Palladian Partners LP & Ors v The Republic of Argentina & Anor

This is a case which, as will appear, turns on the construction of a single provision (the so-called ‘Adjustment Provision’) in an admittedly long and complex contractual framework.

I have concluded that the Claimants, four institutional and corporate investors (Palladian Partners LP, HBK Master Fund LP, Hirsh Group LLC and Virtual Emerald International Limited - collectively, ‘the Claimants’), are right in what they have to say on that construction issue and, it follows, that it is unnecessary to go on to consider an alternative case advanced by the Claimants that the First Defendant (‘the Republic’) acted in bad faith in producing certain Gross Domestic Product (‘GDP’) data in March 2014.

I have concluded, indeed, not only that it is unnecessary to address that alternative case but that it would be undesirable to do so given that it only arises in the alternative, given furthermore the seriousness of the allegations which it entails and given that those allegations are levelled against a sovereign state. Accordingly, I say no more about this alternative case.

Not having to deal with the alternative case has the additional advantage that this judgment can be shorter than would otherwise be the case. This is because, combined, the written submissions ran, in total, to some 800 or so pages, with more than half of those submissions being taken up with the construction issue, including what each side has suggested is factual matrix. In truth, however, as will appear, I have not been persuaded that the factual matrix is either as substantial as has been suggested or, in any event, that it is as helpful as has also been suggested.

I shall endeavour in what follows to be as concise as possible whilst, in doing so, seeking to address as many of the matters raised as is sensible. I have decided, indeed, that this is a case, in which, despite the length of the submissions advanced, the issues which need to be determined in order to resolve the dispute do not require a judgment which is overly long. Suffice it to say, therefore, that if I do not deal with a particular point, it should not be assumed that I have failed to take it into account; on the contrary, I have considered every matter raised.

Before saying something about the factual background, I should mention that this judgment also addresses a discrete issue which has arisen between the Claimants and the Second Defendant (‘Bank of New York Mellon’ acting as ‘the Trustee’). Again as will appear, I have reached the conclusion that the Claimants are right in what they have to say on this issue also.

The claim relates to certain Euro-denominated securities issued by the Republic in two tranches, in the years 2005 and 2010 respectively, which are linked to its GDP (the ‘Securities’). Specifically, the claim turns on whether an obligation to make a payment under the Securities arose for Reference Year 2013, and all the years after it.

The Securities are governed by English law with an English jurisdiction clause.

There are large sums at issue. If the action succeeds, the Claimants estimate that the amounts potentially due under the Securities for Reference Year 2013 to the majority of holders not party to the action range from c. €758 million to €1.18 billion.

The Claimants originally raised their claims with the Trustee, and in May 2019 served a letter before action on the Republic. After the parties were unable to resolve their dispute in correspondence, a claim was issued on 9 August 2019. The Republic’s Defence was filed on 6 November 2019. This was followed by an application by the Republic for summary judgment. Cockerill J dismissed that application on 21 July 2020: [2020] EWHC 1946 (Comm).

The Securities were first issued as part of a major sovereign debt restructuring launched in 2005, in the wake of a national financial crisis of unprecedented scale. This followed the Republic having built up a considerable structural deficit during the 1990s, which it had funded through heavy borrowing on international capital markets after the Republic established a fixed monetary rule (the ‘convertibility’ regime) pursuant to which the Argentine peso was pegged to the US dollar at one-to-one parity, in response to a recurrence of hyperinflation and uneven growth in 1991-1992. However, following the Russian financial crisis in 1998, which shocked international financial markets, the Republic was forced to turn to domestic borrowing. The Republic, as a result, fell into recession in late 1998.

Two years later, in March 2000, voluntary domestic financing having dried up, the Republic signed a Standby Agreement with the International Monetary Fund (the ‘IMF’) for lending of US$7 billion.

A year later, in March 2001, the Standby Agreement was augmented to US$14 billion as a part of an international support package of US$40 billion.

Shortly after that, in June 2001, faced with unsustainable debt servicing costs, the then Minister of Economy, Mr Domingo Cavallo, launched a “Mega-Swap” of 52 government bonds with imminent payment obligations for new bonds with longer maturities. This involved postponing US$12.6 billion in cash payments for five years at an additional cost of US$22.1 billion, thereby significantly worsening the Republic’s debt sustainability.

In late November 2001, following a substantial run on deposits, Mr Cavallo introduced crippling withdrawal limits at Argentine banks (the “Corralito”) in an effort to retain enough money in the banking system for the Republic to pay the interest falling due on its sovereign debt. This saw people being forbidden from withdrawing more than the equivalent of US$200 weekly from their bank accounts and led to widespread protests.

As a result, President Fernando de la Rúa resigned on 21 December 2001. There then followed the short interim presidency of President Adolfo Rodríguez Saá, who announced that the Republic would suspend payment of over US$80 billion of debt. At that time, this was the largest sovereign debt default in history.

These developments were followed by an intense economic depression in 2002, which paralysed economic activity, pushed inflation above 40% and tipped more than 50% of the Argentinian population below the poverty line.

By mid-2002, with a debt-to-GDP ratio of more than 130%, it was clear that the Republic needed to restructure its sovereign debt, since it would have been impossible to restart payments on its bonds as they stood. It was in this context that President Duhalde, who succeeded President Rodríguez Saá, appointed a new Minister of Economy, Roberto Lavagna, on 27 April 2002. Mr Lavagna and his team were charged with undertaking a detailed debt sustainability analysis to assess the necessary level of debt reduction and to model the level of payments which the Republic could afford to make in the future.

On 25 May 2003, Néstor Kirchner was sworn in as President. President Kirchner made implementing an effective sovereign debt restructuring and achieving debt sustainability a cornerstone of his administration’s economic programme. He kept Mr Lavagna and his team in post at the Ministry of Economy (‘MECON’) in order to see through this important programme. At that time, it was “the highest priority” of MECON to restore stability to the financial system and to return Argentina to good standing in the financial markets. This required the Republic to ensure that its debt burden was sustainable.

Before launching the exchange offer in 2005, the Republic held over 80 meetings with bondholders, including with retail investors and their brokers and representatives, as well as numerous meetings with institutional stakeholders such as the IMF, the World Bank and government officials from various countries, to negotiate and develop its restructuring proposals.

It emerged from these meetings that certain groups of bondholders believed that the Republic’s economy would grow at a higher rate than that projected by MECON and that the Republic could, therefore, sustain higher levels of debt payments. Mr Katz explained in his evidence that some of the creditors thought that the Republic’s potential output could be higher than 3%. It was for this reason, he explained, that the Republic introduced the concept of GDP-linked securities, which would provide bondholders with additional payments if the Republic did, indeed, grow at a higher than projected rate.

The first iterations of the proposed GDP-linked securities contained only one payment condition, which was that the level of GDP in the relevant year should be above what was known as the base case scenario for that year. However, it became clear that such a condition would not necessarily restrict payments to years when GDP growth was above trend. It was for this reason that the Performance Condition was introduced.

Once the proposed debt exchange offer was finalised, the Republic embarked on a series of roadshow presentations, explaining the “pre-determined growth trendline” by reference to which payments would be made under the GDP-linked securities.

In January 2005, and again in April 2010, the Republic launched voluntary debt exchanges, allowing creditors to trade in 152 different series of non-performing bonds for new types of performing bonds with lower interest rates, reduced principal and/or longer maturities, which would be more sustainable for the Republic over the long term.

Together, the 2005 and 2010 debt exchanges represented the largest sovereign debt restructuring in history. Approximately 76% of the aggregate value of eligible securities was exchanged under the 2005 debt exchange, and the 2010 debt exchange raised the participation rate to over 91%.

