Dexia Crediop S.P.A. v Comune Di Prato

The claimant, Dexia Crediop SpA (“Dexia”), is an Italian bank. The defendant, Comune di Prato (“Prato”), is an Italian local authority with responsibility for the municipality of Prato in Tuscany.

In the spring of 2002 Dexia applied, as one of a number of tenderers, to become Prato’s adviser on debt restructuring and interest rate swaps. Debt restructuring, which may involve merely varying the terms of existing debt or may involve extinguishing debt and replacing it with new debt, is as old as debt itself. Interest rate swaps are a modern development. They fall within the type of financial instrument known as “derivatives”: contracts that derive their value from the performance of an underlying entity. Market participants often abbreviate the term “interest rate swap” to “IRS”.

From 1985 onwards ISDA has produced standard terms, including definitions, for use in the swaps market. ISDA is an association of market participants which was incorporated under the name International Swap Dealers Association, Inc. and which later changed its name to International Swaps and Derivatives Association, Inc. From 1987 onwards it has published standard form master agreements which, once entered into between two parties, enable those parties to make future swaps simply by agreeing upon the particular commercial terms of the swap in question.

Dexia’s application was successful: by the autumn of 2002 Dexia was effectively acting as Prato’s adviser in this regard. Prato wrote to Dexia on 25 November 2002, formally appointing it to act as Prato’s adviser, and Dexia formally accepted the appointment by letter dated 28 November 2002. I shall refer to the agreement thus created as “the 2002 advisory agreement”. Both sides agree that the 2002 advisory agreement was validly made. In this judgment I describe and determine issues between the parties as to the legal effect of what was done and not done during the course of the 2002 advisory agreement.

When writing to Dexia on 25 November 2002 Prato also sent a document that it had signed and sealed that day. This was a proposed master agreement based on the ISDA 1992 Multicurrency – Cross Border form, and including a schedule supplementing and amending the form, among other things by incorporating the 2000 ISDA Definitions and the 1998 ISDA FX and Currency Option Definitions. The document set out terms which were to govern interest rate swap transactions between Dexia and Prato, each transaction to be evidenced by a confirmation. This document was signed on behalf of Dexia on 29 November 2002. I shall refer to the agreement thus created as the “master agreement”.

The master agreement stated that it and each confirmation together constituted a single “Agreement”. Part 4(h) of the schedule to the master agreement stated that the Agreement was governed by and to be construed in accordance with the laws of England. By section 13(b) of the master agreement this had the consequence that each party irrevocably submitted to the jurisdiction of the English courts.

Subsequent interest rate swap transactions (“the swaps”) were entered into, each of them evidenced by a confirmation and said to form part of the Agreement contemplated by the master agreement. As explained below, the master agreement and the swaps are now said by Prato to be invalid or not to be enforceable against Prato.

During the period of the 2002 advisory agreement Prato’s public debt was restructured, first in 2004 (“the 2004 bond restructuring”), and second in 2006 (“the 2006 bond restructuring”). As part of the 2004 bond restructuring, numerous fixed interest loans initially denominated in lire, which by that time had become denominated in euros, were repaid. Also as part of the 2004 bond restructuring, new bonds were issued in 2004 (“the 2004 bonds”). The 2004 bonds comprised a first tranche of floating rate notes issued on 30 November 2004 in the sum of €27,870,000 (“the November 2004 bonds”) and a second tranche of floating rate notes issued on 29 December 2004 in the sum of €37,553,000 (“the December 2004 bonds”). The 2006 bond restructuring extended the maturity of the 2004 bonds.

As to the master agreement and the swaps, Prato admits that documents evidencing them were signed, but contests their validity or enforceability. I said earlier that the master agreement was “created” and that the swaps “were entered into”. In this judgment, unless the context otherwise requires, references to the master agreement and the swaps, and to them being created, authorised, approved or entered into, are without prejudice to the questions whether they were valid or enforceable. Similarly I use the language of rights and entitlements, and duties, liabilities and obligations, under the master agreement and the swaps without prejudice to those questions.

The swaps comprised:

on 4 December 2002, an interest rate swap with an initial notional sum of €83,824,626.88 (“swap 1”);

on 6 August 2003, an interest rate swap with an initial notional sum of €113,105,592.42 (“swap 2”), which terminated swap 1;

also on 6 August 2003, an interest rate swap with an initial notional sum of €13,055,932.44 (“swap 3”);

on 30 December 2004, an interest rate swap with an initial notional sum of €27,870,000 (“swap 4”) and an interest rate swap with an initial notional sum of €37,553,000 (“swap 5”); swaps 4 and 5 together terminated swap 2;

on 29 June 2006, an interest rate swap with an initial notional sum of €67,524,044.17 (“swap 6”), which terminated swaps 3, 4 and 5.

All obligations under swaps 1 to 5 were performed by each side as and when those obligations fell due. Netting off of payment obligations under those swaps almost always resulted in a payment by Dexia to Prato. The same was true in relation to swap 6 until 30 June 2009, when Prato started to become liable to make payments to Dexia. On 13 December 2010, Prato wrote to Dexia stating an intention to commence administrative self-redress procedures in relation to swap 6.

Obligations under swap 6 falling due on and from 31 December 2010 have not been met by Prato. In that regard Prato relied initially on self-redress procedures adopted in late December 2010, and subsequently relied additionally on various other defences. On the first day of the present trial Prato noted that its entitlement to rely on the self-redress procedures, while pleaded by way of defence in the present proceedings, had been litigated between Prato and Dexia in the Administrative Court in Italy and had been resolved by that court against Prato. An appeal by Prato is pending in Italy. Nevertheless, Prato said that it acknowledged that the English court would not at the present trial come to a decision contrary to that of the Administrative Court in Italy, and that for this reason it would not rely on the self-redress procedures at the trial.

In this judgment, unless the context otherwise requires, references to the self-redress procedures, and to them being authorised, approved, exercised, undertaken or entered into, are without prejudice to the questions whether they were valid or could be relied upon by Prato. Similarly I use the language of nullifying earlier resolutions or decisions, and similar language, without prejudice to those questions.

In the present proceedings Dexia’s substantive monetary claim is for sums it says should have been, but have not been, paid by Prato under swap 6 on and from 31 December 2010, pleaded as €6,504,878.35 as at 31 December 2013. Dexia also seeks declaratory relief as to the effect, validity and enforceability of swap 6. If swap 6 is invalid or unenforceable, Dexia seeks an alternative declaration that Prato is bound by the terms of swaps 3, 4 and 5 (on the footing that in such circumstances they would not have been terminated pursuant to swap 6).

Prato says that it is not bound by swap 6, nor by swaps 1 to 5. It says alternatively that obligations arising under the swaps cannot be enforced against it. Reliance is placed on 5 defences:

the swaps were null and void in English law by reason of Prato’s lack of capacity to enter them under Italian law (“the capacity defence”);

Prato’s alleged obligations under the swaps are unenforceable in English law in circumstances where enforcing them would require Prato to act illegally under Italian law in Italy (“the illegality defence”);

Prato’s alleged obligations under the swaps are unenforceable by reason of mandatory rules of Italian law, which must be given effect under the Rome Convention, Article 3(3) (“the mandatory rules defence”);

Prato has under English law rescinded swap 6 by reason of Dexia’s actionable misrepresentation, or alternatively is entitled to rescind swap 6 for this reason (“the misrepresentation defence”);

in any event, if found liable to pay any sum to Dexia, Prato will seek to set off that liability against the sums it counterclaims (“the set off defence”).

In its counterclaim Prato seeks substantive relief in the form of rescission of swap 6, restitution of sums paid under swap 6, and damages in respect of all the swaps. Prato describes its substantive counterclaims in broad terms in this way:

As regards sums it paid on and after 30 June 2009, on the basis that swap 6 is invalid and/or unenforceable, Prato seeks restitution under Italian law, alternatively English law (“the restitution counterclaim”);

Prato claims damages in respect of Dexia’s breaches of the Italian financial services regulatory regime (“the regulatory counterclaim”);

Prato claims damages in respect of Dexia’s breaches of the 2002 advisory agreement under Italian law (“the advisory counterclaim”);

Prato claims damages in respect of Dexia’s alleged misrepresentations under Italian law (“the misrepresentation counterclaim”).

Dexia’s replies to these defences and Dexia’s defences to these counterclaims include assertions, both of them denied by Prato, that:

a contractual estoppel bars Prato from raising certain points (“the estoppel reply”); and

it is entitled to rely upon a release (“the advisory release”) from any liability relating to its performance as Prato’s advisor.

In the next three parts of this section I describe the written evidence relied on by the parties, the course of the trial, and market concepts concerning types of interest rate swap, “mark to market”, and “hidden” or “implicit” costs.

I then describe in section B key features of the background and history, and in section C some general aspects of Italian law. In section D defences asserting contravention of Italian local government law are examined. In section E defences asserting contravention of Italian financial services and civil law are examined. Section F is concerned with the immediate consequences of my conclusion on the main provision of Italian financial services law relied upon. Section G explains why this gives rise to a need for further submissions in relation to Prato’s counterclaim and Dexia’s alternative claims. Section H summarises my conclusions.

Each side relied on two witnesses of fact and on three expert witnesses. A summary of their written evidence appears as Annex II to Prato’s written closing submissions. I transpose it here, with minor revisions.

The first factual witness relied upon by Dexia was Mr Riccardo Sommavilla. His witness statement (“Sommavilla 1”) was dated 28 November 2013. Mr Sommavilla was Dexia’s Client Relationship Manager with Prato throughout the duration of the 2002 advisory agreement, and thus at the time when the swaps were entered into.

The second factual witness relied upon by Dexia was Mr Samir Belarbi. His witness statements were dated 25 November 2013 (“Belarbi 1”), 21 March 2014 (“Belarbi 2”), 13 May 2014 (“Belarbi 3”) and 29 May 2014 (“Belarbi 4”). Mr Belarbi was Head of Hedging Solutions (based in Rome) during Prato’s entry into swaps 2 and 3, and then Head of the Debt Management Desk in Dexia’s public finance division (also based in Rome), in which role he was involved in Prato’s entry into swaps 4, 5 and 6.

The first factual witness relied upon by Prato was Mr Davide Zenti. His witness statements were dated 22 November 2013 (“Zenti 1”) and 9 May 2014 (“Zenti 2”). Mr Zenti has been Prato’s Manager of Financial Resources since 1 January 2011.

The second factual witness relied upon by Prato was Ms Graziella De Castelli. She is the holder of a doctorate, but is referred to in Prato’s written closing submissions as “Ms De Castelli”. In this judgment I shall do the same. Her witness statement (“De Castelli 1”) was dated 25 November 2013. Ms De Castelli was Prato’s Manager of Financial Resources between 1996 and 2007. She was Prato’s primary point of contact with Dexia.

As to Italian administrative and local government finance law:

Dexia relied on expert reports from Professor Giulio Napolitano, dated 21 March 2014 (“Napolitano 1”), 13 May 2014 (“Napolitano 2”) and 13 June 2014 (“Napolitano 3”). Professor Napolitano is Professor of Administrative Law at the University of Roma Tre;

Prato relied on expert reports from Professor Salvatore Dettori, dated 21 March 2014 (“Dettori 1”), 12 May 2014 (“Dettori 2”) and 16 June 2014 (“Dettori 3”). Professor Dettori is Professor of Administrative Law at the University of Teramo;

Professor Napolitano and Professor Dettori agreed a joint memorandum dated 17 April 2014 (“the Administrative Joint Memorandum”).

As to Italian financial regulatory and civil law:

Dexia relied on expert reports from Professor Aurelio Gentili, dated 21 March 2014 (“Gentili 1”), 13 May 2014 (“Gentili 2”), 20 May 2014 (“Gentili 3”) and 13 June 2014 (“Gentili 4”). Professor Gentili is a professor of Italian private law at the University of Rome III;

Prato relied on expert reports from Professor Antonella Sciarrone Alibrandi, dated 20 March 2014 (“Sciarrone Alibrandi 1”), 12 May 2014 (“Sciarrone Alibrandi 2”), 16 June 2014 (“Sciarrone Alibrandi 3”) and 25 June 2014 (“Sciarrone Alibrandi 4”). Professor Sciarrone Alibrandi is Vice-Rector and Professor of Banking Law and Financial Markets Law and Professor of Italian Private Law at the Università Cattolica del Sacro Cuore, Milan;

Professor Gentili and Professor Sciarrone Alibrandi agreed a joint memorandum dated 23 April 2014 (“the Civil Joint Memorandum”).

As to valuation of the swaps:

Dexia relied on expert reports from Mr Pawan Malik dated 21 March 2014 (“Malik 1”), 13 May 2014 (“Malik 2”) and 20 May 2014 (“Malik 3”). Mr Malik currently works for Navigant Consulting (Europe) Ltd;

Prato relied on expert reports from Dr Enzo Faro dated 20 March 2014 (“Faro 1”) and 12 May 2014 (“Faro 2”). Dr Faro is a Senior Financial Consultant at Finance Active Italia Srl;

Mr Malik and Dr Faro agreed a joint memorandum dated 16 April 2014 (“the Valuation Joint Memorandum”).

At the trial Mr Richard Handyside QC and Mr Rupert Allen, instructed by Allen & Overy, appeared for Dexia. Mr Jonathan Davies-Jones QC and Mr Christopher Burdin, instructed by Seddons, appeared for Prato. I am grateful to the legal teams on both sides for the considerable assistance given to me before and during the trial.

