into those Transactions. The second is whether, if so, the lack of such a power is to be characterised in English law as meaning that Venice lacked capacity. As to the first step, I consider that, if the judge had been correct that the Transactions were speculative and/or fell foul of Article 119(6), he would have been entitled to find that, on that basis, Venice would have lacked power under Italian law to enter the Transactions rather than Article 119 imposing a prohibition against the exercise by Venice of a substantive power conferred on it. As Mr Cox KC correctly submitted, the Supreme Court in Cattolica decided that Article 119 conferred a power on local authorities to enter swaps provided that they were not speculative. It is no answer that other legislation gave local authorities general powers to contract. The position is akin to that in Haugesund, where section 50 of the relevant Norwegian law was found by Tomlinson J and the Court of Appeal to limit the power of the local authority to enter a swap albeit against the backdrop of a more general power to enter contracts.

In my judgment, although one has to be cautious about reading too much into a translation from the Italian, the wording of Article 119(6): “They may have recourse to indebtedness only for the purpose of financing investment expenditures” more naturally points to a limit on the power of local authorities to enter swaps, rather than a prohibition on their exercising a substantive power conferred under some other statute.

On the basis that Article 119 limits the power of local authorities to enter swaps rather than being a prohibition on the exercise of a power, as a matter of English law the Article is to be characterised as going to capacity. As Tomlinson J said in Haugesund at [123] of his judgment in a passage approved by the majority of the Court of Appeal in [59] of the judgment of Aikens LJ: “the conclusion cannot I think be escaped that a lack of substantive power to enter into an agreement can only properly be characterised as going to capacity.” Once it is established that the provision goes to capacity, then it must follow, on the basis of the decision of the majority of the Court of Appeal in Haugesund, that the Transactions would have been void.

Ground 4 of the Banks’ appeal contends that the judge was wrong to apply the law as stated in the Supreme Court decision in Cattolica in May 2020 to the issue of the validity of the Transactions entered into in 2007 because the decision in Cattolica was not reasonably foreseeable and constituted a retrospective change in the law. This point was not argued below and the Banks require permission to appeal to argue it. If the point had been necessary to decide to determine the outcome of the Banks’ appeal, I would have been inclined to permit them to argue it and to give permission to appeal. Contrary to Mr Cox KC’s submissions, it is not a point on which fresh evidence would need to be adduced and Venice would not suffer any prejudice through it being argued.

However, as already said, it is not necessary to decide it and anything which this Court said would be obiter. The issue whether the decisions of a foreign highest court should be given retrospective effect by an English court is a difficult, albeit interesting, one on which there is no authority of direct relevance. It does not seem to me that the issue could simply be decided by reading across the decisions in Lynch and/or Adams. In those circumstances, it seems to me better to refuse permission to appeal and to leave the issue for decision in another case where it would be determinative of the appeal. In this context, I have in mind the salutary and wise observation of Mummery LJ in Housden v The Conservators of Wimbledon and Putney Commons [2008] EWCA Civ 200; [2008] 1 WLR 1172 at [30]:

“It is unnecessary to decide the issue for the purpose of disposing of the appeal. In general, it is unwise to deliver judgments on points that do not have to be decided. There is no point in cluttering up the law reports with obiter dicta, which could, in some cases, embarrass a court having to decide the issue later on.”

I propose to deal with Ground 5 of the Banks’ appeal and the two Grounds of Venice’s appeal which collectively relate to the restitution counterclaim by Venice in the same order as did the judge, since logically, the first issue which should be determined in this context is whether the applicable law for the restitution claims is English law or Italian law (Ground 1 of Venice’s appeal). It is common ground that the correct legal test is that the law which applies is the law of the jurisdiction with which the unjust enrichment claim has its closest and most real connection. Whilst Mr Cox KC is no doubt right that, at common law, there is not an inflexible rule that, where the unjust enrichment claim is for repayment of monies paid under a contract which is void or invalid, the applicable law is the law which would have governed the contract if it had been valid, in my judgment the judge was correct in concluding that, in a case such as the present, the applicable law for the restitution claim is English law, since that was the law which governed the Transactions. I agree with Ms Tolaney KC that the judge’s analysis at [390] of his judgment as to why the applicable law is English law is correct and cannot be faulted.

There is an obvious very close and real connection between the law which determines whether the Transactions are void which, by virtue of Article 8(1) of the Rome Convention, is English law and the law which determines whether restitution claims can be brought as a result of finding that they were void. The restitution claim arises directly from the invalidity of the Transactions. The connection is all the stronger since, applying the decision of the majority of the Court of Appeal in Haugesund, it was English law, not Italian law, which determined that the correct categorisation of what Cattolica decided was that the Transactions were beyond Venice’s capacity and, therefore, void. It would be odd if, that issue having been determined by English law, the proper law of the Transactions, there was a switch to Italian law to determine what, if any, claims were available to Venice once English law had determined the Transactions were void.