Of the 13 new securities issued, four were GDP-linked securities: (i) US-Dollar denominated GDP-linked securities (governed by New York law); (ii) US-Dollar denominated GDP-linked securities (governed by Argentine law); (iii) Euro-denominated GDP-linked securities (i.e. the Securities); and (iv) Argentine Peso-denominated GDP-linked securities. The same terms (albeit governed by different systems of law) applied to each of the GDP-linked securities. There were, in addition, Japanese Yen-denominated GDP-linked securities.

The GDP-linked securities were initially attached to the discount, par and quasi-par bonds, which made up the primary offering. After 180 days, the GDP-linked securities became detached from the underlying bonds and were traded independently in the secondary market.

Since their issuance in 2005, the Republic has made payments under the Securities totalling approximately US$10 billion. The Republic made no payments in relation to Reference Years 2009 and 2012 when (as is common ground) its growth rate was respectively 0.85% and 1.9%.

It will be appreciated that the dispute which has given rise to the present proceedings concerns Reference Year 2013. The particular context in which that dispute has arisen concerns something known as a “rebasing” from base year 1993 to base year 2004. It is convenient, however, to address this later once various GDP-related concepts (including rebasing) have been explained.

Before turning to the Securities, I should say something concerning the witnesses who gave evidence at trial, albeit only briefly given that I am not proposing to address the Claimants’ alternative (bad faith) case and only some of the factual witnesses gave evidence which appropriately falls to be addressed in relation to the construction issue. I am conscious that it would not be helpful were I to get into in any detail in relation to evidence given by witnesses in relation to that alternative case if it were later to prove necessary to give a judgment which addresses that case.

I start with the evidence which was given by the Republic’s factual witnesses.

Mr Gustavo Rodríguez had worked on Argentina’s national accounts in various roles since 1988, when he joined the Central Bank of Argentina as head of Division in the National Accounts Directorate. Between February 2008 and June 2012, he was Director of Consumption, Investments, and the External Sector within the National Directorate of National Accounts (the area within the Instituto Nacional de Estadística y Censos de la República Argentina (‘INDEC’) with responsibility for overseeing the production of the Republic’s GDP statistics). In June 2012, he assumed the role of National Director of National Accounts at INDEC. In this position, which he held until January 2016, he headed the National Directorate of National Accounts. As such, he gave evidence in respect of INDEC’s rebasing of GDP. That evidence was, accordingly, focused on the Claimants’ alternative (bad faith) case.

Dr Guillermo Nielsen was the Republic’s Secretary of Finance from May 2003 to November 2005. In that capacity, he led the restructuring of the Republic’s debt that led to the issuance of the Securities. He is now the Republic's Ambassador to Saudi Arabia and other countries in the Middle East.

Dr Nielsen was an impressive witness, who gave evidence on a number of matters. These included the rationale behind the Republic’s entry into the Securities. He explained, in particular, that the aim “was to restructure Argentina’s debt so that it was sustainable … The idea was simple: not overpromising what Argentina could pay and paying what Argentina said that it would pay on time”. He explained also that this is what led him to appoint a team to analyse the Republic’s debt profile and to model “the kind of payments that Argentina would be able to afford to make in the future”. Specifically, referring to a presentation which he gave to potential investors in Dubai on 22 September 2003, he said this in his witness statement:

“I am reminded by the slides accompanying my presentation in Dubai on 22 September 2003 (at pp.3-4) that in 2003, we initially assumed GDP growth of 5.3% for 2003 and then 4% for the 2004-2006 period. As those slides also show, our macroeconomic working assumptions in approaching the restructuring (at p.25) were that the Argentine economy would grow, on average, 3.8% for the next five years (i.e. 2004 through 2009) and 3% thereafter. These assumptions for the growth of Argentina’s economy, which projected repayment capacity, did not change very much from the inception of the process.”

Mr Sebastián Katz was Undersecretary of Economic Programming at MECON from February 2004 to April 2006. In that capacity, he reported to Dr Nielsen. Now employed by the Central Bank of Argentina, his evidence was primarily concerned with his role in conducting a debt sustainability analysis and macroeconomic modelling which was used as part of the 2005 debt restructuring. It was his evidence (to which I shall return) that it was this that led to the adoption of a 3% long-term growth rate as the threshold for the Performance Condition, explaining the concept of “potential GDP” which, he explained, underpinned the table in the definition of Base Case GDP.

Given the prominence which Mr Valentin KC gave to the evidence given by Mr Katz, it is worth setting out what, specifically, he had to say on this topic in paragraphs 12 to 28 of his witness statement.

He started with some background:

“When I arrived at MECON, discussions about a proposed restructuring of the Republic’s debt had already begun. The Republic’s aims in the restructuring were to align its payment commitments with the likely payment capacity of the Argentine economy. What we wanted was for Argentina to be able to enter a process of moderate but sustained growth, avoiding the recurrent episodes of volatility and pronounced macroeconomic imbalances which had characterized the growth of the Argentine economy in the past. …

The restructuring negotiations with creditors and other international institutions were relatively informal. The process was run by the Secretariat of Finance, which acted as the representative of Argentina in this context. I was not formally part of the negotiations, although I participated in certain meetings at the request of then-Finance Secretary Guillermo Nielsen, to explain the macroeconomic scenarios that supported the restructuring proposal to different groups of investors and creditors.”

He continued as follows:

“When I joined MECON in early 2004, I was briefed by my colleagues on the status of the discussions. I recall being advised that a key difficulty in the negotiations was that there were diverse groups of creditors with different views, some of whom expressed their opinions frequently and firmly. I understood from my discussions with colleagues on the negotiations team that the main disputed issue in the negotiations was that the Republic estimated that its GDP could grow at 3% per year, whereas some groups of creditors said that GDP could grow at 4.5 or 5% per year, which would increase the Republic’s payment capacity. We needed to bridge this gap and reassure the creditors that our projections were reliable. That was when the idea started to emerge that we could offer something similar to an equity instrument. The central idea was to design an instrument that would share the upside with the creditors if the economy did indeed perform better than the Republic’s projections. That is where the idea of the GDP-linked bonds came from. …

I do not remember the exact origin of this idea, or whether it was first proposed by the Republic or by one of the banks. However, I participated in the process of analysing that idea. The team under my charge was in charge of preparing the macroeconomic scenario, which was the analysis of the rate at which Argentina’s GDP could be expected to grow safely. I was not involved in the drafting of the GDP-linked bonds, but I saw many iterations of the instrument that the Ministry was starting to consider.

I also participated in technical discussions with a group of domestic and international banks, which advised the Republic through the process of developing the GDP-linked bonds. These banks, and the individuals who represented them, are identified in a presentation dated 11 March 2004, which I was shown by S&C, and I remember them well. My main interaction was with José M. Barrionuevo (who was then at Barclays) from the Economic Modelling Team that managed the sustainability and macroeconomic model of the group of advisory banks. Mr Maia, Mr Simpson and I had several meetings with him. The advisory banks and Economic Modelling Team had their own economic models but they interacted and exchanged views with me and with the team that I supervised. The local advisory banks in Argentina also had their own models, and Mr Sebastián Vargas of Banco Galicia was a member of those teams. I recall that Mr. Barrionuevo’s analysis was very aligned with our projections. Some of the external advisors started out with slightly different views. But, in the end, all the advisory banks were aligned with MECON’s projections. … .”

He, then, explained what specific role he played:

“Prior to my arrival at the Argentine Ministry of Economy, Mr Maia had been working with Mr Nielsen’s team on the macroeconomic analysis required for the proposed debt restructuring. In early 2004, I took over this process and built upon their work to generate a reasonable macroeconomic scenario under which the restructuring could take place. Mr Maia and his team remained involved in the process alongside me, and I was ultimately responsible for supervising the work that was done.

By a ‘reasonable macroeconomic scenario’, I mean a scenario that would allow for sustainable growth, without the extremely high volatility and recurrent episodes of over-indebtedness and inflationary acceleration of the past; a scenario in which payment capacity would arise from a sustainable growth process not punctuated by recurring episodes of crisis, so that the Republic’s financial commitments would be manageable.