The trial occupied 18 hearing days. On day 1 leading counsel for each side made opening speeches. In the course of those speeches a question arose as to whether, in relation to misrepresentation, it was permissible for Prato to advance factual assertions which went beyond its pleaded case that the alleged representations were implicit in a failure to comply with duties of disclosure. I concluded that this was not permissible and gave a ruling to that effect.

Oral evidence was then heard on days 2 to 16 inclusive. A summary of that evidence appears in Prato’s written closing submissions. I transpose it here, with minor revisions:

Mr Sommavilla gave oral evidence on Day 2; Mr Belarbi gave oral evidence on Days 2 and 3;

Mr Zenti gave oral evidence on Day 3; Ms De Castelli gave oral evidence on Days 8 and 9;

Professor Napolitano gave oral evidence on Days 4 and 5; Professor Dettori gave oral evidence on Days 6 and 7;

Professor Gentili gave oral evidence on Days 10 and 11; Professor Sciarrone Alibrandi gave oral evidence on Days 12, 13 and 14;

Mr Malik gave oral evidence on Day 15; Dr Faro gave oral evidence on Day 16.

Day 17 was devoted to Mr Handyside’s oral closing submissions on behalf of Dexia. On day 18 Mr Davies-Jones made oral closing submissions for Prato, and Mr Handyside made oral reply submissions.

There are many different types of swap contract. In section A4.2 I describe some of those which feature in the present case. In section A4.3 I discuss a calculation commonly used in financial markets. It is referred to as “mark to market” or MTM. In section A4.4 I note aspects of what has been said about “hidden” or “implicit” costs.

The swaps in the present case all derive their value, at least in part, from the performance of rates of interest which are determined daily for certain types of loan as the rate at which banks are prepared to lend to each other. With the arrival of the euro market participants established mechanisms for determining the Euro Interbank Offered Rate (“Euribor”) for loans made between banks and denominated in euros. Euribor is determined for loans of particular durations. A shorthand is commonly used under which rates for loans with a maturity of n months are known as “nM Euribor” or “EURnM”.

A swap involving Euribor may be as simple as a promise by one side (“the fixed leg”) to pay a fixed rate of interest annually on a notional sum in exchange for a promise by the other side (“the floating leg”) to pay 1M Euribor monthly on the same notional sum. I shall refer to a swap of this kind as a “simple interest rate swap” or “simple IRS”.

What are commonly described as interest rate swaps may be more complex than a simple IRS. In the examples below I refer to a party which is to receive a particular advantage as “the buyer” and the party making a promise to provide that advantage as “the seller”. The consideration provided by the buyer to the seller in return for the seller’s promise may take various forms, but is commonly described as “the premium”. Using terminology adopted generally in the swaps market, and assuming that the swap derives its value from 1M Euribor:

in an interest rate cap a buyer and a seller agree a cap rate (also referred to as a ceiling, as a strike price or as a threshold or barrier rate) for 1M Euribor; if 1M Euribor exceeds the cap rate for a particular month the seller makes a payment to the buyer, usually calculated by applying the amount of the excess to the notional sum, thereby protecting the buyer from the danger that the buyer in relation to that sum may have to pay to its own counterparty 1M Euribor going beyond the cap rate;

in an interest rate floor a buyer and a seller agree a floor rate (also referred to as a strike price or as a threshold or barrier rate) for 1M Euribor; if the floor rate exceeds 1M Euribor for a particular month the seller makes a payment to the buyer, usually calculated by applying the amount of the excess to the notional sum, thereby protecting the buyer from the danger that the buyer in relation to that sum may receive from its own counterparty 1M Euribor lower than the floor rate;

an interest rate collar is a financial instrument that involves both an interest rate cap and an interest rate floor. The party who is seeking the collar (“the end user”) may be looking to buy from a counterparty an interest rate cap with a defined ceiling, and may seek to reduce the premium otherwise payable by selling an interest rate floor to the same counterparty. In these circumstances the collar protects the end user against the danger that 1M Euribor may exceed the collar’s ceiling, but sacrifices profits it would otherwise gain when 1M Euribor drops below the floor. Alternatively, it may be that the end user is looking to buy from a counterparty an interest rate floor, and may seek to reduce the premium otherwise payable by selling an interest rate cap to the same counterparty. In these circumstances the collar protects the end user against the danger that the end user may receive from the counterparty 1M Euribor lower than the floor rate, but sacrifices profits it would otherwise gain when 1M Euribor rises above the ceiling.

Market participants use a type of calculation known as “mark to market”, commonly abbreviated to “MTM”. Issues arise in the present case as to whether the MTM of the swaps had any, and if so what, relevance to their validity or enforceability. I discuss those issues in subsequent sections. For present purposes I note that Dr Faro and Mr Malik agree that MTM is generally understood in its simplest form to mean the present value of the expected cash-flows, calculated according to a series of generally accepted conventions.

How this works can be seen by starting from a theoretical base in relation to the two legs of the simple IRS postulated in section A4.2. The present value of future cash flows is obtained by discounting them at market rates. If, on inception, each rate is the same as the current market discount rate then the swap is theoretically at par – each leg has a present value of zero because the promised rates equate to what can, in theory at least, be obtained in the market. In this theoretical example the MTM on inception will be zero for both sides, because the present value of what will have to be paid by the fixed leg is neither higher nor lower than the present value of what will have to be paid by the floating leg.

However if the annual discount rate in the market differs from the fixed rate under the swap, then the present value of the fixed rate leg will no longer be zero. Mr Malik gives an example where the swap is for a period of a year with a notional sum of €100. Under a notional loan of €100 the notional repayment by a fixed rate borrower at the end of the year will be €105, comprising the principal of €100 and interest of €5. If the annual discount rate goes up from 5% to 6%, then the party paying the fixed rate will be paying in a year’s time interest of €5 while the market would now be willing to promise to pay 6% at the end of a year. That entitlement to pay less than the market rate, when applied to a notional sum of €100, gives the fixed rate leg a positive present value of €0.95 – because at the rate of 6% that the market would give, it would be necessary only to invest €99.05 in order be entitled in a year’s time to a repayment of €105. Making a further theoretical assumption that 1M Euribor is unchanged, the floating leg would continue to have a value of zero. The result will be thus be that this simple IRS for a term of a year on a notional sum of €100 will have a positive revised MTM for the fixed rate payer of +€0.95, and a negative revised MTM for the floating rate payer of -€0.95.

More commonly a transaction will be more complex, involving floors or caps or other components. If so, the MTM of the transaction will be the sum of the MTM of each component.

In practice there will be numerous other complexities to take account of. One such will be the spread between bid and offer rates. In relation to any financial product traded between banks, what a bank will be prepared to pay will be less than what it will offer to receive. This difference is the spread charged by the bank for acting as market maker. One way of taking account of it is to calculate MTM on the basis of a mid-market rate half way between the two.

Aspects of this case also involve allegations of what have been described as “hidden costs” or “implicit costs”. Unless the context otherwise requires, the present judgment uses these expressions in the sense described by Dr Faro in paragraph 21 of Faro 1:

For reasons given in section D below, I conclude that the swaps involved no contravention of Italian local government finance law. Defences mounted by Prato in that regard, including assertions that such contraventions deprived Prato of capacity to enter into the swaps, accordingly fail.

For reasons given in section E below, however, I conclude that, under English conflict of law principles, mandatory rules of Italian financial services law applied to the swaps. Also in section E below, after considering expert evidence of Italian law, I conclude that under those rules Prato was entitled to a 7 day cooling off period after entering into each swap. During the 7 day period Prato had a right of withdrawal. It follows from the relevant mandatory rule that Dexia’s forms were required to state that right of withdrawal. The relevant mandatory rule sets out the consequence for failure to do so: such a failure results in “the related contracts being null and void, with only the client having the right to enforce this provision”.

Because Dexia’s forms did not state the right of withdrawal, so far as Dexia’s main claim is concerned the relevant mandatory rules have the consequence that the contract in question (here, swap 6) was null and void, with only Prato having the right to enforce that provision. In its defence in these proceedings Prato has enforced that provision. It follows that Dexia’s main claim, which is based on swap 6, necessarily fails.

My conclusion in section E makes it necessary to consider whether Prato is entitled to relief under its restitutionary counterclaim in relation to money that it paid Dexia pursuant to swap 6. It may also be necessary to consider whether that conclusion may entitle Dexia to advance alternative claims. It seemed to me that, before proceeding to consider the restitutionary counterclaim and alternative claims, I needed additional submissions from the parties as to how their reasoning applied specifically to the mandatory rules in question. When informing the parties of this, I added that I also needed submissions on how my conclusion on the particular mandatory rules involved, and my eventual conclusion on the restitutionary counterclaim and alternative claims, would affect the remaining counterclaims.

In my view it is important that my conclusions and reasoning, as thus far made known to the parties, should be set out in a public judgment. This will enable me to hear oral submissions in public on appropriate directions to deal with my requests for additional submissions and on such other consequential orders as may be appropriate at this stage. It is for that reason that I now deliver the present judgment.

Much of the background and history is not in dispute. In this section I summarise the main features. For this purpose I draw upon historical accounts given by the two sides.

The constitutional structure of Prato is that it consists of a body of elected representatives, the elected Consiglio Comunale (“the Council”). Responsibility for executive functions lies with the Giunta (“the Executive”). As noted earlier, Ms De Castelli was Prato’s Manager of Financial Resources. Other general background concerning Italian legal structures is set out in section C below.

Prato’s public finances were under considerable pressure from 1996 onwards. By the late 1990s derivative products, including those used for interest rate hedging, had become common features of European financial markets. As at 31 December 2001, Prato’s total borrowing stood at around €177 million: 88% at a fixed interest rate (all rates at or above 5.5%), the remaining 12% at a variable rate. The majority of Prato’s fixed rate borrowing (68%) was with Cassa Depositi e Prestiti (“CDP”). Prato was concerned at the level of its debt burden. Proposals for the use of interest rate swaps as an instrument of debt management had been made to Prato by Dexia in the summer and autumn of 2001. However, Prato itself had no experience of interest rate hedging or other derivative products. It sought assistance from external experts, who recommended that it should seek an adviser on strategies, including interest rate swaps, for optimising the cost of its debt.

On 24 April 2002, Prato issued a tender notice for the appointment of an adviser. The tender notice stated:

By decision 1453 of 22 May 2002, Prato resolved to set up a technical committee to consider and evaluate the tenders. The technical committee was made up of Ms De Castelli (chair), Ms Rappuoli and Ms Belli from within Prato, along with external members. The external members were the previous external experts, who had drafted the tender notice. They comprised two professors of economics, Professor Nigro and Professor Bompani, and a former senior banker, Mr D’Agliana.

The technical committee considered proposals from 7 tenderers. Among them was a presentation submitted by Dexia by letter dated 16 May 2002. Dexia said, among other things, that:

its proposals were “meant to reduce the borrowing cost and the exposure to the interest rate risk”;

it was “preferable to include derivative transactions as instruments for pure hedging against interest rate risk”;

it had “carefully examined” the legal position, “making it possible for its clients to obtain the maximum possible benefits over time”;

its proposal was based on Dexia’s “constant involvement”, providing continual assistance “within the scope of a long-term partnership”;

its services would include (in bold typeface) “monitoring of the market situation…with commitment to provide on a periodic basis and nevertheless upon [Prato’s] request … the performance of the transaction in mark-to-market terms”; and

there would be “no additional fee or charge” if Dexia were appointed as Prato’s advisor.

The external members advised that Prato should look at short-term transactions of no more than 5 years. The tenderers were therefore asked to re-submit proposals in this regard, complying with certain stated conditions so as to enable easier comparison. Accordingly, Dexia put together a revised proposal on 17 June 2002, including possible structures with 5, 4 and 3-year terms.

The technical committee met again on 21 June 2002, and resolved to appoint Professor Nigro and Mr D’Agliana to study the proposals that had been received. Professor Nigro and Mr D’Agliana reported back to the technical committee in a meeting on 10 July 2002. The technical committee recommended the appointment of Dexia as Prato’s advisor, because its offer was considered to be the most advantageous.

Prato’s Executive approved the technical committee’s report by resolution number 524 on 25 July 2002, and decided to appoint Dexia as advisor: “for the definition of debt transformation strategies and for assistance, advice and management in relation to the interest rate swap transaction”. The resolution also referred to Prato’s requirement that the proposed structures involved minimal and contained risk. Ms De Castelli was given responsibility for finalising the appointment.

The technical committee was at this point defunct. It did not meet again and did not advise Prato further.

A resolution number 2331, dated 29 July 2002, stated that Dexia was appointed as Prato’s financial advisor.

By resolution number 140 of 3 October 2002, the Council approved Prato’s “debt restructuring policy” which included the following:

In October and November 2002 revised proposals for what would become swap 1 were discussed by Mr Sommavilla and Ms De Castelli. As noted in section A above, by the end of November 2002 each of Prato and Dexia had signed the 2002 advisory agreement and the master agreement. On 4 December 2002 the terms for swap 1 were set out in Prato’s decision 3842/2002 authorising an irrevocable proposal by Prato to Dexia in those terms. Also on 4 December 2002 that irrevocable proposal was duly notified by Ms Castelli to Dexia and duly accepted by Dexia.