I find the analysis of Joanna Bird in Restitution and the Conflict of Laws at pp 123-4 compelling:

“Legal systems may have an internal balance which is destroyed if one part of what is essentially a single problem is dealt with by one legal system and another by a different legal system…where the parties have chosen the law to govern their contract…it is almost inconceivable that the parties expected another law to apply to unjust enrichment actions which flow from the contract.”

This is essentially the same point as the judge was making in [390(vi)], that it would be the natural expectation that English law would apply to issues relating to security of receipt and rights of recovery.

I do not consider that there is anything in Mr Cox KC’s suggestion that the judge failed to consider the various factors which pointed towards Italian law and weigh those against the factors pointing towards English law. The judge had referred two paragraphs earlier, at [388], to the judgment of Walker J in Prato which found that the restitution claim was governed by Italian law and in [389] to the judgment of Cockerill J in Busto distinguishing Prato, so he evidently had the Italian factors well in mind. In any event, I agree with Ms Tolaney KC that the Italian factors on which Mr Cox KC relied are of little, if any weight. Whilst the parties were Italian, performance was in Italy and the unjust enrichment was in Italy, the performance and the unjust enrichment occurred because payments were made pursuant to what both parties considered were contractual obligations under an English law contract which turned out to be void as a matter of English law. Accordingly, any issues of payment or unjust enrichment seem to me to have a much closer connection with English law. As Ms Tolaney KC submitted, the mandate agreement may have been governed by Italian law, but it preceded the Transactions and was of no relevance to the terms of the Transactions. It is unclear whether it remains in force. The fact that the Banks had regulatory duties under Italian law is irrelevant, as they also concerned the period before the Transactions were entered.

In all the circumstances, I consider that the judge was right to conclude that Venice’s unjust enrichment claims were governed by English law. Logically, the next issue must be whether the judge was wrong to hold that, by reason of section 32(1)(c) of the Limitation Act 1980, Venice’s claims in restitution for payments made more than six years before the issue of the claim form were not time barred. The test as to when a mistake of law could with reasonable diligence have been discovered is that laid down recently by the Supreme Court in FII. As set out at [112] above, the Supreme Court found that time runs from the point in time when the claimant discovered, or could with reasonable diligence have discovered, his mistake, in the sense of recognising that a worthwhile claim had arisen or that he had been mistaken with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking advice and collecting evidence.

The judge recognised that this was the correct test at [429] of his judgment, but I consider that he then misapplied the test in concluding at [430] that Venice could not with reasonable diligence have discovered that it had a worthwhile claim prior to the decision of the Italian Supreme Court in Cattolica. I agree with Ms Tolaney KC that it is difficult to see what relevance his points (i) and (ii), which attached significance to the fact that England was the contractual forum, had to the limitation issue, given that any defence Venice had of lack of capacity would be based in the first instance on Italian law. As Ms Tolaney KC also said, in (iii) the judge focused on the likely position in England after the decision of Walker J in Prato in 2015, but that was the wrong time at which to assess discoverability. Applying the FII test properly, a local authority in the position of Venice exercising reasonable diligence would have recognised that it had a worthwhile claim justifying preliminary steps towards issuing proceedings in the Commercial Court from around the time when Prato issued its claim in 2010. Many Italian local authorities did precisely that. It is nothing to the point that Walker J subsequently decided that Prato’s claim failed, which seems to have been what influenced the judge in (iii). Venice did not need to know that its claim would succeed, just that it was able to plead a proper case, which it could have done in 2010 in the same way as Prato did.

Since, as Ms Tolaney KC put it, Venice hitched its limitation wagon entirely to the date of the Supreme Court decision in Cattolica in May 2020 and did not put forward any alternative earlier date for the purposes of section 32(1)(c), the judge should have concluded that the case based on that subsection failed and that, in so far as Venice’s restitution claim related to payments made before 15 August 2013, it was time barred.

The final issue concerned with the restitution claim is that raised by Venice’s second ground of appeal, whether the judge was wrong to conclude that a change of position defence was available to the Banks in principle in respect of payments by the Banks under their back-to-back hedging swaps. So far as concerns the stultification argument, that to allow such a defence would stultify the policy that contracts which are ultra vires should not be enforceable, I consider that argument to be mistaken for a number of reasons. To begin with, none of the authorities on the availability of a restitution claim where there has been a mistake of law consisting of a mistaken belief that a contract was valid when it was not, suggests that a change of position defence to such a claim is not available to a defendant where the invalidity or voidness of the contract is because it was ultra vires as opposed to for any other reason. There would be no sensible basis in public policy or otherwise for drawing such a distinction.