In this context, it was important for the restructuring negotiations and policy decisions of the Argentine Ministry of Economy to be informed by a sustainability analysis, i.e. an assessment of the likely future growth of the Republic’s economy under various conditions and a corresponding analysis of the level of debt that it could reasonably be expected to bear.

My team prepared a reasonable macroeconomic framework for the restructuring, which emerged from our analyses of different economic scenarios. I have been shown a press release dated 27 April 2004 which refers to a ‘sustainability model’. Whilst I did not draft this press release, I understand the reference to a ‘sustainability model’ to be a reference to the macroeconomic framework prepared by my team. There was no formal ‘model’ as such (in fact, nothing that was prepared for publication); these were internal working models based on different macroeconomic scenarios.

The sustainability analysis involved considering the future evolution of the Argentine economy. This emerged from an analysis of long-term growth trends and a concept that economists call ‘potential output’ or ‘potential GDP’, based on the notion of production function. Potential GDP is a method of estimating the medium and long-term growth capacity of an economy. It is a calculation of the aggregate potential output capacity of the economy, based on the accumulation over time of factors of production (physical and human capital, labour, natural resources) and the evolution of the aggregate productivity of those factors, arising from technical progress. The production function takes account of these components and enables an aggregate growth rate to be determined in a context of full employment and efficient allocation of the factors. All of this stems from econometric analyses which may be carried out on the time series of the Argentine economy in order to estimate the rate of evolution of each of these components of the aggregate production function (i.e., the resources or factors of production and the total productivity of the factors).

Ultimately, we came to the conclusion that 3% was a reasonable growth rate for Argentina to commit to as a medium- to long-term projection. Of course, growth was expected to be higher in the years immediately following the recession because following an adverse economic shock, such as the one that Argentina had experienced and which had created a very significant negative gap between actual and potential GDP, short term growth is generally faster than the longer-term outlook. Following the years of economic recovery, we concluded that growth would then taper down to a stable 3%.”

He went on to explain this:

“When one looks at the past history of the Argentine economy, average growth during the 20th century as a whole is around 3%. Of course, the Argentine economy had good moments and bad moments during this period and the growth rate was higher or lower than 3% at different stages and in different growth cycles. But, over the course of a century, the average has not been much higher than 3% per year, and, indeed, in the last quarter of the last century the rate was lower. That is the reason why, provided the country deployed sound economic policies, 3% was considered a reasonable upper limit for sustainable growth for the purposes of Argentina’s medium- and long-term commitments.

This projected growth rate that we arrived at formed the base case growth scenario for the purposes of assessing the Republic’s payment capacity in the context of the restructuring. Our projections for what would constitute a future base case growth scenario were based on real growth and were not linked to any particular base year or series of prices.”

He continued by saying this:

“The draft prospectus for the restructuring was consistent, in its macroeconomic assumptions, with our internal working models that projected Argentina's capacity to pay. I remember that Mr Nielsen asked me to confirm this, which I did, because the prospectus reflected the estimates and the projections that we had made in the macroeconomic analysis team. S&C has shown me a note I drafted and sent to Mr Nielsen on 30 November 2004 to this effect. I remember this note and the analysis it was based on.

The base case growth scenarios which my team developed were also reflected in the roadshow presentations being delivered to creditors at that time. I attended some of these presentations and I remember travelling to the United States to attend a roadshow presentation in January 2005. S&C has shown me a slide deck from this presentation and the base case figures (on p. 19) correctly reflect what I remember of our work on the macroeconomic analysis team. There may possibly have been some final corrections to the decimals by the time the final prospectus was published - I have not checked whether these figures match up exactly - but the figures in this presentation are aligned with the projections we had and our thinking on the macroeconomic analysis team at that time.

I have also been shown an excerpt from a presentation on debt sustainability, which appears to be undated. Based on the typeface and the language used, I believe that I drafted it, although I do not specifically recall doing so. If I remember correctly, it is a presentation we made at a public seminar in Argentina in 2005 after the public presentation of the offer. I think the event was organised by the newspaper El Cronista Comercial at the Hilton Hotel or the Hyatt in Buenos Aires. The contents of the presentation are an example of my team's work on the macroeconomic scenario and reflect our view that 3% was a sustainable base case growth rate for the Argentine economy.

Based on my team’s macroeconomic analysis, I believed that Argentina would be in a condition to pay its debts following the restructuring. This was on the assumption, implicit in our analysis, that the macroeconomic environment, economic policies and the external scenario would not evolve in an adverse way.”

I found Mr Katz, like Dr Nielsen, to be an impressive witness. I am in little doubt that, in giving the evidence which he did, he was doing his best to assist the Court. What I am less clear about, however, as I later explain, is that his evidence amounts to appropriate factual matrix, at least to the extent to which Mr Valentin KC contended.

Mr Santiago Wright also gave evidence. He is currently an independent economist but previously he worked in the National Office of Public Credit at MECON between February 2000 and December 2020. In June 2007, he became Head of the Financial Information Unit before, then, becoming Head of the Financial Risks Analysis Unit in October 2014 and Director of Programming and Financial Information in October 2018. In those three roles, his responsibilities included determining whether payment was due under the Securities for any given Reference Year.

He gave evidence in relation to the application of the Adjustment Provision and the determination that no payment was due in respect of the 2013 Reference Year. That evidence is not, however, admissible in relation to the construction issue.

Mr Pablo López is currently the Minister of Treasury and Finance of the Province of Buenos Aires. However, from November 2013 to December 2015, he was Secretary of Finance at MECON, having previously held the positions of National Director of Macroeconomic Policy and a member of the Board of the Argentinian Central Bank. His role as Secretary of Finance was a senior role, at the Head of the Secretariat of Finance. He had responsibility for the ONCP, which was responsible for the calculations on the Securities.

Like Mr Wright, his evidence was focused on the Claimants’ alternative (bad faith) case; it was of no assistance in relation to the construction issue.

As for the Claimants, their only factual witness was Mr Fausto Spotorno, who is the Chief Economist at Orlando J. Ferreres & Asociados (‘OJF’), a corporate finance and economic consulting firm in Buenos Aires. He has shown interest in the Securities since around the time that they were first issued in 2005, having built one of the first models analysing their value (at his previous employer Delphos Investment). He, then, at OJF, wrote a report in February 2007 on their value, which he described as a milestone in his career. His evidence was similarly of no real assistance on the construction issue.

Turning to the experts, for the Claimants Mr Michael Davies was formerly Head of Economic Statistics at the Australian Bureau of Statistics and a member of various key expert groups, including for the IMF and the OECD, with a role providing Australia’s input into the overarching System of National Accounts (the ‘SNA’: an international standard maintained by the United Nations, the IMF, the World Bank, the Organisation for Economic Co-operation and Development (‘OECD’) and Eurostat) and acting as a consultant to various other countries, including for the Bank of Chile, various former members of the USSR, and the UK’s Office of National Statistics. He gave evidence both about GDP statistics and also about INDEC’s actual figures and processes and how those figures and processes compared to international practice.

Mr Valentin KC submitted in closing that Mr Davies’s evidence was notable for his dogged reluctance to accept that he had no practical experience of any aspect of a rebasing exercise such as the one undertaken by Argentina in the years leading to March 2014, given that, in 1998, Australia had switched from a fixed base year system for measuring GDP to chain-measured GDP. Much of his evidence, Mr Valentin KC went on to suggest, involved seeking to read across experience in other fields (points which had not appeared in any of his reports) to offer a personal evaluation of whether what INDEC appeared to have done was in line with his views of how matters might have been better conducted. It was also submitted that Mr Davies had a tendency to answer questions in a generalised or speculative way that was more in the nature of advocacy of the Claimants’ case than the objective evidence of an independent expert.

Although Mr Davies did display at least a hint of this tendency, my overall assessment is that he did his best to assist and to give independent expert evidence.