Swap 1 was for a period of 10 years from 30 June 2002 to 30 June 2012 and was based upon an initial notional sum of €83,824,626.88 amortising over that period. It derived its value from the performance of 6M Euribor. So long as 6M Euribor was 5.70% or less, Prato promised to pay Dexia interest on the notional sum at a rate which was fixed at 5.92% per annum. If 6M Euribor exceeded 5.70%, however, Prato promised to pay Dexia interest on the notional sum at 6M Euribor plus 1.90% per annum. In exchange for these promises, Dexia promised to pay Prato interest on the notional sum at a fixed rate of 6.36%. All this can be expressed more shortly by saying that Prato received a net differential of 0.44% from Dexia provided that 6M Euribor was no higher than a threshold of 5.70%, but once 6M Euribor exceeded the threshold Prato would have to make a net payment of 1.24% plus the amount by which the threshold was exceeded. Thus while 6M Euribor did not exceed 5.70% Prato would, so long as Dexia remained solvent, receive periodical payments that it could use to finance part of its fixed-interest payments to lenders, but as soon as 6M Euribor exceeded 5.70% it would find itself not only liable to make those fixed interest payments to lenders but also to make payments of something in excess of 1.24% to Dexia.

In addition under swap 1 Dexia paid Prato an upfront premium of €165,000.

On 6 December 2002 Dexia entered into a back to back agreement with an associated company based in France. This meant that it hedged its interest rate risk. So long as 6M Euribor did not exceed the threshold, Dexia received from its hedge counterparty a net payment of 0.12% more than the net payment it would have to make to Prato. The practical effect was that in exchange for this margin, it bore a credit risk in relation to its hedge counterparty. Once 6M Euribor exceeded the threshold Dexia would receive from Prato 0.85% more than the net payment it would have to make to its hedge counterparty. The practical effect was that in exchange for this margin, it bore a credit risk in relation to Prato.

In addition under the back to back agreement Dexia was entitled to receive from its hedge counterparty the same premium as it was bound to pay to Prato under swap 1.

In May 2003 Dexia prepared a document for Prato described in translation as a “Hypothesis for active debt management”. After describing Dexia’s active debt management services, it said, among other things, that the market expected low interest rates in the short term, and predicted possible increases in the medium term in view of a recovery in the economic cycle. It also set out a hypothesis for “the conversion of three portions of [Prato’s] debt” involving 3 possible transactions, said to lead to Prato obtaining “benefits, certain or potential, in the first years”.

On 22 July 2003, Mr Sommavilla emailed Ms De Castelli a further proposal from Dexia, dated 21 July 2003. The proposal stated that the market expected protracted stability of short-term interest rates at current levels, with a possible increase in the medium term. It proposed transactions that were precursors to swaps 2 and 3.

On 4 August 2003, Prato passed resolution 2407, recording that “in the framework of its constant monitoring of the transactions carried out by [Prato]”, Dexia had proposed that the existing transaction be unwound and replaced with what would become swaps 2 and 3 concerning Prato’s fixed rate and variable rate borrowing respectively. Prato resolved to accept Dexia’s proposals.

Also on 4 August 2003, Dexia emailed Prato drafts of two “irrevocable proposals”. These were signed by Ms De Castelli and returned to Dexia by fax on the same day. Dexia accepted the irrevocable proposals on 6 August 2003. The irrevocable proposals and the acceptance constitute the confirmations for swaps 2 and 3.

Under swap 2, Prato received interest from Dexia at a fixed rate of 5.54% on an initial notional sum of €113,105,592.42, amortising until maturity on 30 June 2013. Prato paid Dexia interest at:

a floating rate of 6M Euribor plus 1.69% if 6M Euribor was greater than the upper barrier (30 June 2003 to 2005: 4.40%; 2005 to 2007: 5.00%; 2007 to 2013: 5.70%);

a fixed rate (30 June 2003 to 2005: 4.85%; 2005 to 2013: 5.08%) if 6M Euribor was between the upper barrier and the lower barrier; or

if 6M Euribor was lower than the lower barrier (30 June 2003 to 2007: 1.79%; 2007 to 2013: 1.89%), a fixed rate (the rate of the lower barrier) plus the difference between the lower barrier rate and 6M Euribor.

Swap 2 also terminated swap 1.

The underlying debt corresponding to the notional sum for swap 2 had been borrowed by Prato at a fixed rate. The fixed rate received by Prato under swap 2 (5.54%) corresponded to the average rate charged on the underlying borrowing. The best possible result for Prato under the transaction would be a net gain of 0.69% between 30 June 2003 to 2005 and 0.46% thereafter, which would occur if 6M Euribor fixed within the corridor defined by the upper and lower barriers: namely, the difference between 4.85% (or, after 30 June 2005, 5.08%) and its underlying cost of borrowing (5.54%).

If 6M Euribor fixed below the lower barrier, Prato’s position would deteriorate depending on how low it fixed, because Prato would pay the lower barrier rate plus the difference between it and 6M Euribor. Conversely, once 6M Euribor went above the upper barrier, there was no limit to Prato’s potential liabilities because Prato would have to pay the floating rate plus a 1.69% spread.

Under swap 3, Prato received interest from Dexia at a floating rate of 6M Euribor flat on an initial notional sum of €13,055,932.44, amortising until maturity on 30 June 2013. Prato paid Dexia interest at:

a fixed rate (30 June 2003 to 2005: 1.55%; 2005 to 2013: 4.01%) if 6M Euribor was lower than the barrier (30 June 2003 to 2005: 1.89%; 2005 to 2013: 4.90%); or

if 6M Euribor was greater than the barrier, a floating rate of 6M Euribor minus a spread (30 June 2003 to 2005: 0.40%; 2005 to 2013: 0.05%).

The notional sum under swap 3 corresponded to actual borrowing by Prato at floating interest rates indexed to Euribor. If 6M Euribor fixed below the barrier, Prato would benefit from lower costs on its underlying liabilities, but would be required to pay the fixed rate under the swap, which would become a net loss if 6M Euribor fell below the fixed rate (e.g. if, after 2005, 6M Euribor fell below 4.01%, and therefore below the barrier at 4.90%, Prato’s net loss would be the difference between the 6M Euribor flat it received from Dexia and the fixed rate it was required to pay at 4.01%).

If 6M Euribor fixed above the barrier, Prato would pay a higher rate on its underlying floating rate borrowing, whilst receiving a net benefit under the swap (i.e. the difference between the 6M Euribor flat it received from Dexia and 6M Euribor minus the spread: 0.4% for the first two years and then 0.05%).

In order to hedge its interest rate risk on swap 2, Dexia entered into a back to back swap with Deutsche Bank on 5 August 2003. The fixed rate received by Dexia under that agreement was 5.8175% (compared with the 5.54% received by Prato). Deutsche Bank additionally paid Dexia an upfront payment in the sum of €204,000.

In order to hedge its interest rate risk on swap 3 Dexia entered into a back to back swap with Morgan Stanley on 5 August 2003. Under that trade, Morgan Stanley agreed to pay Dexia 6M Euribor plus 0.0825% (compared with the 6M Euribor flat paid to Prato).

Discussions took place between Ms de Castelli and Mr Sommavilla in about September 2004 concerning, among other things, a need for Prato to make savings that year of around €1.5m.

Dexia formulated a proposal, dated 23 September 2004, entitled “Proposal for the active management of [Prato’s] debt”. It said that “Long-term rates have decreased” and that interest rates should not be “subject to substantial changes”. Dexia suggested that replacing fixed rate debt with floating rate debt, along with a corresponding replacement for swap 2, would lead to an “overall discounted saving” to Prato of €2.535m.

On 23 September 2004 Prato passed resolution 178. This approved the issue of a bond in the nominal amount of €6,148,999 amortising over a 15-year term at an annual rate of 6M Euribor plus a 0.093% margin.

There were then further revised proposals submitted by Dexia to Prato. On 18 October 2004 Dexia emailed Prato a draft of what should be included in a resolution authorising the proposed bond issue. Prato duly passed resolution 669 on 19 October 2004, by which it resolved to repay specified fixed rate loans and approved the refinancing scheme proposed by Dexia. It also instructed Ms De Castelli to bring the swaps into line with the new borrowing position, “in order to reduce the risks associated with fluctuations in the interest rate applied to [Prato’s] borrowing”.

Following receipt of drafts sent by Dexia, on 28 October 2004 Prato passed resolution 214. Prato resolved to approve the issue of a variable rate bond in the sum of €67,229,000 over a 15-year term. “Parallel to the bond issue”, a swap transaction was to be carried out “to reduce the risks associated with fluctuations of the interest rates applicable to [Prato’s] borrowing…and taking into consideration that the transaction may not under any circumstances be carried out for speculative purposes”.

On 3 November 2004 Dexia emailed Ms De Castelli requests to Banca d’Italia relating to Prato’s proposed bond issue. On 26 November 2004 Dexia emailed Ms De Castelli a draft resolution, which was said to be subject to the approval of Banca d’Italia. This explained that it was not possible within the existing timeframe to repay certain borrowing from CDP. The proposal was therefore a bond issue split into two tranches. On 30 November 2004 Prato passed resolution 3471, which resolved to proceed on the basis suggested by Dexia.

On 30 November 2004, Prato issued the November 2004 bonds in the sum of €27,870,000, amortising over a maturing period of 15 years, indexed to 6M Euribor plus 0.15%.

On 13 December 2004 Dexia emailed Ms De Castelli “an update of the proposal for extinguishment and refinancing of the [CDP] loans and related swap transactions”, dated 13 December 2004. The email also attached a draft council resolution relating to the issue of the December 2004 bonds.

On 14 December 2004 Dexia emailed Ms De Castelli a further update, proposing a two-swap scheme for restructuring swap 2, reflecting the two tranches of the 2004 bonds. On 23 December 2004, the Council resolved to proceed with the second bond issue, in the sum of €37,553,000. It was noted that early repayment penalties applied in the sum of €2,802,434.77. It was further stated that “in parallel with the bond issue…the existing swap agreement [swap 2] will be modified through the execution of two swap transactions”.

On 29 December 2004, Prato issued the December 2004 bonds in the sum of €37,553,000, amortising over a maturity period of 15 years, indexed to 6M Euribor plus 0.15%.

Dexia acquired the entirety of the 2004 bonds in private placements upon issue.

Meanwhile on 27 December 2004 Dexia emailed Ms De Castelli a “technical report” relating to the proposed derivatives. Dexia said that it had proposed “the most appropriate instrument – permitted by the regulations in force – in order to bring the cost of the debt to market levels”. It also said that the proposed scheme met the objective of “contain[ing] the cost of the debt without prejudice to [Prato’s] budget balances in the medium and long term”. The report set out tables suggesting a net advantage to Prato from the proposed transactions, noting further that they would “have a positive impact on [Prato’s] medium and long term budget balances”.

On 28 December 2004 Prato passed resolution 904, which approved Dexia’s “technical report” and authorised Ms De Castelli “to achieve the assigned strategic objective in relation to debt restructuring”.

On 29 December 2004 Prato passed resolution 3956, which resolved to unwind swap 2 and to submit irrevocable proposals in respect of the two new swaps proposed by Dexia. The irrevocable proposals were duly submitted, and were dated 29 December 2004. Dexia accepted them by fax dated 30 December 2004. On 31 December 2004, Dexia emailed Prato asking for it to fax a countersigned version of its letter of acceptance. The irrevocable proposals and the acceptance constitute the confirmations for swaps 4 and 5.

Under swap 4, Prato received interest from Dexia at a floating rate of 6M Euribor plus 0.15% on an initial notional sum of €27,870,000, amortising until maturity on 30 November 2019. Between 30 November 2004 and 2006, Prato paid Dexia interest at a floating rate of 6M Euribor flat. Between 30 November 2006 and 2019:

if 6M Euribor was greater than the upper barrier (30 November 2006 to 2008: 7.00%; 2008 to 2019: 7.90%), Prato paid Dexia interest at a fixed rate (upper barrier plus 0.09%, i.e. 7.09% or 7.99%);

if 6M Euribor was between the upper and lower barrier, a floating rate of 6M Euribor plus 0.09%; or

if 6M Euribor was lower than the lower barrier (30 November 2006 to 2008: 1.99%; 2008 to 2019: 4.10%), a fixed rate (lower barrier plus 0.09%, i.e. 2.08% or 4.19%).

In the first period (30 November 2004 to 2006), Prato was guaranteed a 0.15% net gain under the swap, which corresponded with the 0.15% margin on the First and Second Tranches of the bond issue. After 30 November 2006, the swap turned into a collar structure, under which Prato could not pay less than the floor (2.08% or 4.19%) or more than the cap (7.09% or 7.99%). If 6M Euribor fixed within the corridor, Prato would make a gain of 0.06% (the difference between the 0.15% margin paid by Dexia and the 0.09% margin paid by Prato). To benefit from the cap, 6M Euribor would need to rise above 7.00% or (between 2008 and 2019) 7.90%. If interest rates fell below the floor, the swap would lead to a loss for Prato, depending on the 6M Euribor rate. For example, if after 2008 6M Euribor fixed below 4.04% (i.e. 4.19% minus 0.15%), Prato would suffer a net loss according to the difference between the floor rate of 4.19% it paid (i.e. the 4.10% lower barrier plus 0.09%) and the floating rate of 6M Euribor plus 0.15% that it received.