Furthermore, it is now well-established that where a bank has paid a local authority under a swap contract which has turned out to be invalid or void because it was ultra vires the local authority, the bank is entitled to recover payments made by way of a claim in restitution for unjust enrichment. That claim is not one based on implied contract nor is it one which seeks indirectly to enforce an ultra vires contract. This is clearly the effect of the decision of the House of Lords in Westdeutsche [1996] AC 669, overruling the reasoning in its own earlier decision in Sinclair v Brougham [1914] AC 398. This was explained by Aikens LJ in [62]-[88] of his judgment in Haugesund, expressly approved recently by Lord Burrows giving the judgment of the Privy Council in Trinsalvage at [44]. At [87], Aikens LJ concluded:

“My conclusion is that the majority of the House of Lords in the Westdeutsche Landesbank case did depart from the decision in Sinclair v Brougham that a lender under a borrowing contact that is void because ultra vires the borrower, cannot recover the sum lent in a restitutionary claim at law. As a result of the decision of the majority of the House of Lords in the Westdeutsche Landesbank case such a claim can be advanced. It is, of course, not a claim based on any implied contract or promise and it does not indirectly enforce an ultra vires contract, for the reasons given by Lord Goff in Westdeutsche Landesbank Girozentrale v Islington London Borough Council [1996] AC 669, 688G-H. Moreover, I respectfully agree with him that any such restitutionary claim must be subject, where appropriate, to any available restitutionary defences, including any that can legitimately be based on public policy. If I am correct then Sinclair v Brougham can “fade into history” as Lord Goff hoped it would.”

If such a claim by a bank in restitution is not seeking to enforce indirectly an ultra vires contract, it is difficult to see how, when the restitution claim is the obverse, that is a claim by the local authority for restitution in respect of payments made by it to the bank or banks, a defence to that claim of change of position is somehow seeking to enforce indirectly an ultra vires contract. In my judgment it is not and there is no reason of public policy why such a defence should not be available in principle. As Ms Tolaney KC submitted, the claim in restitution by Venice is a private law claim and what was said by the Privy Council in Conway at [111] is applicable in those circumstances:

“…it is the private interests of the two parties which are mainly in issue, and the equitable considerations requiring the benefit to be returned to the plaintiff can be cancelled out by equitable considerations arising from a change of position on the part of the defendant.”

I also accept that the availability of a change of position defence to a restitution claim in respect of an ultra vires contract has been recognised in the House of Lords in the passages from speeches referred to at [146] above. Perhaps the clearest statement is that of Lord Goff in Kleinwort Benson, who, in recognising the availability of a restitutionary claim in respect of payments made under swap contracts which were void because they were ultra vires, expressly recognised that there could be a change of position defence to such a claim.

So far as the alternative case advanced by Mr Paul is concerned, that the defence is only precluded in an ultra vires context in cases of anticipatory reliance, this is equally mistaken. To the extent that Hobhouse J decided such a defence was not available in Westdeutsche, that was a decision on its own facts, as was the judgment of Clarke J in Svenska. The law of restitution has moved on since those cases were decided and I doubt whether they would be decided in the same way today. I note that in Busto, Cockerill J said that she would have declined to follow those decisions. In Dextra, which post-dated both first instance decisions, the Privy Council recognised that a change of position defence was available in cases of anticipatory reliance. Foxton J himself set out at [473] to [478] of his judgment in School Facility Management a compelling analysis of why anticipatory reliance was not seeking indirectly to enforce an ultra vires contract and why the change of position defence should be available in such a case. That analysis is equally applicable here.

I also agree with Ms Tolaney KC that, in any event, Venice failed to lay the groundwork at trial for this stultification defence. It was not pleaded and no Italian law expert evidence was adduced on the point. It followed that Venice did not call any evidence to suggest that Italian public policy would have precluded a defence of change of position in the present case. The position is akin to that in Haugesund referred to at [145] above. I also agree with her that this point was effectively conceded by Venice in its closing submissions at trial as set out in the same paragraph.

In relation to Venice’s second sub-ground, that the judge should have followed the decision of Hobhouse J in Westdeutsche and decided that the change of position defence was not available because the taking out of the hedging swaps was an independent choice by the Banks, which took the risk of non-payment by Venice under the swaps, that argument is also misconceived. It is clear that a change of position defence will not be available to a defendant who receives the money knowing that he will have to repay it at some stage in any event, as Goss v Chilcott demonstrates. However, that is not this case, since the Banks received payments under the Transactions believing the Transactions to be valid so that the money was theirs to keep.

As for the suggestion that the hedging swaps were wholly independent of the Transactions, taken out by the Banks for their own purposes. and therefore a change of position defence was not available, in my judgment the judge’s reasons for rejecting that argument at [411] to [413] of his judgment (summarised at [51] above), are unimpeachable.

It follows that, for the reasons I have set out, if it had been necessary to decide Venice’s appeal, I would have dismissed it. I would have allowed the Bank’s appeal on Ground 5 and dismissed its appeal on Ground 3. I would refuse permission to appeal on Ground 4. However, ultimately, interesting though the arguments are on the Bank’s other Grounds and on Venice’s appeal, they are academic because I have decided that the Banks’ appeal succeeds on Grounds 1 and 2.

I agree.

I also agree.