The same assessment applies to Professor Glenn Hubbard, the expert called by the Republic both to deal with GDP statistics and national accounts as well as with quantum. He, too, was a highly intelligent witness who, on occasion, strayed into the territory of advocacy which, perhaps counter-intuitively, entailed him, more than once, declining to express a professional opinion despite my directly asking that he do so.

As for the other experts, Dr Willem Buiter gave expert evidence on the Claimants’ behalf in relation to GDP-linked securities. He is a highly decorated economist, having held important academic positions at Princeton, Yale, Cambridge and Columbia, as well as major private sector roles (as Chief Economist at Citi) and, indeed, also public sector roles (as a member of the Bank of England’s Monetary Policy Committee). His evidence was considered and careful. It was also even-handed since he acknowledged from the outset that there are multiple possible ways of dealing with a rebasing in the context of the Securities and, in particular, that the Adjustment Provision could have been written in different ways.

Dr Eduardo Borensztein, who was called by the Republic as its GDP-linked securities expert, was equally impressive as a witness: fair and independent. He, too, was frank, acknowledging, in particular, that the words of the Adjustment Provision were not clear and did not assist in identifying an overlap year which, as will appear, the Republic’s construction requires there to be.

Lastly, Mr Richard Caldwell also gave evidence. He did so on behalf of the Claimants in relation to quantum. Mr Caldwell is a Principal of Brattle Group, an international consulting firm specialised in economic and financial analysis, and head of the firm’s global international arbitration practice. I am satisfied that, on the whole, he did his best to assist the Court. My only reservation concerns his unwillingness to accept that the analysis in his first report involved a mistake concerning his understanding of Professor Hubbard’s Inflation-Adjustment Method, specifically in failing to understand that in Professor Hubbard’s first report of GDP statistics the method used had involved taking GDP in (as will appear) what are known as 2004 base prices and deflating it to 1993 prices. That error resulted in what Mr Valentin KC rightly described as some strange calculations in Mr Caldwell’s first report, which were then corrected by Mr Caldwell in his second report but only with Mr Caldwell wrongly suggesting that Professor Hubbard was in some way responsible for the error which Mr Caldwell had himself made. I suspect that it was professional pride which saw Mr Caldwell adopt this approach but it would have been better if he had more readily acknowledged that it was not Professor Hubbard who had made the mistake.

The total issue in 2005 had a nominal value of EUR 11.8 billion and the total issue in 2010 was EUR 7.1 billion. The Securities mature in 2035.

There were three other series of GDP-linked securities issued at the same time (referred to above), two denominated in US dollars, and one in Argentine pesos. They were on substantially similar terms but not subject to English law and jurisdiction.

The structure of the Securities involved two key instruments: a Trust Indenture (dated 2 June 2005, although a Supplemental Indenture was executed in 2010) executed by the Republic and by the Trustee (‘the Indenture’), which set out various key terms in relation to the Securities (including in relation to defaults and remedies); and a Global Security issued on 2 June 2005, which contains the various Terms and Conditions which are at issue in these proceedings.

As will appear, the Securities contain a series of terms which link payment to GDP. In particular, there are three key conditions to a payment in any given year (the ‘Payment Conditions’), two of which (the ‘Level Condition’ and the ‘Performance Condition’) are linked to what is described as Real GDP whilst the third (the ‘Payment Cap Condition’) provides that total payments under the Securities shall not exceed 48% of the nominal amount of the Securities.

There are also provisions dealing with the calculation of a ‘Payment Amount’ to be paid if the Conditions are met, which is also linked to GDP.

Before setting out the provisions of the Securities which are relevant to the proceedings, it is convenient, first, to say something about GDP and, then, specifically, rebasing. There is no issue about what follows. Indeed, as will appear, much of the terminology is reflected in the Securities themselves.

GDP is a critical macroeconomic indicator that measures a country’s economic output. It is an estimate of total production of goods and services within a country’s borders over a given time period, most commonly a year. Expressed in the currency of the relevant country, GDP is, to repeat, only ever an estimate; measurements of GDP in different series are, therefore, different estimates of the same thing, and no estimate is the true or real economy. It follows that the way in which GDP is estimated is important, because the methodology fundamentally affects the figure.

GDP figures may be presented in current price (or nominal) and constant price (or real) terms.

Current Price GDP (or “GDP in current prices”) is the estimate of the value of the final goods and services produced using the current (contemporaneous) prices of those goods and services. It is an estimate of the present value of GDP.

Constant Price GDP (or “GDP in constant prices”) is expressed in the prices of a particular “Year of Base Prices” (‘YOBP’). The YOBP is sometimes referred to as a “base year”. It is used principally to compare GDP between different time periods and to measure growth between those time periods. To exclude the effects of price inflation/deflation and to produce comparable estimates, it is necessary to use a set of “constant” prices for the goods and services and a consistent measurement methodology. At the very highest level of abstraction, it involves the volume of goods and services produced in the relevant year being multiplied by the constant prices for those particular items. However, other complex methodologies are also applied. This is what the Securities call “Real GDP”.

In Argentina the base year for Real GDP at the time the Securities were issued was 1993.

Figures for different years in the same YOBP are commonly referred to as a “series” - so for Argentina’s Real GDP there is a “1993 series” which states Real GDP for all years from 1993 to 2012, measured in 1993 YOBP.

Nominal GDP is the measure of GDP calculated by measuring the current value of goods and services. In order to isolate the real changes in quantities of production, economists define real GDP, also known as “GDP in volume”, by designating a particular year as a base year and using the fixed prices of goods and services in the base year to calculate the value of goods and services in all other years.

In the Securities the Current Price GDP and Constant Price GDP of Argentina for a given year are referred to, respectively, as “Actual Nominal GDP” and “Actual Real GDP”. “Actual” is used to distinguish them from the benchmark numbers contained in the Securities.

An example put forward by Mr Valentin KC serves to illustrate the point quite nicely. If the base year is 2012, real GDP for 2022 would be calculated by using (i) prices of goods and services from 2012 and (ii) quantities from 2022. Nominal GDP would simply be the market value of the same goods and services but calculated using (i) current prices and (ii) quantities in 2022. By keeping prices constant, changes in real GDP capture changes in quantities, such that GDP is essentially a weighted average in which the weights are given by the prices of the base year. Accordingly, when analysing an economy’s production and growth from year to year, economists almost always limit consideration to real GDP.

Since Real GDP in a year is given by the quantity of goods and services in that year, valued at the prices of the base year, and nominal GDP in a year is given by the quantity of goods and services in that year, valued at current prices in that year, the difference between real and nominal GDP in that year is driven by changes in price level across total production.

The ratio of nominal GDP to real GDP at a point in time can be used to calculate a measure of the price level across total production known as a GDP deflator. In other words, by identifying and accounting for inflation, the GDP deflator does not purport to measure real economic performance. Rather, as Dr Buiter explained and was not in dispute, the deflator translates real economic performance into current prices for any particular year.

As I say, these concepts find expression in the Securities themselves, not only through the definition of “Actual Real GDP” (meaning, for any Reference Year, “the gross domestic product of Argentina for such Reference Year, measured in constant prices for the Year of Base Prices, as published by INDEC”) but through also the definition of “Actual Nominal GDP” (meaning, for any Reference Year, “an amount equal to Actual Real GDP for such Reference Year multiplied by the GDP Deflator for such Reference Year”) and the definition of “GDP Deflator” (meaning, for any Reference Year, “the number that results from dividing (i) the gross domestic product of Argentina for such Reference Year measured at the current prices of such Reference Year, as published by INDEC, by (ii) the Actual Real GDP for such Reference Year”).