The economic terms of swap 5 were broadly identical to those of swap 4. Under swap 5, Prato received interest from Dexia at a floating rate of 6M Euribor plus 0.15% on an initial notional sum of €37,553,000, amortising until maturity on 29 December 2019. Between 29 December 2004 and 2006: Prato paid Dexia interest at a floating rate of 6M Euribor flat. Between 29 December 2006 and 2019: if 6M Euribor was greater than the upper barrier, Prato paid Dexia interest at a fixed rate (upper barrier plus 0.09%); if 6M Euribor was between the upper and lower barrier, a floating rate of 6M Euribor plus 0.09%; or, if 6M Euribor was lower than the lower barrier, Prato paid a floor rate (lower barrier plus 0.09%).

The overall effect of swap 4 and swap 5 was that Prato would receive a net differential of 0.15% for the first two years. Thereafter, Prato would receive a net differential of 0.06% provided that 6M Euribor fixed between the cap and the floor, and would pay the cap rate or the floor rate (as appropriate) if 6M Euribor fixed outside the range. Prato’s exposure to interest rate risk outside the range was removed, and it would enjoy an effective reduction in the cost of borrowing under the 2004 bonds if 6M Euribor fixed above the floor. If 6M Euribor fixed above the floor, Dexia would (in effect) fund part of the interest on the 2004 bonds since Prato would be the net recipient of differentials under swap 4 and swap 5.

Dexia entered into back to back swaps in respect of these trades with Deutsche Bank on 30 December 2004. Dexia received from Deutsche Bank 6M Euribor plus 0.699% (compared with the 0.15% spread it paid to Prato).

Swaps 4 and 5 also terminated swap 2. At the time, and in the light of movements in expectations of interest rates, the value of swap 2 was positive to Prato.

In February 2006, Mr Sommavilla and Ms de Castelli discussed a problem that had arisen on swap 3: Prato had been unable to pay a negative differential that had fallen due in January 2006. Moreover, there would be a further negative differential due under swap 3 at the end of June 2006. Ms de Castelli added in oral evidence that Prato’s budget was under “particular strain” and that it had “enormous problems of budget balance”.

On 9 May 2006 Dexia sent a presentation to Prato with suggestions for extending the maturity of the 2004 bonds and restructuring of swap 3, swap 4 and swap 5 by a new swap.

On 9 June 2006, Mr Giampiero Poddighe of Dexia emailed Prato a revised presentation, dated 31 May 2006. On the basis of forward rates at 31 May 2006 it indicated that a revised swap structure could lead to net positive results for Prato up until 2010. This was on the footing that the floor rate for the longest period of the proposed swap was indicated as 4.69%.

On 12 June 2006, Dexia emailed Prato a draft decision to authorise the bond renegotiation. The draft decision would also authorise management to adopt:

The preamble set out in the draft included recitals a to d concerning the 2004 bond restructuring, and continued:

On 13 June 2006 Mr Poddighe emailed Ms Brilli a document titled “Management Scenarios of [Prato’s] debt”. This referred to the proposed collar structure, which was said to include:

On the same day, in an internal email, Mr Poddighe was asked whether Prato was a “qualified operator”. Mr Poddighe replied: “Qualified operator? [Prato] do not even know what interest rates are”.

On 14 June 2006, there was a meeting of Prato’s standing committee number 2, concerned with planning and organization. In attendance was Ms de Castelli, along with Mr Sommavilla and Mr Poddighe. It was explained that Prato wanted to reduce current expenditure and that the Council had adopted a strategy that saved €1.5 million over the previous 4 years and would release up to €7 million over the next 3 years, “easing pressure on the budget”. Mr Sommavilla said that Dexia had proposed solutions “to manage interest rate risk” and that “short and long rates were still close to an all-time low”.

By resolution 101 of 15 June 2006 the Council authorised the 2006 bond restructuring and the restructuring of interest rate swap agreements.

The terms of the November 2004 bonds and the December 2004 bonds were duly renegotiated in June 2006. In respect of the November 2004 bonds, from 30 May 2006 the changes resulted in a new maturity date (30 June 2026, instead of 2019), a different amortisation profile, and different fixing and payment dates (30 June and 31 December yearly). In respect of the December 2004 bonds, from 30 June 2006 the changes resulted in a new maturity date (30 June 2026, instead of 2019) and a different amortisation profile.

By decision 1691 of 28 June 2006 (“Prato’s swap 6 decision”), Prato resolved to enter into what was to become swap 6, and to cancel swaps 3, 4 and 5. A draft irrevocable proposal in that regard was emailed by Dexia to Prato that day and was signed by Prato later that day. Dexia accepted Prato’s irrevocable proposal by fax dated 29 June 2006. The irrevocable proposal and the acceptance constituted the confirmation for swap 6.

What Dexia promised to pay to Prato under swap 6 comprised:

between 30 June and 31 December 2006, a fixed rate of 3.063% in respect of a notional €25,083,278.70 and 3.240% in respect of a notional €33,798,075.53;

from 1 January 2007, interest at a floating rate of 6M Euribor flat on an initial notional sum of €67,524,044.17, amortising until maturity on 30 June 2026.

What Prato promised to pay to Dexia under swap 6 comprised interest on an initial notional sum of €67,524,044.17, amortising until maturity on 30 June 2026:

if 6M Euribor was greater than the upper barrier (30 June 2006 to 2008: 4.55%; 2008 to 2010: 5.55%; 2010 to 2026: 6.74%), at a fixed rate (upper barrier minus 0.05%);

if 6M Euribor was between the upper and lower barrier, at a floating rate of 6M Euribor minus 0.05%; or

if 6M Euribor was lower than the lower barrier (30 June 2006 to 2008: 2.95%; 2008 to 2010: 3.80%; 2010 to 2026: 4.80%), at a fixed rate (lower barrier minus 0.05%).

Once swap 6 was in place and had terminated swaps 3, 4 and 5, the overall effect produced a collar similar to that produced by the cap and floor in swaps 4 and 5. If 6M Euribor fixed within the corridor between the upper and lower barriers, Prato stood to make a net gain of 0.05%, this being the difference between the 6M Euribor flat it received from Dexia and the 6M Euribor minus 0.05% it paid. By contrast, Prato would pay the cap rate or the floor rate (as appropriate) if 6M Euribor fixed outside the range. However, as the corridor between the upper and lower barriers changed over time (4.55% to 2.95%; then 5.55% to 3.80%; then 6.74% to 4.80%), the floor rate increased, and thus the size of a potential disadvantage to Prato increased.

Dexia entered into a back to back swap with Barclays effective from 30 June 2006. Under that transaction Barclays agreed to pay Dexia 6M Euribor plus 0.568% (compared with the 6M Euribor flat paid by Dexia to Prato). That transaction was altered so that, after 31 December 2007, there would be no 0.568% margin, and instead Barclays paid Dexia €2.65 million.

Dexia sent Prato a document dated 17 October 2007, entitled “Evaluation of the Existing Swaps”. This put the market value to Prato of swap 6 on that date at minus €1.444 million.

A document dated 21 July 2008 sent by Dexia to Prato was entitled “Mark to Market Communication”. It stated the value of swap 6 to Prato as minus €1,320,500.

Dexia sent Prato a further document entitled “Mark to Market Communication” dated 3 March 2009. This stated the MTM value to Prato of swap 6 as minus €5,204,856 as at 30 January 2009.

Prato appointed a financial expert, Professor Cherubini, to review the swaps. In his report of 18 November 2009 he identified negative “loadings” (i.e. MTM values) at inception of each of the swaps.

By decision 2723 of 8 October 2010, Prato resolved to take legal action as part of its “administrative responsibility to avoid further prejudice, if not actual financial and capital losses for [Prato’s] budget”. Just under a month later, by decision 98 of 4 November 2010, Prato further resolved that the Executive should “take every action to protect the sound financial position of [Prato]” and to instigate “administrative proceedings for the purposes of exercising powers of self-protection”.

On 13 December 2010 Prato wrote to Dexia indicating that it intended to commence administrative self-redress procedures in relation to swap 6. There was a hearing at which Dexia protested at the short notice it had been given. Then, by resolution 4142 of 31 December 2010, Prato exercised its right of self-redress under Italian administrative law to annul (with retroactive effect) resolution 1691 of 28 June 2006, by which it had resolved to enter into swap 6, and to cease any future payments under swap 6.

On 6 April 2011 the Guardia di Finanza (Italian Finance Police) issued a preliminary report, making a number of adverse findings in relation to Dexia’s conduct in the sale of the swaps.

On 23 November 2011 Prato’s application to declare swap 6 invalid was dismissed by the Regional Administrative Court for Tuscany for want of jurisdiction. The court did not regard Prato’s swap 6 decision, which Prato claimed to have nullified, as an administrative act capable of falling within the self-redress jurisdiction; rather it regarded it as a private law act subject to the jurisdiction of the relevant civil court. On the same basis, the court rejected Dexia’s application challenging Prato’s decision to exercise its rights of self-redress and to cease payments under swap 6.

By resolution 30 of 19 April 2012, Prato further exercised its right of self-redress under Italian administrative law to annul in so far as they related to the swaps (with retroactive effect): resolution 101, which related to swap 6; resolution 140, which approved Prato’s use of derivative instruments; and resolution 214, which related to swaps 4 and 5.

By decision 249 of 5 June 2012, Prato resolved to annul in so far as they related to the swaps (with retroactive effect): decision 745, which authorised the signing of the master agreement; and decision 669, which related to swaps 4 and 5.

By decision 1625 of 2 July 2012, Prato resolved to annul (with retroactive effect) its prior decisions enabling entry into the master agreement and the swaps.

On 21 February 2013, the Regional Administrative Court for Tuscany ruled (i) that, to the extent that the nullified acts were not administrative acts, they could not be the object of self-redress and (ii) that, to the extent that the nullified acts were administrative acts, Prato’s exercise of self-redress was time-barred because it had not occurred within 3 years of the relevant acts, leading to the court pronouncing the cancellation of decisions 30 and 249. Prato is appealing the decision.

On 7 June 2013, by order of a criminal judge, criminal proceedings against Dexia and Mr Sommavilla were formally commenced. These proceedings are continuing.

This case involves important questions of Italian law. Each is the subject of strong contest. This section of my judgment is not intended to resolve those contests: to the extent necessary or appropriate, they are resolved in other sections.

I am concerned in this section to set out some uncontentious aspects of general matters relevant to the English court’s approach to, and understanding of, Italian law. To the extent that there is any dispute on such general matters I do not seek to resolve that dispute here.

Each side identified propositions concerning the English court’s approach to questions of Italian law. In most respects they were not disputed by the other side. I set out some of them below, modified so as to limit them to what was common ground.

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.

Issues of foreign law are to be proved as a fact by expert evidence. Both parties make positive (rival) cases as to the content of Italian law. Accordingly, both parties bear the burden of proving the propositions of Italian law on which they rely. The court will therefore need to decide which evidence it prefers. In other words, this is not a case in which the court can hold that Italian law is not sufficiently proved and therefore presume that Italian law is the same as English law: both parties agree that it is not, and lead evidence as to the respects in which it differs from English law.

When there is conflicting evidence as to what foreign law is, “the court should approach the conflict in the same way as it approaches other conflicts of fact”, which means evaluating the evidence of the expert witnesses “in much the same way as [the judge] would evaluate the evidence of any witness of fact”.

Professor Dettori agreed with what was said by Professor Napolitano at paragraph 3 of Napolitano 1:

Professor Dettori agreed with what was said by Professor Napolitano at paragraph 4 of Napolitano 1:

There is no formal system of precedent in Italy. It is common in Italy for first instance and appellate courts to reach conflicting decisions on the same issues.

Insofar as it is necessary to resolve any such conflicts, it is necessary to evaluate the cogency of the reasoning in the cases in the light of the decisions of the highest courts in similar cases and to decide which approach would be likely to be adopted by the highest courts. In that regard, there may be more than one stream of previous decisions, each stream pointing to particular conclusions. If so, and the point is one which the English court considers it necessary or desirable to resolve, then the English court will need to determine which stream is correct. As with any dispute of fact, this is a test that the court will apply by weighing the evidence before it, assisted by the Italian law experts.

Article 12 of the Civil Code states:

Professor Napolitano said, and I do not understand it to be in dispute, that Italian courts must:

It is common ground that ministerial circulars are not themselves a source of law. Nor do they have the force of law. They will nevertheless be regarded as binding on the public officials to whom they are addressed.

In this section it is necessary to refer to four Italian legislative provisions concerning local government, and in particular local government finance. They are:

Prato says that each was contravened by one or more of the swaps. In so far as Prato says that such contraventions give rise to defences, I refer below to those alleged contraventions as “the local government legislation contraventions”.

In addition, Prato asserts a general prohibition, not expressly found in these legislative provisions, on speculative transactions by the State, including local authorities. In so far as Prato says that contravention of this suggested general prohibition gives rise to a defence, I refer below to that alleged contravention as “the general prohibition contravention”.

In section D2.2 I examine the parties’ contentions on article 119 of the Constitution and the allied provisions in paragraph 17 of Article 3 of Law 350/2003. In section D2.3 their contentions on article 41 of law 448/2001 are examined. In section D2.4 their contentions on article 3 of ministerial decree 389/2003 are examined. That section completes my examination of the local government legislation contraventions.

I then turn in section D2.5 to examine the parties’ contentions on the alleged general prohibition.

Article 119 of the Constitution regulates the financial autonomy of regions and other local authorities. During the period 18 October 2001 to 20 April 2012, it stated, in translation and with the addition of Roman numerals conventionally used to identify sub-clauses:

These provisions are of a high order of generality. The crucial restriction for present purposes is the second to last sentence quoted above. Did any of the swaps constitute a resort by Prato to “indebtedness” within the meaning of that restriction? The opinion of Professor Dettori is that all, or at least some, of them did. Professor Dettori cites, among others, two decisions: one of the Court of Appeal of Bologna and the other of the Court of Auditors. If he is right, then Dexia does not suggest that the swap or swaps in question will have complied with the restriction, and Prato will have established that the swap in question was contrary to Italian local government finance law.

By contrast, Professor Napolitano asserts that none of the swaps involved “indebtedness” within the meaning of the crucial restriction. He says that the two decisions noted earlier do not represent Italian law, which he says is correctly stated in a different decision of the Court of Auditors and in an opinion published by the MEF. If he is right, then it will be necessary to examine other ways in which Prato says that the swaps were contrary to Italian local government finance law.

Professor Napolitano and Professor Dettori both refer to decision 425/2004 of the Constitutional Court. In that decision the court held that what is meant by indebtedness and investment in the crucial restriction in article 119 could not be determined on the basis of the constitutional provision alone. Instead it was up to the legislator to decide the content of those concepts, which it has done in definitions in article 3, paragraphs 16-19, of Law 350/2003. The court characterised these definitions as arising from financial and economic policy choices of the legislator made in close correlation with the constraints of the European treaties and in accordance with political-economic and technical criteria adopted by the organs of the European Union. In translation, the court is recorded as saying that the notion of “indebtedness” had been influenced by criteria adopted elsewhere in Europe to try to control public deficits, and reflected “all the income that cannot be deducted from the deficit calculated in order to comply with Community parameters”.

If I have correctly understood the Constitutional Court’s decision, then Italian law is that whether there is “indebtedness” within the meaning of the crucial restriction depends on whether the swap in question falls within the definition of “indebtedness” in article 3 of Law 350/2003. The relevant definition was set out in paragraph 17 of that article. Until December 2004 paragraph 17 included a final sentence enabling the MEF by decree to change the definition: this was held invalid in decision 425/2004 of the Constitutional Court, and it was removed by legislative amendment. Relevant parts of the definition, from enactment until 1 January 2009, have been translated as:

As a matter of first impression, this appears to be a specific list of the transactions which will constitute indebtedness for the purpose of the crucial restriction, subject only to the proviso that they will not do so if they fall within the last sentence cited above. Again as a matter of first impression, the list does not appear to include derivatives.

This first impression is consistent with the subsequent legislative history of paragraph 17: from 1 January 2009 onwards paragraph 17 was amended so as to add transactions which involved premiums received upon entering into derivatives. At first sight this suggests that the amendment was made because the legislature wished derivatives to be included, but only if they were entered into on or after 1 January 2009 and only if, upon entering into the derivative in question, the local authority received a premium.

In decision 49/2011 the Court of Auditors, Joint Sections, had to consider, amongst other things, a finance lease taken out by a public authority in respect of public infrastructure works. Under this transaction real property was granted to the public authority in the form of a lease against payment of a periodical fee for a specific number of years, at the end of which the authority had the right to acquire ownership by payment of a pre-established sum. The court pointed out that this was a mixed contract which included important financing components, involving a contractual procedure for the return of a notional loan for a sum corresponding to the value of the financial transaction put into effect. It drew attention to the need to ensure compatibility with provisions, among others, concerning the EU internal stability pact.

The court’s judgment noted that paragraph 17 “establishes in relation to local authorities which financial transactions constitute borrowing”, and that the list in paragraph 17 did not expressly include financial leases. The judgment then stated:

In Dettori 2 Professor Dettori asserted that this decision showed that paragraph 17:

Professor Napolitano disagrees. I am persuaded by him that he is right to do so. The Court of Auditor’s decision expressly acknowledges that it is paragraph 17 which establishes what constitutes borrowing. As regards that paragraph (and not as regards article 119 more generally) the court contrasts a formal, literal approach and an approach which looks at the substance of what is referred to in paragraph 17. It is the words “the taking out of loans” that are focussed upon by the court, but that is because the lease in question was in substance the taking out of a loan. The court’s reasoning would be equally applicable to something which was in substance the issuing of a bond, or in substance one of the particular types of securitization mentioned in paragraph 17. Equally the court’s reasoning will apply where a transaction does not literally fall within the proviso in the last sentence of the definition, but in substance does so: such a transaction will accordingly not be subject to the crucial restriction.

Prato relied upon a reference by the court in its judgment to the presence of an element of risk. I consider that this misunderstands the judgment. As Dexia observed, in identifying the substance of the transaction as one in which the public authority was being lent money in order to acquire an asset, the court was concerned with the risks relating to the completion and management of the project to which the finance leasing contract related rather than the risk of being required to make payments under the contract. The point the court was making was that a finance leasing contract is a form of loan (rather than e.g. merely a lease) where the public administration also bears the risk of completing and managing the project to which it relates.

In the course of Professor Dettori’s oral evidence it seemed to me that he may have been rowing back to a defensible analysis under which the question would be whether in substance the transaction was of a kind defined in paragraph 17 as constituting “indebtedness” for the purpose of the crucial restriction. Such an analysis would recognise that article 119 cannot contemplate that every occasion when a local authority incurs an obligation to pay money will be subject to the crucial restriction. When it employs staff it incurs such an obligation. Article 119 cannot have intended that the employment of staff will be subject to the crucial restriction. More analogous to the present case, by entering into a contract of insurance the local authority incurs an obligation to pay premiums. Article 119 cannot have intended that insurance contracts will be subject to the crucial restriction.

I add that, applying the analysis in decision 49/2011, I do not accept a suggestion by Professor Dettori that the change in the law as regards derivatives involving payment of a premium upon entry into the derivative, which took effect from 1 January 2009, was no more than clarification. In practice, for reasons given below, none of the swaps in the present case in substance involved a loan which Prato was to repay. I would agree that in theory if, prior to 1 January 2009, it could be demonstrated that a premium receivable by the local authority under a particular derivative was in substance a loan which the local authority was to repay, then the transaction would fall within the words “taking out of loans” in paragraph 17. What changed on 1 January 2009 was that it was no longer necessary to undertake this theoretical task. From that time onwards if the derivative in substance involved a premium payable upon entering into the derivative, then as a result of the legislative amendment it would, subject only to the proviso in the last sentence of the definition in paragraph 17, constitute “indebtedness” for the purposes of the crucial restriction.

The decision of the Court of Appeal of Bologna relied on by Prato is number 734/2014. It concerned parties whose names were anonymised. I shall refer to it as Municipality of C. I acknowledge that it is inconsistent with the analysis above: it found that swaps between a bank and the municipality were a form of current or potential debt, and that for this reason, among others, the swaps were null. Merely on the footing that an interest rate swap has an aleatory nature, involving current or potential debt, the swap is said to fall within the crucial restriction. As Professor Napolitano rightly observes, the judgment is expressed in terms which are vague and uncertain. It does not analyse paragraph 17 or seek to explain how derivatives generally can be fitted within that paragraph. I accept Professor Napolitano’s evidence that it does not accurately state Italian law.

Professor Napolitano is right also to say that decision 596/2007 of the Court of Accounts (Lombardy) of 26 October 2007, relied on by Prato, suffers from the same defects. By contrast I agree with Professor Napolitano that decision 1891/2009 of the Court of Auditors (Sicily) contains a clear and detailed analysis in support of its conclusion that an upfront premium did not, prior to the amendment of paragraph 17 on 1 January 2009, fall within the crucial restriction.

I note also that:

there is no suggestion in the present case that the swaps need to be treated as “indebtedness” in order to comply with either the form or the substance of European criteria;

there was a suggestion by Professor Sciarrone Alibrandi that article 119(VI) prohibited Italian public bodies from entering into speculative transactions and/or transactions that may expose the public entity to unjustified financial risks; this was, however, the subject of a reasoned answer by Professor Napolitano in Napolitano 2, and that answer was not challenged in cross examination.

Thus on a proper analysis the only question which arises on article 119 is whether any of the swaps in substance constituted a transaction falling within paragraph 17. Prato asserted that all the swaps were to be regarded as “forms of indebtedness”, but for the reasons given above, that is not the test. As regards swaps 2 and 6, Dr Faro considers that the pricing of these swaps absorbed the costs of closing prior swaps. Those costs, he says, are paid back over several years through an increase in what would otherwise be the interest rate (in the case of swap 2) or the level of the floor rate (in the case of swap 6). Accordingly he asserts that in economic terms the effect is equivalent to the giving of a loan to fund the unwind costs and the termination of the relevant prior swaps.

I do not agree that swaps 2 and 6 had an effect equivalent to a loan. Dexia points out that the value of the local authority’s existing position is not fixed or crystallised, that there is no advance by the bank to the local authority, and there is no certain obligation to pay any sum to the bank. It adds that the restructuring of a swap in this way does not create any debt that the local authority must pay to the bank. These features in my view rule out any conclusion that in substance any of the swaps involved a loan by Dexia to Prato.

For all these reasons I consider that none of the swaps in substance involved a loan which the local authority was to repay, and accordingly that there was no breach of article 119 of the Constitution. I turn to the second way in which Prato said that the swaps were contrary to Italian local government finance legislation, namely by allegedly contravening article 41 of law 448/2001.

Even if the swaps did not fall within the crucial restriction in article 119 of the Constitution, Prato says that they were nevertheless impermissible because they contravened article 41 of law 448/2001. If Prato is right, then a question arises as to the consequences of any such breach.

By contrast, Dexia asserts that none of the swaps contravened article 41 of law 448/2001. If it is right, then it will be necessary to examine the remaining ways in which Prato says that the swaps were contrary to Italian local government finance law.

At material times article 41 of law 448/2001 provided as follows:

In its defence Prato asserted that article 41.1 imposed a requirement that any derivative entered into by local authorities must satisfy a purpose which Prato described as “the limiting or containment of the costs of debt.” However the section of Prato’s closing submissions identifying alleged breaches of article 41 did not allege any specific breach of article 41.1. The upshot is that Prato’s complaints under article 41 all depend upon whether or not article 41.2 imposed requirements in relation to a particular swap, and whether those requirements were broken.

The second sentence of article 41.2 permits local authorities to provide for certain transactions under “refinancing conditions” that meet a particular requirement. The particular requirement has been translated as being that the refinancing conditions “allow a reduction of the financial value of the total liabilities to be paid by the bodies themselves, net of fees”. Italian legal scholars, judges and lawyers have used varying terminology for this requirement. A common feature of that terminology in the original Italian is the use of the words “convenienza economica”. Thus in ruling 5628/2011 (“Pisa I”) the Council of State referred to evaluation “ai fini della” [for the purposes of] “convenienza economica”. The expression “convenienza economica” has been translated in different ways. In a translation of the judgment in Pisa I it appears as “economic fitness”. In a translation of decision 1937/2014 of the Criminal Division of the Court of Appeal of Milan (“Arosio”) it appears as “cost-effectiveness”.

Translations of the reports of experts for the present proceedings used a coinage which was new to me, “economic convenience”. I suspect that whoever coined this expression assumed that “convenienza” in Italian and “convenience” in English are what translators know as “true friends”: words which look alike and mean the same thing. However, the coinage “economic convenience” is a dangerously misleading translation of “convenienza economica”, for to an English speaker in this context the word “convenienza” is a false friend. The word “convenience” in English is commonly used to describe something that is suitable or useful or does not cause a problem. The Italian “convenienza economica” describes something rather different. An example was given by one of our translators early on in the trial: if she finds the same shoes in two shops, but in the second shop they are available at a lower price than the first, then the second shop offers “convenienza economica”. What the second shop offers is financial advantage, not mere convenience. This is consistent with the agreed translation of the words actually used in article 41.2, “a reduction of the financial value of the total liabilities to be paid by the bodies themselves, net of fees”. The English coinage “economic convenience” is not apt to describe the concept envisaged by those words. The expression “financial advantage” is a better description of that concept. In this judgment I shall use that expression, and when quoting documents which translate “convenienza economica” in some other way I shall substitute “[financial advantage]”.

At the start of the trial Dexia’s primary stance was to acknowledge that there was a financial advantage requirement in article 41.2, but to assert that it did not apply to derivatives. By the close of the trial Dexia no longer advanced that primary stance. Instead Dexia accepted that the financial advantage requirement applied to derivatives, but only subject to four qualifications:

The first qualification asserts that the requirement only applies where two conditions are met: there must be a debt refinancing transaction and that transaction must involve new debt.

The second qualification asserts that in cases where it applies to derivative transactions, the requirement involves a consideration of the effective cost of the local authority’s refinanced debt, taking into account for this purpose the effect of any derivative which forms an integral part of a debt refinancing transaction.

The third qualification asserts that in cases where it applies to derivative transactions, the requirement does not involve examination of a derivative transaction in isolation: the overall benefit of the whole debt refinancing transaction is what matters for the purposes of article 41.2.

The fourth qualification asserts that in cases where it applies to derivative transactions, the requirement does not involve taking account of either the initial MTM or the so-called “implicit costs” of the derivative, which are irrelevant since they are not actual costs. Instead the local authority should make a projection as to the effect of the derivative on its cost of debt using forward rates as at the date of the transaction.