As Mr Valentin KC observed, accurate, up-to-date measurement of GDP is a matter of real importance, for many reasons. As Mr Davies, the Claimants’ GDP statistics expert, put it in cross-examination, “It’s the main indicator used for determining policy: industry policy, fiscal policy, monetary policy. It’s probably the most important macroeconomic indicator. It is the most important single figure where you wrap up an enormous amount of information in one figure that can be graphed on one A4-page and make some sense. So if you want to know how an economy is travelling, it’s – and the more up-to-date, the more recent and the more accurate and the less biased, the better.”

Mr Valentin KC highlighted also how having up to date GDP measurement is important for international organisations like the IMF because, as Dr Buiter and Dr Borensztein agreed, “using an up to date base year of national accounts is conducive to accomplishing the economic functions of a GDP-linked security by tying payment obligations to a more accurate measure of the economy”. So, Mr Valentin KC submitted, an interpretation of the Adjustment Provision which results in using an outdated base year would not be conducive to accomplishing the economic functions of a GDP-linked security, because it would tie payment obligations to a less accurate measure of the economy, and thus detract from the intended function of the security.

Although only directly relevant to the alternative bad faith case which this judgment does not address, in Argentina, state measurement of GDP is undertaken by INDEC. INDEC published Nominal and Real GDP figures quarterly, approximately 3 months in arrears. This means that Q4 of a given year would be published in late March of the following year (as would GDP for the Full Year, which is computed by simply adding together GDP for the four quarters). INDEC also published a separate index called EMAE, which was produced monthly and was designed as an advance indicator of GDP growth. EMAE was expressed as an index, rather than a peso figure, which reflected that its principal purpose was to measure the change in economic activity (i.e. growth) from month to month.

Another concept of importance, for reasons which will become apparent, is that of ‘rebasing’ since it is appropriate periodically to update the YOBP by which Real GDP is measured in order to achieve a better measurement of the real economy through the making of adjustments for price, product and quality changes.

There can, in short, be distortions over time as the economy evolves and changes structure from the one represented by the base year. These distortions become more pronounced as time passes from the base year and the actual economy evolves further away from the year of base economy, notably because prices and composition of goods may have changed relative to each other, and certain goods may have experienced quality changes.

Rebasing also enables methodology updates and the incorporation of improved data sources or compilation methods to be introduced. In consequence, it is common ground between the experts that a rebased measure of GDP is likely to be a closer estimate of true GDP than an older measure.

The SNA recommends that:

“In general, constant price series should not be allowed to run for more than five, or at the most, ten years without rebasing.”

As such, when the Securities were issued in 2005 and 2010, it was known that Argentina would probably rebase GDP during the life of the Securities, but not to what year.

As to the effects of a rebasing, there are several points to note. The first is that two key components of the Real GDP estimate will change: the price configuration of goods and services (which will now be in the prices of the new YOBP) and the scope of GDP in the sense of what is measured and how it is measured. Secondly, the GDP experts were agreed that GDP Levels will be different between two different YOBP since the changes in price and weights, and the changes in methodology and data sources (or scope), in a rebasing result in different levels of Real GDP, with Real GDP often being higher in the new YOBP than the old YOBP.

Thirdly, the difference between Real GDP Levels in two different YOBP is not linear, the GDP Experts agreeing that the relationship varies from year to year and is not constant over time because the aggregate Real GDP figure is made up of many parts, each of which is measured differently and different sectors tending to grow at different rates so that their relative contributions to overall GDP will vary. It follows that measuring Real GDP in two different YOBP for Year A, the two estimates will be different. So, too, if the GDP for Year B or Year C is measured, the estimates will be different but not by the same amount or proportion.

Likewise, fourthly, and again as the GDP experts agreed, the growth trajectory of Real GDP in two different YOBP will generally not be the same, in that the relationship between Real GDP growth in two YOBP is not constant over time.

Moving on from the theory, as previously mentioned, the Republic, in due course, engaged in a rebasing process which saw 1993 YOBP replaced by 2004 YOBP.

Again, as previously mentioned, it was Mr Rodríguez who addressed this process in his evidence. In summary, however, it entailed two central information sources for the new base year being used: a national economic census and a survey of household income and expenditures. The results of the economic census and the survey were then reviewed and verified against a range of data from other sources in order to produce the necessary statistics for the new base year. Methodological changes are also implemented during the data processing stage.

The change to base year 2004 took several years to complete. The 2004 economic census data had been collected manually and required to be digitised and checked, a process which continued until the end of 2008.

By that time, the SNA had introduced significant methodological changes. INDEC sought not only to align with these new international standards but also to capture the very significant structural changes which the Republic’s economy had undergone between 1993 and 2004, including the collapse of the convertibility regime and the shutdown of international debt markets, which had pushed the Republic to produce many of the goods it had previously imported.

As INDEC worked on these matters, the 2004/2005 household income survey became outdated and had to be repeated. The new survey results became available at the end of 2013 and the first rebased GDP data (for 2012 and 2013) was published on 27 March 2014, followed by the remaining data (for 2004-2013) and a detailed methodology on 9 May 2014.

Between June 2011 and April 2015, the IMF issued a series of so-called K-1 reports, containing demands, questions and observations regarding the Republic’s CPI and GDP statistics. Prior to the Rebasing, the IMF considered that the Republic was in breach of its obligations under Article VIII, Section 5 of the IMF’s Articles of Agreement due to inaccurate reporting of both CPI and GDP data.

With respect to GDP statistics, in particular, the IMF’s principal concern was that the methodology for calculating GDP was outdated and that GDP was, therefore, overstated. In support of this conclusion, the IMF cited the wide gap between INDEC’s published GDP in the 1993 base series and the (much lower) estimates of private analysts and consultants. Thus, the IMF’s 8 June 2011 K-1 report required the Republic to undertake a series of remedial measures, including to “[u]pdate the base year of national accounts estimates to 2004, from the current 1993 base year” and to publish “revised time series of national accounts (at current and at constant 2004 prices)… reconstructed as far back as possible” by March 2012, with the first release of Q4 data for 2011.

Following further communications between the Republic and the IMF, the deadline for updating the base year was extended to July 2012 and, after that, to 17 March 2013. However, the IMF continued to press the Republic to speed up its processes.

On 17 September 2012, the IMF escalated its demands for the remedial measures to be completed, issuing a Statement of Concern regarding the Republic’s CPI and GDP statistics.

On 1 February 2013, when it was clear that INDEC could not meet the March 2013 deadline, the IMF issued a Declaration of Censure against the Republic, setting a 29 September 2013 deadline for the rebasing and stating in the Annex to the Declaration of Censure that, should the Republic fail to adopt the remedial measures within the specified timeframe, the IMF could adopt “a declaration of ineligibility to use the general resources of the Fund under Article XXVI, Section 2(a) of the Articles of Agreement”.

This led, according to Mr Rodríguez, to a redoubling of effort on the part of INDEC, and the Republic alerting the IMF to the fact that it would not be able to meet the 29 September 2013 deadline because the data would not be ready but making it clear that it had been making progress towards the publication of GDP data in the new base year by March 2014.

In the IMF’s 13 November 2013 Rule K-1 report, the Managing Director noted that, although the Republic had not adopted all of the remedial measures within the timeframe set out in the Declaration of Censure, it had provided an explanation as to why it had failed to do so. In view of the progress made by the Republic, the Managing Director recommended that the Executive Board refrain from taking any further action under Article VIII, Section 5 of the IMF Articles of Agreement at that time, proposing that an overarching deadline of 30 September 2014 be set for the Republic to complete the rebasing of its national accounts.

However, on 3 December 2013, a group of IMF Executive Directors (including the Executive Directors for the US and Germany) produced what is known as a Gray Statement demanding that the deadline for the remedial measures relating to the rebasing be brought forward to 31 March 2014, failing which there should be “immediate consideration of a Declaration of Ineligibility to use Fund resources”.