I start by examining the first qualification and its two conditions: there must be a debt refinancing transaction and that transaction must involve new debt. Article 119 of the Constitution envisages that “indebtedness” falling within that article will have come about in order to finance “investment” within that article. It does not limit an authority’s ability to reduce its “indebtedness”, for example by paying it back early. However, subject to legislative exception, it would not permit a refinancing involving new “indebtedness” in the absence of new “investment”. Professor Napolitano explains that the legislative purpose of article 41.2 is to permit new “indebtedness” as part of a refinancing. It is in this context that the financial advantage requirement arises. Accordingly the two conditions are the logical consequence of this legislative purpose: the aim is to permit refinancing, a permission which is needed because refinancing involves new debt.

In support of the first qualification Dexia relied on what was said in three judgments:

In Pisa I (cited above) the Council of State rejected contentions that derivatives could not engage article 41.2, but its reasons for doing so were that the derivatives in question constituted “the specific instruments” through which a debt restructuring operation was realised. The Council of State in Pisa I did not decide whether there had been a breach of article 41, but instead sought an expert evaluation “of the case in controversy (operation of restructuring debt, financial derivative contract and its effects) according to criteria of economic-financial science”. The appointed expert was Dr Angeletti.

The Council of State considered the report of Dr Angeletti in decision 5962/2012 (“Pisa II”), identified both derivatives and new debt in that case, and proceeded upon the basis that they were “connected and interlinked, as they are structurally geared toward achieving the objective set forth in article 41 of law no 448 of 28 December 2001”.

In Arosio (cited above) the Court of Appeal of Milan said, in a passage which treated the MEF circular as correctly stating Italian law:

I am persuaded by Professor Napolitano that these two conditions represent Italian law. Professor Dettori’s broader approach, in effect, suggested that article 41 was concerned to ensure that every financial transaction of a relevant local authority was subject to the financial advantage requirement. This, however, does not acknowledge that article 41.2 is concerned with permitting local authorities to do specific things, nor does it acknowledge or explain the express reference in the second sentence to the things in question taking place “under refinancing conditions”. The cases he cited similarly offer no such explanation.

In support of Professor Dettori’s broader approach Prato said that Professor Dettori’s was the more teleological approach consistent with containing the cost of debt, drew attention to the width of the words “conversion” and “renegotiation” in article 41.2, and suggested that it made no sense to insist on the financial advantage requirement where new loans are negotiated but, by contrast, to free local authorities from the financial advantage requirement by permitting renegotiation of existing loans. Professor Napolitano accepted that article 41.2 could be read in a way which imposed greater restrictions on local authorities, and that the words “conversion” and “renegotiation” in the second sentence were wide. But those words are followed by the phrase “under refinancing conditions”. To my mind Professor Napolitano’s distinction is logical in so far as it focuses on the question of what would or would not be new “indebtedness” for the purposes of article 119 of the Constitution. That article is concerned to ensure that the proceeds of relevant borrowing will be applied only to relevant investments. Once that has happened, I am persuaded by Professor Napolitano that later variation of the terms of the borrowing is immaterial for the purposes of Article 119.

My conclusion on this first qualification has the consequence that swaps 1 to 3, which involved no debt restructuring, do not form part of a transaction which has to satisfy the financial advantage requirement in article 41 of law 448/2001. Further, as the restructuring associated with swap 6 did not involve new debt, but was no more than a variation of the terms for existing debt, it follows that swap 6 equally does not form part of a transaction which has to satisfy the financial advantage requirement in article 41 of law 448/2001.

Thus the remaining three qualifications are relevant only as regards swaps 4 and 5. Neither the second qualification nor the third is disputed. In these circumstances it is not necessary for me to decide which side is right in relation to the fourth qualification. The reason is that even if the initial MTM, or the so-called “implicit costs”, of swaps 4 and 5 are included in the calculation Prato accepts that the financial advantage requirement is met. In case my reasoning on the first qualification is wrong, however, I briefly examine below the rival arguments on the fourth qualification.

The position in that regard is that, building upon the second and third qualifications, Dexia contends that neither the initial MTM nor the so-called “implicit costs” of the derivative impact upon the effective cost or overall benefit of the whole debt refinancing transaction, and thus it follows that under Italian law the fourth qualification is made good.

Dexia’s case in this regard is supported by Professor Napolitano. On Prato’s side, however, Professor Dettori considers that the issues raised fall outside his expertise.

Dexia’s reasoning is, to my mind, strongly supported by Pisa II. Dexia points to passages to the effect that the so-called “implicit costs” of a derivative “do not by any means constitute an effective cost … but merely stand for the value that the swap could have had in an abstract and hypothetical (but utterly unrealistic and untrue) negotiation”.

Prato acknowledges, as it must, the existence of these passages. It makes six points in response. However in my view none provides a satisfactory answer.

Prato notes that Pisa II follows but did not and could not overturn Pisa I. In my view Dexia is right to say that the passages in Pisa II that it relies on did not seek to overturn Pisa I, for what Pisa I did was to direct a report on numerous matters, among them “hidden costs”, so that the court could consider them.

Prato notes that in Pisa II the passages that Dexia relies on were not determinative. Professor Napolitano acknowledges that this is right, in the sense that even if the MTM were taken into account the overall transaction was found to satisfy the financial advantage requirement. Nonetheless the Council of State went out of its way to state its opinion on the present aspect of the matter.

Prato seeks to derive support from the way in which Professor Napolitano commented on what the Council of State had said. However nothing in those comments was inconsistent with his oral evidence.

Prato, with some support from Professor Sciarrone Alibrandi, says that the Council of State gave undue weight to a criminal decision of the Court of Cassation. I accept that the criminal decision involved a context different from the present, but it nonetheless appears to me significant that the court considered that the MTM did not express a practical and current value.

Prato, with support from Dr Faro and Professor Sciarrone Alibrandi, said that the MTM had a continuing effect. Mr Malik’s evidence, however, demonstrates that the true position is that it is the commercial terms which have a continuing effect. It does not matter for this purpose whether the impact of those terms is computed by calculating an MTM, whether the MTM is described in accounting terms as a “cost”, or whether the MTM expresses market expectation as to the likely outcome of a swap. Nor does it matter whether the MTM or any particular factor entering into pricing was “hidden”. Whether the financial advantage requirement is met will depend on whether the commercial terms of the overall transaction would “allow a reduction of the financial value of the total liabilities to be paid” by Prato, net of fees.

Prato describes it as “a nonsense” to interpret Pisa II as a judgment which excludes implicit costs. This is a variant on Prato’s second point. It is true that the Council of State was prepared to accept an expert report which gave negative weight to “implicit costs”. The context for that, however, was that even after taking this course the expert concluded that the benefits of the transaction outweighed the costs. There was thus no need to say anything about the validity of giving negative weight to “implicit costs”: yet the Council of State went out of its way to do so.

Dexia’s reasoning is also, to my mind, supported by Arosio. The Court of Appeal of Milan, when discussing the Pisa II judgment, stated:

Prato makes two points in response on this aspect of Arosio. However, in my view neither point demonstrates that the Court of Appeal disagreed with the Council of State:

In Arosio the Court of Appeal said it was fruitless to determine whether implicit costs were virtual or real. However, as pointed out by Professor Gentili, the court was saying this in a context where it was holding that non-disclosure by the banks was legal.

The Court of Appeal stressed repeatedly that public authorities should assess the value of derivatives and take account of implicit costs. However on this aspect Professor Gentili commented that the court was focussing on what public authorities should do.

In these circumstances I turn to the last remaining ways in which Prato says that the swaps were contrary to specific Italian local government finance legislation, namely alleged contraventions of article 3 of ministerial decree 389/2003.

Article 41 of law 448/2001 contemplated that a decree of the Ministry of Economy and Finance would approve the rules on, among other things, the use of derivatives by local government. This was duly implemented by ministerial decree 389/2003. The final aspect of Italian local government law relied on by Prato concerns article 3.2 of that decree. It came into force on 4 February 2004, and thus applies only to swaps 4, 5 and 6. At material times Articles 3.2 and 3.3 stated:

Even if article 119 of the Constitution was not contravened, and whether or not article 41 of law 448/2001 was contravened, Prato says that article 3 of ministerial decree 389/2003 was contravened. It identifies such a contravention as occurring because one or more of swaps 4, 5 and 6 contravened one or more of articles 3.2(d), 3.2(f) and 3.3. If Prato is right, then a separate question will arise as to the consequences of the particular contravention.

By contrast, Dexia asserts that there was no contravention of article 3. If Dexia is right, then in the light of my conclusion that there has been neither a contravention of article 119 of the Constitution nor a contravention of article 41 of law 448/2001, the consequence will be that none of Prato’s specific Italian local government legislation contraventions has been established.

The alleged contravention of article 3.2(d) is not obvious at first sight. As can be seen above, it permits purchase of an interest rate collar in which the buyer is guaranteed an interest rate to be paid, fluctuating within a predetermined minimum and maximum. These words on their face describe the collar which Prato purchased as part of swaps 4 to 6. In each case the maximum interest rate to be paid by Prato is identifiable as a cap, and the minimum interest rate to be paid by Prato is identifiable as a floor.

Prato notes that article 3 does not envisage a floor being “sold” by an authority in the absence of a cap, and advances a proposition that under article 3.2(d) local authorities are prohibited from agreeing to swaps with collars unless the MTM of the floor at inception is equal to the MTM of the cap at inception. It claims to derive support for this from an MEF circular dated 27 May 2004. As Professor Napolitano points out, the circular simply does not say this. What it says is translated as follows:

Professor Napolitano had no difficulty in accepting that this passage in the circular represents Italian law. As he made plain in cross examination, what he could not discern was how either article 3.2(d) or the circular supported a suggested requirement that there must be equivalence or equilibrium between the value of the cap and the floor.

The evidence relied upon by Prato in support of such a requirement came from Professor Sciarrone Alibrandi. Her opinion was that such a requirement was needed in order to ensure that article 3.2(d) was “in line with Article 41 [of law 448/2001] which has a higher value than that of Decree 389 …”.

I am persuaded by Professor Napolitano that Professor Sciarrone Alibrandi’s opinion in this regard does not represent Italian law. As Professor Napolitano observed in Napolitano 1, decree 389/2003 is concerned to implement article 41.1 of law 448/2001. Nothing in article 41.1 calls for an equivalence of the kind asserted by Professor Sciarrone Alibrandi. Moreover, as it seems to me, Dexia rightly adds that there is no inconsistency between the law on the one hand excluding the possibility of a local authority selling a floor on its own but on the other hand permitting the sale of a floor as part of a collar transaction even though the MTM of the floor is greater than the cap.

As regards article 3.2(f), Prato says it was breached by swaps 4, 5 and 6 in two respects. Both involve an alleged contravention of two of the four elements of the second sentence of article 3.2(f), a sentence which can properly be described as opaque. I set it out below with the elements numbered in square brackets, the two allegedly contravened being elements [3] and [4]:

Professor Napolitano said that Article 3(2)(f) only applies to derivatives that “restructure debt” by exchanging capital payments in addition to interest payments. As is rightly pointed out by Prato, and as was put to Professor Napolitano in cross-examination, article 3(2)(f) does not say this. Professor Napolitano’s response was to say that “a forecast ex ante of payments”, this being contemplated in element [3], is possible only if capital is involved. This reasoning is difficult to follow. Professor Napolitano acknowledged that he was not an expert in finance and commercial law. It seems to me that the proposition he advanced went beyond his expertise.

In relation to element [3], the circular stated in translation:

The evidence of Mr Malik was that it was possible to prepare net cashflow profiles for each swap using “market implied rates at the time of execution of the swap”. If this were done, over the life of each of the swaps there would be a period of net cash receipts, followed by a period of net cash payments, followed by a period of net cash receipts. Calculations prepared by Dr Faro identified only a few occasions on which the expected value of Prato’s payments increased. Neither of these appears to me to show “an increasing profile” (using the words in element [3]), and they certainly do not show a concentration of payments close to maturity (using the words of the circular).

As to element [4], the circular explains that discounts or premiums are permitted to cater for restructurings in market conditions which have changed from those at the time when the debt was underwritten, but those discounts or premiums may not be greater than 1% of the face value of the underlying debt. As Mr Malik observes, none of swaps 4, 5 or 6 involved the payment of such a discount or premium. Dr Faro claims to identify it in the MTM, but there is nothing in either the wording of element [4] or the circular to suggest that an MTM constitutes a discount or premium: on the contrary, what is contemplated is something which is paid.

The final alleged breach concerns article 3.3. What it says is that the derivative transactions identified as permissible in earlier parts of article 3 “are allowed only in the presence of real liabilities due.”

The alleged contravention of article 3.3 is not obvious at first sight. Swaps 4, 5 and 6 all took place in the presence of real liabilities due from Prato to bondholders.

Prato asserts that if there has been a variation in the underlying debt of a derivative instrument, any consequent change to the derivative instrument must not involve a loss to the local authority. It is difficult to see how this can be derived from article 3.3. There is some support for it, however, in the circular, which states in translation:

Dexia objects that it is common ground that a ministerial circular is not a source of law and cannot extend article 3.3. This is indeed common ground: see section C4.3 above. In many cases the circular offers a helpful explanation of what is implicit in the decree, but in this particular instance the circular seems to me to go well beyond what is said in article 3.3.

Dexia also objects that, while Prato’s claimed “loss” lies in the initial negative MTM or alleged hidden costs, there is no evidence of Italian law to that effect. This seems to me to be a complete answer.

For all these reasons I conclude that there were no contraventions of article 3. It follows that none of Prato’s allegations of local government legislation contraventions is sustainable.