Nonetheless, on 6 December 2013, after the IMF had, on 4 December 2013, asked the Republic to identify which CPI and GDP measures then listed for completion by September 2014 could be completed by March 2014, the IMF Managing Director concluded that it would not be “feasible to advance the deadline for any of the four specified actions proposed for end-September 2014 to end-March 2014 due to the time required for verification”, but that there should be “a new specified action on the public release of revised GDP data with an updated base year be added for end-March 2014, in keeping with the authorities’ proposed schedule”.

It is the Republic’s case that it was pursuant to the deadline set in the December 2013 K-1 report that, on 27 March 2014, INDEC published GDP data for 2012 and 2013 in the rebased 2004 GDP series. This was followed by publication of a full set of national accounts some six weeks later, on 9 May 2014, along with the detailed methodology for the rebasing which had been carried out.

Following the March 2014 publication of “preliminary” GDP data in the 2004 base year, further releases of “provisional” and “provisory” data were made in June and September 2014, as was customary in the Republic following the rebasing. According to this data, GDP growth in 2013 in 2004 prices was 2.93%.

In compiling the 2014 public debt budget during August and September 2013, MECON had made provision for a payment under the Securities in respect of Reference Year 2013, based on the limited macro-economic projections available at that time, which had been prepared in 1993 prices. However, following the rebasing, the assessment was made by the Republic that no payment was due.

The determination that no payment was due in respect of the Reference Year 2013 was made by the Secretariat of Finance based on analysis undertaken by the Financial Risks Analysis Unit within the ONCP. It was then signed off by the Minister of Economy, who was not involved in the underlying analysis or the calculations, but was briefed on the approach taken and approved the determination made by the Secretariat of Finance.

Once INDEC had adopted a new base year, it ceased measuring or publishing GDP data in 1993 prices (following standard international practice). Initial GDP data in 1993 prices was published up to and including the third quarter of 2013. The last year for which GDP in 1993 prices exists is fiscal year 2012.

Against this background, I now turn to the provisions of the Securities which matter for present purposes, starting with the three Payment Conditions as set out in Clause 2(b) of the Terms and Conditions.

That provision is in these terms:

“Notwithstanding anything to the contrary hereunder, Holders of this Security shall not be entitled to receive any payment pursuant to this Security in respect of any Reference Year unless (i) Actual Real GDP for such Reference Year is greater than Base Case GDP for such Reference Year [the Level Condition], (ii) Actual Real GDP Growth for such Reference Year is greater than Base Case GDP Growth for such Reference Year [the Performance Condition], and (iii) the aggregate amount of all payments made by the Republic hereunder, when added to the amount of such payment, does not exceed the Payment Cap [the Payment Cap Condition].”

It can be seen, therefore, and as previously mentioned, that there are three conditions and, furthermore, that the Base Case GDP is critical to the first two of those conditions, namely the ‘Level Condition’ and the ‘Performance Condition’.

Specifically, the Level Condition involves comparing Actual Real GDP to Base Case GDP. As to that, Actual Real GDP is defined, for any Reference Year, as “the gross domestic product of Argentina for such Reference Year measured in constant prices for the Year of Base Prices, as published by INDEC”, whereas “Base Case GDP” is defined as being:

“for any Reference Year, the amount set forth in the chart below for such year:

Reference Year	Base Case GDP (in millions of constant 1993 pesos)	Reference Year	Base Case GDP (in millions of constant 1993 pesos)
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019	287,012.52 297.211.54 307,369.47 317,520.47 327,968.83 338,675.94 349,720.39 361,124.97 372,753.73 384,033.32 395,554.32 407,420.95 419,643.58 432,232.88 445,199.87	2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034	458,555.87 472,312.54 486,481.92 501,076.38 516,108.67 531,591.93 547,539.69 563,965.88 580,884.85 598,311.40 616,260.74 634,748.56 653,791.02 673,404.75 693,606.89

provided that, if the Year of Base Prices employed by INDEC for determining Actual Real GDP shall at any time be a calendar year other than the year 1993, then the Base Case GDP for each Reference Year shall be adjusted to reflect any such change in the Year of Base Prices by multiplying the Base Case GDP for such Reference Year (as set forth in chart above) by a fraction, the numerator of which shall be the Actual Real GDP for such Reference Year measured in constant prices of the Year of Base Prices, and the denominator of which shall be the Actual Real GDP for such Reference Year measured in constant 1993 prices.”

Leaving to one side, for the moment, the wording which follows the word “provided”, not least because that is the crucial wording for the purposes of these proceedings, what is important to note for now is that the chart sets out a level of Base Case GDP, expressed “in millions of constant 1993 pesos”, for each Reference Year until 2034 (in other words, Real GDP figures, expressed in 1993 YOBP) which can be compared to Actual Real GDP for the relevant year. Both GDP Securities Experts agree that Base Case GDP represents a “baseline path of real GDP” for the life of the Securities, so reflecting something close to “expected future real GDP in 1993 Base Year Prices”.

As to the Performance Condition, this involves comparing Actual Real GDP Growth and Base Case GDP Growth, the relevant definitions being as follows:

“‘Actual Real GDP Growth’ means, for any Reference Year, the percentage change in Actual Real GDP for such Reference Year, as compared to Actual Real GDP for the immediately preceding Reference Year; …”.

“‘Base Case GDP Growth’ means, for any Reference Year, the percentage change in Base Case GDP for such Reference Year, as compared to Base Case GDP for the immediately preceding Reference Year, except that solely for purposes of determining Base Case GDP Growth for Reference Year 2005, the Republic shall assume a Base Case GDP for the year 2004 equal to Ps.275,276.01 (in millions of constant 1993 pesos).”

Where there has been a rebasing, that comparison needs to take into account what the Actual Real GDP Growth definition goes on to say, namely:

“provided that, if the Year of Base Prices employed by INDEC for determining Actual Real GDP for such Reference Year and for the immediately preceding Reference Year shall differ, then Actual Real GDP for the immediately preceding Reference Year shall for this purposes be measured using constant prices for the Year of Base Prices applicable to the Reference Year in respect of which Actual Real GDP Growth is being determined.”

Accordingly, as the Republic accepted by the time of trial, the role of Base Case GDP in both the Level Condition and the Performance Condition is affected by a rebasing, in particular given the wording following the word “provided” in the definition of Base Case GDP (after the table): the Adjustment Provision. Put differently, it is common ground that the structure of each of the Level Condition and the Performance Condition means that, following a change to the YOBP, the relevant comparisons (of level and growth in Real GDP) require a comparison of Real GDP (level and growth) in the new YOBP, to Base Case GDP (level and growth) as adjusted by the Adjustment Provision. The Adjustment Provision is, in short, critical.

The third condition, the Payment Cap Condition, is not important for present purposes and is not impacted by the Adjustment Provision.

The other aspect which matters for present purposes is the Payment Amount. This is a freestanding concept which exists whether or not the Payment Conditions for any given Reference Year are met. Accordingly, in a given Reference Year, the Payment Amount can still be computed even if one or more of the Payment Conditions is not met.

The terms and conditions provide as follows in relation to the Payment Amount:

“for any Payment Date, an amount equal to (i) the Available Excess GDP (converted into euro) for the Reference Year corresponding to such Payment Date, multiplied by (ii) the notional amount of this Security outstanding as of such Payment Date; …”.

“Available Excess GDP” is defined elsewhere as being 5% of “Excess GDP” for the given Reference Year multiplied by a “Unit of Currency Coefficient” (0.015387), whilst “Excess GDP” is defined as being “the amount (expressed in billions of Argentine pesos), if any, by which Actual Nominal GDP for such Reference Year exceeds the Nominal Base Case GDP for such Reference Year”.

Excess GDP is, accordingly, a nominal (current prices) GDP concept, which is based on comparison of Actual Nominal GDP to Nominal Base Case GDP. It follows, as Ms Prevezer KC observed, that the key payment triggers (the Level Condition and the Performance Condition) are determined by reference to Real (constant prices) GDP but that the amount of any payment is determined by reference to Nominal (current prices) GDP. The Securities are, as such, something of a hybrid.