In these circumstances I turn to the last way in which Prato says that the swaps were contrary to Italian local government finance law, namely alleged contraventions of a general prohibition on speculative transactions.

Prato advanced a proposition that under Italian law, even where there is no legislative contravention of the specific rules of article 119 of the Constitution, article 41 of law 448/2001, and article 3 of decree 389/2003, those provisions are to be read as prohibiting speculative derivatives. In my view Dexia are right to say that there is a short answer to this proposition: it was convincingly denied by Professor Napolitano, and only weakly supported by Professor Dettori.

In cross-examination Professor Dettori was asked what he meant by “speculative”. He gave various replies before summarising his position in this way:

Professor Dettori relied upon a decision of the Council of State dated 6 December 2000, where it was said that:

It is by no means obvious that the Council of State here recognised a rule of Italian law under which a court could be asked to decide whether a contract of a local authority was speculative and if so to hold the contract invalid or unenforceable. Nor was I able to discern from Professor Dettori’s evidence how this could be the case. More fundamentally, however, Professor Napolitano in Napolitano 2 set out his reasons for rejecting Professor Dettori’s proposition. Professor Napolitano’s evidence on this point was not challenged in cross-examination.

For the reasons given above, I have concluded that there were no contraventions of Italian local government law. It follows that it is not necessary to investigate whether some or all of the breaches alleged by Prato would have rendered relevant swaps invalid or unenforceable. I consider it undesirable to attempt such an investigation in the present judgment: if there were to be a conclusion that I am wrong in rejecting any particular allegation by Prato of breach of Italian local government law, then a determination as to the consequences of that breach will be likely to depend upon the particular reasons for that conclusion.

In addition to the administrative law provisions discussed in section D above, Prato relies on particular provisions of Italian financial services law and Italian civil law. Ordinarily, provisions of Italian law of this kind would not be applicable to the swaps, for the parties have agreed in the schedule to the master agreement that English law is to govern. However in the present case the contractual dealings between the parties all took place during the period in which the EEC Convention on the Law Applicable to Contractual Obligations (“the Rome Convention”) was incorporated into the law of the United Kingdom by the Contracts (Applicable Law) Act 1990 (“the 1990 Act”). Under article 3(3) of the Rome Convention what are called “mandatory rules” of the law of Italy will apply if, apart from the choice of English law and jurisdiction:

In order to benefit from this exception Prato acknowledges that it must show that, with the exception of the choice of law and jurisdiction, no “elements relevant to the situation at the time of the choice [of law]” are connected with a country different from Italy. It observes in that regard that Italy was where both parties were incorporated, it was where the parties communicated with each other, it was where the swaps were entered into, and it was where the obligations under the swaps had to be performed.

At the start of the trial Dexia identified objections, in the form of two reasons for suggesting that there were “elements relevant to the situation” which were not connected with Italy:

the master agreement was in the standard ISDA form, drafted by international working groups for routine use in derivative transactions in the international capital markets. Dexia submitted that the master agreement was designed to promote certainty, and that the significance and global nature of ISDA had been recognised by English courts;

in the case of each of the swaps Dexia entered into a back to back hedging swap with a bank outside Italy in the international market using the same industry standard documentation.

To my mind, Prato is right to say that both these points are misconceived. As to the master agreement, it is true that it is an international standard form, but it does not follow from this that it is an “element in the situation” which is connected to a country other than Italy. It is of course designed to promote certainty, but that does not give it a connection to a country other than Italy. Nor does the significance and global nature of ISDA. Even if the standard form itself were shown to have a connection with another country, that would not in the present case be an “element relevant to the situation” as it existed at material times. Throughout the relevant period everything relevant to the use of the form happened in Italy. As to Dexia’s decision in each case to choose a non-Italian counterparty for its back to back hedging swap, that does not appear to me to be an element relevant to the situation as between Prato and Dexia. Whether or not Dexia entered into a hedging swap is a matter for Dexia alone: to Prato it is immaterial. There was no contemplation that a non-Italian entity would take over obligations of either party. Dexia’s choice to use a non-Italian counterparty is something which is completely external to “the situation” at the time that choice of law was agreed.

For these reasons I conclude that neither of Dexia’s objections is valid. Accordingly Prato is entitled to rely upon such provisions of Italian law as constitute “mandatory rules” for the purposes of article 3(3) of the Rome Convention.

The provisions of Italian financial services law relied upon by Prato in this regard are conceded by Dexia to constitute “mandatory rules”. They are found in a legislative decree and in regulations. The legislative decree is number 58 of 24 February 1998. This decree is Italy’s consolidated law on financial markets and investment services, known in Italian as Testo Unico della Finanza, commonly abbreviated to “TUF”. Relevant regulations were made by the Italian Securities and Exchange Commission, Commissione Nazionale per le Società e la Borsa (more commonly known as “Consob”).

As to civil law, Prato relies upon the Italian civil code (“CC”). It says that the swaps were subject to requirements under article 1418, paragraph 2, of the code, but failed to comply with two such requirements. The first is that a contract must have a lawful “causa”. The second is that it must also have a determinable “oggetto”.

Section E2 below examines whether the swaps contravened provisions of Italian financial services law. Section F then deals with the consequences of my conclusion in that regard.

Prato relies upon two sets of provisions concerning financial services. The first comprises article 30 TUF, which is concerned with “off-site offers”, and article 32 TUF, which is concerned with distance marketing techniques. As regards both these articles, there is a requirement that forms used by those offering financial services must state the right of the investor to withdraw within seven days of making an investment.

The second set of provisions concern article 23 TUF and article 30 of Consob regulation 11522/1998 (“Consob 11522/1998”). These two provisions together require that the contract between an investor and a financial service provider must contain specific provision in relation to three aspects of the relationship between the investor and the service provider. These aspects concern the giving of orders and instructions by the investor, reports by the service provider on the activity carried out, and (as regards contracts for trading and the reception and transmission of orders) “providing and replenishing the means for carrying out or guaranteeing the transactions ordered…”.

Articles 30 and 32 TUF provide:

Professor Gentili’s opinion is that article 30 does not apply to the swaps. His reasoning, both in Gentili 1 and in Gentili 2, was diffuse. In cross-examination he accepted that essentially his reasoning relied on two points. The first was that article 30 only applies to offers to the public; and the second was that article 30 only applies to unsolicited contact.

Professor Gentili derived support for both these points from decision 2065/2012 of the Court of Cassation dated 14 February 2012. The case concerned bonds sold to an investor by Banca Fideuram S.p.A., and I will refer to it as “Fideuram”. It was the investor who appealed to the Court of Cassation. The appeal failed because that court agreed with the reasoning of the Court of Appeal of Milan: the bonds had been purchased as part of the implementation of a framework agreement between the investor and the bank, whereas the reference to “placement” service in article 30.6 was characterised by an agreement by an issuer or offeror and the intermediary in charge of the placement, aimed at an offer of financial instruments to an undetermined public, issued subject to predetermined time and price conditions. I shall refer to this interpretation of “placement” as “the Fideuram primary interpretation”. It was in line with Consob guidelines, among them a notice in 1997 defining “placement service” in this way, and article 35 of Consob 11522/1998 which imposed requirements on those providing a placement service.

The Fideuram primary interpretation of “placement” inevitably entails that article 30.6 only applies to offers to “an undetermined public”. In addition, however, the Court in that case added:

The Court of Cassation noted that the decision under appeal, in an analysis “not contrasting with this consideration”, ruled out on the facts that this hypothesis [exposure to the risk of making poorly considered decisions] was manifest, the purchase of bonds having been connected with the pre-existence of a previous relationship between the investor and the bank. Thus, said the Court of Cassation:

I am persuaded by Professor Gentili that these passages involve separate and distinct reasoning that there will not be “placement” within the meaning of article 30.6 if the purchase does not take place out of the investment provider’s initiative. The reasoning is that in those circumstances there will be no effect of “surprise”. I shall refer to this interpretation of the word “placement” as “the investment provider initiative interpretation”.

However the Fideuram case was not an end of the matter. Because there had been differing rulings on the meaning of “placement”, and because there continued to be academic debate, arrangements were made for a further case to be the subject of a hearing before the Joint Chambers of the Court of Cassation. It was a case where, in the absence of a “placement” within the Fideuram primary interpretation of that word, Banca Mediolanum S.p.A. had dealt with an individual investor using forms which did not include any statement of a right to withdraw. I shall refer to the resultant decision 13905/2013 of the Joint Chambers of the Court of Cassation of 3 June 2013 as “Mediolanum”. In Mediolanum the Court of Appeal of Palermo had held that the investor could claim nullity under article 30.6 and 30.7 where the investor had subscribed for bonds following requests from a representative of the bank. It took the view that obligations under article 30.6 were not only applicable within the public offer of financial instruments that the broker has placed, as a result of being appointed by the issuer or offeror, but also “in any other negotiation event of such instruments off the site of the broker”.

The Joint Chambers noted that “placement” could be used in the narrow sense of an agreement by an issuer or offeror and the intermediary in charge of the placement, aimed at an offer of financial instruments to an undetermined public. They described this as a “placing service”. However the use of the word “placement” to embrace activities within article 30.1(b) suggested that, as a whole, the term “placing” in article 30 had been used by the legislator with much wider and more general meaning; almost as a synonym for any transaction aimed at issuing financial products or investment services on the market.

After observing that the literal text alone did not seem capable of providing a satisfactory answer, the Joint Chambers said in their decision at paragraphs 4 to 7, to which I have added sub-paragraph numbers in square brackets:

It was common ground that the only case in which Mediolanum had been considered was the decision of the Court of Appeal of Bologna in Municipality of C, discussed in section D2.2 above. As noted in section D2.2, the swaps in that case were held to be null because, among other reasons, they were a form of current or potential debt and violated the crucial restriction in article 119 of the Constitution. It was accordingly unnecessary to deal with an argument by the Municipality that it was entitled to avoid the swaps under article 30.6 and 30.7 TUF. The Court of Appeal dealt with the matter very shortly. It quoted from paragraphs 4[.2] and 4[.3] of Mediolanum. Citing those passages, the Court of Appeal of Bologna said that the Municipality:

Professor Gentili acknowledged in cross-examination that, as a decision by the Joint Chambers, Mediolanum is particularly authoritative. He also acknowledged that paragraph 7 of the decision expressly stated that earlier decisions of the Cassation Court (including Fideuram) could no longer be followed. What that meant, he said, was that the Joint Chambers wanted article 30 to be applied not only with regard to the narrow “placing service”, but also with regard to any investment service rendered.

In cross examination Professor Gentili accepted that articles 30.6 and 30.7 applied the right of withdrawal, the duty to state that right in relevant forms, and the consequence of nullity at the discretion of the investor where no such statement appeared, to contracts falling within article 32 as well as those falling within article 30. He added that he agreed that article 32 was applicable whether or not the investor had solicited the making of the investment. He was asked why, on his view of the matter, these consequences should apply in relation to article 32 whether or not the investor had solicited the making of the investment, but in relation to article 30 should be excluded in cases where the investor had solicited the investment. His reply was that under article 30 it was necessary to consider the intention of the legislator, which was to protect the investor from surprises. He did not, however, explain why the relevant words in articles 30.6 and 30.7, which only appeared once in article 30, could have been intended to have different meanings, depending upon which of the two provisions was engaged in the particular case.

As regards the differing interpretations prior to the decision of the Joint Chambers, Professor Gentili said that the divergence of views had been solely on the question whether the narrow interpretation of “placement” in the sense of “placement service” (i.e. the Fideuram primary interpretation) was right. There had, he said, been no disagreement about the need for the contract to have come about at the initiative of the investment provider, for there had been only a single judgment of the Court of Cassation on that aspect, namely the Fideuram judgment.

Professor Gentili was then taken to the paragraphs which I have numbered 4[.1] and 4[.2] in the judgment of the Joint Chambers. It was suggested to him that these paragraphs addressed a point that unsolicited contact might be the usual context for article 30, but it was not the exclusive context for article 30. Professor Gentili replied that they had to be interpreted with a different slant. Where a contract was entered into elsewhere than on the premises of the investment provider, it meant that the investment provider had tried to find the investor. In those cases the investor was considered to be taken by surprise. Professor Gentili added that if the investor had been taken by surprise, but not because of the place where the contract was signed, rather because the investor had not had time to reflect on it, then the investor should have seven days to think about it. In his opinion, what the Joint Chambers meant was that if there was solicitation by the investor, regardless of where the contract was entered into, the investor had thought about it and did not need seven days to think about it or change his mind. However Professor Gentili did not explain how it could be that an investor who entered into the contract on the investment provider’s premises could be entitled under article 30 to have a cooling off period of seven days even though the contract had not been entered into “off-site”. As to Professor Sciarrone Alibrandi’s view that it was entering into the transaction off-site that triggered article 30.6, Professor Gentili reiterated that the site where the contract was entered into “will not warrant in any way the necessity to re-think about it.” The question was whether the investor had entered into the contract that the investor wanted and thus did not need to think about it: that would not depend on the site, it would depend on the surprise element. He suggested that paragraph 4[.2] was actually stating this.