The provision as to Payment Amount, then, goes on as follows:

“The Payment Amount shall be determined by the Ministry of Economy on the Calculation Date preceding the relevant Payment Date. All calculations made by the Ministry of Economy hereunder shall be binding on the Trustee, the Registrar, the trustee paying agent and each other trustee paying agent and all Holders of this Security, absent bad faith, willful misconduct or manifest error on the part of the Ministry of Economy.”

The “Calculation Date” is 1 November of the following year, and the “Payment Date” is 15 December of the following year.

To repeat, this case turns on the construction which the Court concludes is appropriately to be afforded to the Adjustment Provision.

In closing, both Ms Prevezer KC and Mr Valentin KC sought to characterise the other’s submissions as entailing what might politely be described as a lack of realism. Thus, Mr Valentin KC suggested that the Court needs to decide between a construction of the Adjustment Provision which, as he put it, “writes the Adjustment Provision out of the Securities entirely and ties the Payment Conditions and the Payment Amount to a historic and outdated year of base prices and which writes in an unprecedented obligation to continue publishing outdated GDP figures for the term of the Securities (the Claimants’ interpretation)” and a construction “which gives effect to the agreed economic functions of GDP-linked Securities, as well as meaning to the language in the Adjustment Provision by adjusting the Base Case GDP levels specified in the Securities to account for the fact that the way in which GDP is estimated has been updated by the rebasing, but preserving the pre-set Base Case GDP Growth percentages which were chosen and fixed at the time of issue of the Securities and used to set the original Base Case GDP levels (the Republic’s interpretation)”.

Ms Prevezer KC, for her part, submitted that the “essence of its [the Republic’s] case appears to be that the Adjustment Provision cannot possibly mean what it says, because no rational investor or issuer could ever want to do anything other than tie the bargain entirely to the New YOBP - even if this tilts the commercial balance”.

The Claimants, through Ms Prevezer KC, say that the Adjustment Provision is clear both in its terms and in its effect. Labelled the ‘Annual Adjustment Construction’, the Claimants’ case is that, following a rebasing, Base Case GDP for each Reference Year is adjusted by a ratio, being the ratio of Actual Real GDP in such Reference Year measured in the New YOBP, and Actual Real GDP in such Reference Year measured in 1993 YOBP. Ms Prevezer KC submitted, apparently on the basis that a process of construction is not in all cases required, that so clear is the wording of the Adjustment Provision that the Claimants’ construction does not really entail a construction at all but merely a reading of words which should be afforded their plain and natural meaning.

Ms Prevezer KC submitted that, furthermore, the Annual Adjustment Construction sees the Adjustment Provision give effect to what she described as “entirely sensible economic theory”, taking into account the key GDP concepts to which reference has already been made but, most importantly, the fact that different YOBP series have different scopes or bases (the methodology of what is measured and how), different price configurations (weightings) and different price levels, meaning that different series will produce different Real GDP levels and growth rates and that there is a non-linear relationship (the ratios between different GDP series will vary from year-to-year, often materially).

The Republic’s case, the ‘One-Off Overlap Construction’, entails a one-off adjustment to the levels of Base Case GDP, rescaling the entire series into the new year of base prices. This, Mr Valentin KC submitted, preserves the pre-set percentages for Base Case GDP Growth which were chosen as part of the creation of a sustainable debt programme for Argentina, those pre-set percentages having been used to calculate the Base Case GDP levels in the Securities which were specified in 1993 base prices as that was the measure of GDP at the time of issuance.

Specifically, the One-Off Overlap Construction requires that Base Case GDP for each Reference Year is adjusted using a single fixed fraction applied to every year going forward - this is sometimes referred to as a ‘Fixed Fraction Approach’ (or ‘Fixed Factor Approach’ or ‘Constant Factor/Fraction Approach’). This construction says that such a Fixed Fraction is to be determined based on ‘the last Reference Year in which all necessary data is available’ (‘the Overlap Year’).

The numerator of the fixed fraction is Actual Real GDP for the Overlap Year in the New YOBP (here, 2004 YOBP) and the denominator of which is the Actual Real GDP for the Overlap Year in the prior YOBP (here, 1993 YOBP). The Republic says that the Overlap Year in this case is, therefore, 2012 and, accordingly, Base Case GDP for Reference Years 2013, 2014, 2015 etc (until any further rebasing) should be adjusted by the ratio derived for Reference Year 2012, rather than separately by reference to the ratio for that Reference Year.

The Republic, in short, takes the position that the Adjustment Provision should be regarded as though it stated as follows:

“… provided that, if the Year of Base Prices employed by INDEC for determining Actual Real GDP shall at any time be a calendar year other than the year 1993, then the Base Case GDP for each Reference Year shall be adjusted to reflect any such change in the Year of Base Prices by multiplying the Base Case GDP for such Reference Year (as set forth in chart above, or as previously adjusted) by a fraction, calculated for the last Reference Year for which official INDEC data is available, the numerator of which shall be the Actual Real GDP for such Reference Year measured in constant prices of the Year of Base Prices, and the denominator of which shall be the Actual Real GDP for such Reference Year measured in constant 1993 prices (or, if INDEC has effected more than one change, the previous Year of Base Prices).”

The method of adjustment under the Republic’s interpretation, Mr Valentin KC went on to submit, ensures that, following a rebasing, Base Case GDP is updated in such a way that the Securities remain linked to real world GDP, so that when the Republic’s economic performance is strong (as measured by the most reliable estimate of GDP available) it will be required to pay under the Securities (and vice versa). In contrast, Mr Valentin KC submitted, the Claimants’ Annual Adjustment Construction calls for the recalculation of Base Case GDP by a different factor each year, with the consequence that Base Case GDP Growth for the Performance Condition would inevitably vary widely for each Reference Year. This would mean, Mr Valentin KC observed, that the Securities become divorced from the real economic performance of Argentina and instead are tethered for their 30-year term to an outdated and obsolete measure of GDP (in 1993 base year prices), which, therefore, also requires the continued publication of that outdated measure of GDP until 2035 (an obligation which is not found anywhere in the text of the Securities). As a result, he explained, the Republic would find itself having to make payments under the Securities even if the economy is growing weakly or is in recession.

The Republic, however, has an alternative case. This is that the Adjustment Provision operates not by reference to figures in constant 1993 prices published by INDEC, but by a method whereby the figure published yearly in the New Year of Base Prices is adjusted for inflation only. This construction, described as the ‘Hubbard Deflator Construction’, entails the replacement of the Adjustment Fraction Actual Real GDP for the relevant Reference Year in 1993 Year of Base Prices with an Actual Real GDP figure in 2004 Year of Base Prices which has been deflated.

As has now become the norm, both Ms Prevezer KC and Mr Valentin KC advanced submissions by reference to a wealth of authority dealing with the process of contractual construction. It is unnecessary, in the circumstances, to rehearse all of those submissions. I prefer, instead, to focus on certain key points.

The first of these is that, as is very well known and as Lord Clarke explained in Rainy Sky SA v Kookmin Bank [2011] 1 WLR 2900 at [14]:

“the ultimate aim of interpreting a provision in a contract … is to determine what the parties meant by the language used, which involves ascertaining what a reasonable person would have understood the parties to have meant”.

Lord Hodge made essentially the same point in Wood v Capita [2017] UKSC 24 at [10] when he described the exercise as being to ascertain the “objective meaning of the language which the parties have chosen to express their agreement”.

Contractual construction has been described as a unitary or iterative exercise, which involves checking rival meanings against the provisions of the contract and testing the commercial consequences: see In re Sigma Finance Corpn [2009] UKSC 2, [2010] 1 All ER 571 at [12] per Lord Mance.