Professor Gentili was then taken to paragraph 5[.4]. He acknowledged that in that paragraph the Joint Chambers said that a pre-existing relationship would not be sufficient to remove a transaction from article 30.6. Professor Gentili said that he was not asserting that a previous relationship inevitably precluded the possibility that the investor would be taken by surprise. It made that possibility more unlikely, but Professor Gentili’s point was that if it were the investor that asked the investment provider for a certain type of product, then of course the investor was not taken by surprise. Thus the most important element was, who is taking the initiative? In that regard, however, it was pointed out to him that the Joint Chambers in this paragraph had emphasised the importance of looking at the text of the law. Professor Gentili replied that while Consob had taken the view that a pre-existing framework agreement would suffice to exclude article 30, the Joint Chambers regarded it as “relevant but not determinative and it goes back to the problem of who is taking the initiative”. However Professor Gentili did not identify any passage in the decision of the Joint Chambers stating that the determinative question was, as had been suggested in Fideuram, whether the investment provider had taken the initiative.

Also in cross examination Professor Gentili accepted that in paragraph 6 the Joint Chambers made the point that article 30 should be interpreted in a way which made it effective and which maximised consumer protection. Nevertheless Professor Gentili maintained that he understood the Joint Chambers to be saying that article 30 would only apply “provided it is credible that there was a surprise effect, and that there isn’t if the investor has taken the initiative”. Professor Gentili read from paragraph 4[.2] of the decision of the Joint Chambers, noting that the reason for giving protection to off-site investors was because an off-site investment would mean that the initiative did not usually originate from the investor, and that it was logical to assume in such cases that the investment was the result of an offer from promoters, a solicitation which could have surprised the investor and could have induced the investor to make a trading choice that had not been carefully considered. He then went on to assert that from this he understood the Joint Chambers to say that if the initiative came from the investor then the protection given by article 30 was not necessary.

It was then suggested to Professor Gentili that the aim of article 30.6 was to lay down a clear and workable rule. Professor Gentili initially stressed article 12 of the preliminary provisions of the Italian civil code, requiring the court to look for the intention of the legislature, this being to find the right remedy to the problem considered. If the remedy was a change of mind, then the problem was surprise, because otherwise the remedy would be pointless. As to the position being clear, Dexia knew that when the contract was entered into off-site they had to include the withdrawal clause, but also knew, because this was clarified by the case law of the court, that they could show that there was no surprise because it was the investor that had solicited the very contract which had been entered into. As to the notion that Professor Sciarrone Alibrandi’s view provided maximum consumer protection for retail clients, Professor Gentili would not accept this, and instead claimed that Professor Sciarrone Alibrandi’s view would provide a pointless protection, for it would protect the consumer even where the problem of surprise did not exist.

At the start of her evidence on this aspect Professor Sciarrone Alibrandi was cross examined about paragraph 81 of Sciarrone Alibrandi 1. In that paragraph she had said that the case law had not specifically examined whether articles 30 and 32 TUF applied to cases where a contract relating to the provision of investment services had been entered into following a public tender. She had then commented that in a decision of the Tribunal of Rimini dated 12 October 2010 article 30 had been applied to derivative contracts signed by the local authority after a public tender.

It was suggested to Professor Sciarrone Alibrandi that the cases on article 30.6 had never considered contracts which came into being during the course of an advisory relationship between the parties. Professor Sciarrone Alibrandi replied that the judgment of the Joint Chambers had thought that article 30.6 could be applied to a case where there was a current relationship between the intermediary and the investor. The Joint Chambers had thus not considered article 30.6 to be a rule aimed only at where there is a first contact between the intermediary and the investor: on the contrary it was a rule that was compatible with granting the protection that it conferred in circumstances where there was an ongoing relationship between the intermediary and the client.

It was put to Professor Sciarrone Alibrandi that the cases on article 30 had never considered contracts which came into effect through the investor making an irrevocable offer to the intermediary which was then accepted by the intermediary. Professor Sciarrone Alibrandi replied that this was quite common, adding that she did not think it very important with regard to the application of article 30.6.

Professor Sciarrone Alibrandi was then taken to what had been said in each of Fideuram and Mediolanum. She was asked to assume that Prato had appointed Dexia after a public tender, that each transaction had been the subject of a prior request by Prato that Dexia produce a proposal aimed at meeting a particular need identified by Prato, that irrevocable offers were sent to Dexia rather than by Dexia, and that each was only sent to Dexia after Prato had passed internal resolutions permitting the proposed transaction. It was suggested to her that those facts would take the case outside the Ratio Legis of the right of withdrawal in article 30. Professor Sciarrone Alibrandi replied that when article 30 referred to an off-site offer it identified the Ratio in “the need to protect the client from not sufficiently thought choices, and the fact that the contracts have been entered into off-site is considered a significant fact…”.

Various reasons were then suggested to Professor Sciarrone Alibrandi as to why the factors that had been suggested on behalf of Dexia would take the case outside the Ratio Legis of article 30. Professor Sciarrone Alibrandi replied that article 30 was a rule which was interpreted in a general way, but was based on:

In response to further questions Professor Sciarrone Alibrandi reiterated that what the Joint Chambers had stated was that it was the fact that the transaction had been carried out outside the premises [of the investment provider] which made it necessary to have special protection, because this meant that generally the initiative had not come from the investor.

Professor Sciarrone Alibrandi was then asked about Municipality of C. Professor Sciarrone Alibrandi acknowledged that the Court of Appeal of Bologna had referred to Mediolanum. She initially accepted that it was consistent with Professor Gentili’s interpretation of article 30. She then qualified this acceptance. The judgment appeared to her to have a more limited content. It did not support Professor Gentili’s wider interpretation under which the question would be whether the contract had been solicited by the investor or had not been solicited by the investor. It was dealing with the fact that there had been previous resolutions and discussion. As to a suggestion that the Mediolanum interpretation of article 30 was correct as a matter of Italian law, Professor Sciarrone Alibrandi maintained that under Italian law the legislator had identified a scope for article 30.6 which was “connected with the off-site offer.” The legislator had considered there to be a need for a higher protection for the client when the client was outside of the premises of the intermediary.

It was then suggested to Professor Sciarrone Alibrandi that the reference in article 30.1 to “the promotion and placement with the public” required the initiative for the transaction to have come from the intermediary, not from the investor. Professor Sciarrone Alibrandi replied that the Joint Chambers had considered those words when reaching their decision. She added that those words had been held by the Joint Chambers to have “a quite wide sense”. They referred to all contracts that were entered into off-site by the intermediary with the investor. As to a suggestion that her interpretation meant that in article 30 the words “with the public” added nothing, Professor Sciarrone Alibrandi replied that the Joint Chambers’ interpretation had taken away the value of “with the public” because the rule in article 30.6 was held by them to be applicable also to an offer made to a specified investor.

Dexia’s closing submissions drew attention to the definition of “off-site offer” in article 30.1, and submitted that they showed that article 30 was concerned only with unsolicited approaches by intermediaries. This, however, ignores the fact that article 30.6 does not address the location of the offer but, for present purposes, addresses whether a contract for the placement of a financial instrument has been executed off-site. Accordingly, as it seems to me, Professor Sciarrone Alibrandi’s reasoning demonstrates that this argument does not assist Dexia.

The closing submissions advanced a general argument that Professor Sciarrone Alibrandi had misunderstood paragraph 4[.2] of Mediolanum. This submission ignored the compelling points which had been put to Professor Gentili in cross examination, none of which in my view was satisfactorily answered. The reasoning of the Joint Chambers in paragraphs 4[.2] to 4[.4] is incompatible with the investment provider initiative interpretation. This is also the case in relation to the discussion in paragraphs 5[.1] to 5[.5] and paragraph 6.

The closing submissions again relied upon the words “with the public”, and criticised Professor Sciarrone Alibrandi’s evidence. However this criticism failed to acknowledge Professor Sciarrone Alibrandi’s point that the Joint Chambers had held that article 30.6 and 30.7 applied not only to an offer made to the public but also to an offer made to a specified investor. If it were otherwise then their decision would have to have been different: Mediolanum concerned dealings with a single investor. This must, on the evidence before me, be treated as an authoritative determination.

Dexia’s closing submissions then placed reliance upon the decision of the Court of Appeal of Bologna in Municipality of C. Dexia acknowledged that the observations of the Court of Appeal of Bologna in that case were not determinative of the decision, but the Court had nevertheless taken the trouble to state them. It is apparent, however, from the very brief reasoning in Municipality of C on this point that the Court of Appeal of Bologna was either unaware of or chose not to deal with the features of Mediolanum which were put to Professor Gentili. Those features have been demonstrated to show that the Joint Chambers considered that the legislature had decided upon a clear rule that the execution of the contract off-site, because of the risks that it carried, should be the criterion which would lead to a requirement for a seven day cooling off period and for this to be stated in the relevant forms.

Dexia’s closing submissions turned to the suggestion that Professor Sciarrone Alibrandi’s interpretation would maximise the level of consumer protection. Reliance was placed on Professor Gentili’s assertion that Prato’s approach would lead to there being a right of withdrawal in circumstances where there was no need for such protection. This, too, fails to address the passages in Mediolanum in which the Joint Chambers make it clear that it is the risks associated with execution of contracts off-site that justify the protection afforded by articles 30.6 and 30.7. What is said by the Joint Chambers in paragraphs 4[.2] and following is simply inconsistent with a notion that there should be an inquiry into the particular circumstances of the case in order to ascertain whether the investor was or was not in fact taken by surprise. Dexia further asserted that Professor Sciarrone Alibrandi had not relied on consumer protection in her reports. I am not sure that this is correct, but even if it were, she was asked about it in cross examination and placed reliance upon it in response to the questions that she was asked. It is plain from the judgment that in Mediolanum the Joint Chambers had consumer protection very much in mind.

Two other matters were raised in Dexia’s closing submissions. First, it was said that the proposals formulated by Dexia for Prato were “tailor-made”. Second, it was said that article 30.6 TUF was particularly unsuited to local authority investors, given the procedures which they were required to follow under local government law before entering the swaps. In both respects it may be thought that the outcome under article 30.6 and 30.7 is particularly hard on Dexia. Nevertheless I consider that Prato has established that these factors are legally irrelevant because the criteria following Mediolanum are clear. Indeed in Mediolanum the Joint Chambers recognised that inevitably the acknowledgement of greater protection in favour of the investor translates into a less advantageous position for the investment provider.

For all these reasons I conclude that under Italian law each of the swaps contravened article 30.6 TUF. It follows that in principle Italian law would entitle Prato to the benefit of article 30.7 TUF.

There is no contention that the allegations of contractual release and contractual estoppel can assist Dexia if article 30 TUF applies.

It is accepted by Dexia that article 30.7 TUF is a mandatory rule falling within article 3(3) of the Rome Convention. It follows that Prato’s defence to the claim succeeds.

In these circumstances it is unnecessary to consider other defences advanced by Prato. To the extent appropriate, I will consider other defences in a further judgment. For present purposes, I turn to Prato’s counterclaims and Dexia’s alternative claims.

As regards sums it paid on and after 30 June 2009, on the basis that swap 6 is invalid and/or unenforceable, Prato seeks restitution under Italian law, alternatively English law. Dexia asserts that Prato’s claim to restitution is governed by English law, and not Italian law. Dexia adds that Prato’s claim would be extinguished by part of sums which would be due from Prato under swaps 3, 4 and 5 if swap 6 is invalid.

Dexia advances alternative claims in relation to swap 1, swap 2, swap 4 and swap 5: if those swaps are invalid, Dexia seeks to recover the net differentials paid to Prato under those swaps. That claim has been pleaded as arising under English law. As I understand it, Prato says that any such alternative claims would be governed by Italian law. If Prato is right, Dexia recognises that it must apply for permission to amend its particulars of claim.

To the extent that the restitutionary counterclaim, and Dexia’s alternative claim for restitution, are governed by English law, a question arises as to whether the opposing party has a defence of change of position.

If English law applies, a further question arises in relation to Dexia’s alternative restitutionary claim. That question concerns limitation, for unless Dexia can bring itself within an exception to the ordinary rule, payments made by it before 7 December 2004 would not be recoverable.

As regards all these issues, the parties’ closing submissions concentrated on what the position would be if Prato were to establish its capacity defences. There was a brief mention in Dexia’s written closing submissions on the question of governing law to the possibility that swap 6 might be voidable rather than void, for example by reason of misrepresentation. However there was no analysis of what the position would be if, as I have held, article 30.7 TUF has the effect that each of the swaps is “null and void, with only the client having the right to enforce this provision”.

I do not criticise anyone in this regard. The closing submissions necessarily had to deal with a large number of possible permutations. The permutations which may arise as a result of my conclusion on article 30.7 TUF had the potential to be quite complicated, and to have an impact on the question of governing law, on the question of whether either party should be allowed to rely upon a change of position defence if English law applies, and on the question whether Dexia can bring itself within relevant provisions of the Limitation Act as regards payments made by it before 7 December 2004. Additional permutations may arise as to the status of what has happened under swaps 1 to 5, and of those swaps themselves. It would have been very difficult to have included all such permutations in the written and oral submissions made at the end of the trial.

Similarly, it does not appear to me that the closing submissions have addressed the potential interaction between my conclusion on article 30 TUF and the remaining counterclaims. Again, I make no criticism. Consideration of potential interactions would have been difficult without knowing the outcome. Among other things, if Prato’s restitutionary counterclaim succeeds, a question arises as to whether those additional counterclaims still arise, and if so, what impact the finding under article 30 TUF has upon them.

For the reasons I have given above, my conclusion is that Dexia’s main claim fails. I propose to make directions to enable me to consider further the precise effect of my reasons in this regard on Prato’s counterclaims and Dexia’s alternative claims. I will hear oral submissions in that regard and in relation to any other order which either party proposes.