Although Ms Prevezer KC was, perhaps, minded to suggest that the iterative process is not engaged where there is no ambiguity in the words under consideration on the basis that in Rainy Sky Lord Clarke stated at [23] that “Where the parties have used unambiguous language, the court must apply it”, I consider that Mr Valentin KC was right when he submitted that still the iterative process should be engaged in even where there is no obvious ambiguity. Indeed, it is worth noting that what Lord Clarke had to say at [23] came after his saying at [21] that the exercise of construction is “one unitary exercise”.

Furthermore, again as Mr Valentin KC pointed out, in Rainy Sky Lord Clarke referred with approval, at [16], to Sir Simon Tuckey’s dissenting judgment in the Court of Appeal at [19], in which this was stated:

“The court must first look at the words which the parties have used in the bond itself. … If the language of the bond leads clearly to a conclusion that one or other of the constructions contended for is the correct one, the Court must give effect to it, however surprising or unreasonable the result might be. But if there are two possible constructions, the Court is entitled to reject the one which is unreasonable and, in a commercial context, the one which flouts business common sense.”

I agree with Mr Valentin KC that Sir Simon Tuckey was here describing needing to look at the words used in the contract more generally rather than merely the words used in the particular provision which is being construed.

It is also to be borne in mind that Lord Clarke went on at [25] in Rainy Sky to refer to what Lord Steyn had had to say when writing extra-judicially in ‘Contract Law: Fulfilling the reasonable expectations of honest men’ at 113 LQR 433, 441 and subsequently in Society of Lloyd’s v Robinson [1999] WLR 756 at page 763, namely:

“Loyalty to the text of a commercial contract, instrument, or document read in its contextual setting is the paramount principle of interpretation. But in the process of interpreting the meaning of the language of a commercial document the court ought generally to favour a commercially sensible construction. The reason for this approach is that a commercial construction is likely to give effect to the intention of the parties. Words ought therefore to be interpreted in the way in which a reasonable commercial person would construe them. And the reasonable commercial person can safely be assumed to be unimpressed with technical interpretations and undue emphasis on niceties of language.”

Moreover, in the Court of Appeal in In re Sigma Finance Corpn [2008] EWCA Civ 1303, [2009] BCC 393 at Lord Neuberger MR, whom Lord Mance in the Supreme Court at [12] described as being right to observe that the process is iterative, had this to say:

“Taking the provision on its own, there is considerable attraction in the Judge's view that the more natural meaning of the phrase is meaning (a) or (b). However, the provision must, of course, be construed not merely by reference to the language used, but also in its documentary and commercial contexts. Ms Prevezer QC, for party D, suggested that it was illegitimate to start by considering the effect of the language of the provision on its own. However, while one is seeking to interpret the document as a whole, the ultimate issue between the parties turns on the meaning of the provision, and, in order to resolve the issue, the reasoning and analysis have to start somewhere. The natural, indeed, I would have thought, the inevitable, point of departure is the language of the provision itself. However, where the interpretation of a word or phrase is in dispute, the resolution of that dispute will normally involve something of an iterative process, namely checking each of the rival meanings against the other provisions of the document and investigating its commercial consequences.”

Any lurking doubt over the matter is dispelled by what Lord Hodge has had to say in two subsequent Supreme Court cases.

First, in Arnold v Britton [2015] UKSC 36, [2015] AC 1619, after agreeing with what Lord Clarke had stated in Rainy Sky at [21], Lord Hodge said this at [77]:

“This unitary exercise involves an iterative process by which each of the rival meanings is checked against the provisions of the contract and its commercial consequences are investigated (Re Sigma Finance Corp ([2009] UKSC 2) [2010] 1 All ER 571, para 12 per Lord Mance). But there must be a basis in the words used and the factual matrix for identifying a rival meaning. The role of the construct, the reasonable person, is to ascertain objectively, and with the benefit of the relevant background knowledge, the meaning of the words which the parties used. The construct is not there to re-write the parties' agreement because it was unwise to gamble on future economic circumstances in a long term contract or because subsequent events have shown that the natural meaning of the words has produced a bad bargain for one side. The question for the court is not whether a reasonable and properly informed tenant would enter into such an undertaking. That would involve the possibility of re-writing the parties’ bargain in the name of commercial good sense. …”.

Secondly, in Wood v Capita Insurance Services Ltd [2017] UKSC 24, [2017] AC 1173, Lord Hodge said this at [10]:

“It has long been accepted that this is not a literalist exercise focused solely on a parsing of the wording of the particular clause but that the court must consider the contract as a whole and, depending on the nature, formality and quality of drafting of the contract, give more or less weight to elements of the wider context in reaching its view as to that objective meaning.”

He continued at [12] by saying this:

“This unitary exercise involves an iterative process by which each suggested interpretation is checked against the provisions of the contract and its commercial consequences are investigated: Arnold para 77 citing In re Sigma Finance Corpn [2010] 1 All ER 571, para 10 per Lord Mance. To my mind once one has read the language in dispute and the relevant parts of the contract that provide its context, it does not matter whether the more detailed analysis commences with the factual background and the implications of rival constructions or a close examination of the relevant language in the contract, so long as the court balances the indications given by each.”

He went on at [13] as follows:

“Textualism and contextualism are not conflicting paradigms in a battle for exclusive occupation of the field of contractual interpretation. Rather, the lawyer and the judge, when interpreting any contract, can use them as tools to ascertain the objective meaning of the language which the parties have chosen to express their agreement. The extent to which each tool will assist the court in its task will vary according to the circumstances of the particular agreement or agreements. Some agreements may be successfully interpreted principally by textual analysis, for example because of their sophistication and complexity and because they have been negotiated and prepared with the assistance of skilled professionals. The correct interpretation of other contracts may be achieved by a greater emphasis on the factual matrix, for example because of their informality, brevity or the absence of skilled professional assistance. But negotiators of complex formal contracts may often not achieve a logical and coherent text because of, for example, the conflicting aims of the parties, failures of communication, differing drafting practices, or deadlines which require the parties to compromise in order to reach agreement. There may often therefore be provisions in a detailed professionally drawn contract which lack clarity and the lawyer or judge in interpreting such provisions may be particularly helped by considering the factual matrix and the purpose of similar provisions in contracts of the same type. The iterative process, of which Lord Mance spoke in Sigma Finance Corpn (above), assists the lawyer or judge to ascertain the objective meaning of disputed provisions.”

As to whether there is ambiguity, I agree with Mr Valentin KC when he submitted that care needs to be taken to avoid adopting too narrow an approach, specifically that the process of ascertaining whether there is an ambiguity which gives rise to rival interpretations is not confined to a consideration of the contested contractual provision in isolation but requires the provision to be considered in its proper contractual and commercial context. It follows, I also agree with Mr Valentin KC, that ambiguity is not confined to linguistic ambiguity since there can be ambiguity of meaning, and so rival interpretations which need to be tested as part of the iterative process, even where there is no strict linguistic ambiguity, but where the meaning of the provision is nonetheless open to question.

As Lewison LJ put it in Napier Park Europe v Harbourmaster [2014] EWCA Civ 984 at [26], on appeal from a decision of the then Chancellor:

“Once an alternative reading emerges as a possible meaning, the interpreter must go on to consider which of two or more possible meanings is the more commercially sensible. The Chancellor did not do that, because he stopped at the first stage; namely by deciding that the language was clear and unambiguous. If the Chancellor had proceeded to the second stage, he should have preferred the interpretation that the downgrade had to be in force at the time when the Reinvestment Criteria had to be satisfied.”

Later, when discussing the iterative process required in construing a contract, Lewison LJ said this at [32]:

“The iterative process thus described is not confined to textual analysis and comparison. It extends also to placing the rival interpretations within their commercial setting and investigating (or at any rate evaluating) their commercial consequences. That is not to say that in a case like this the commercial setting should be derived from considerations outside the four corners of the contractual documents.”

He went on at [33] to say this:

“Thus we must seek to discern the commercial intention, and the commercial consequences from the terms of the contract itself; and that feeds in to the process of deciding whether a particular word or phrase is in reality clear and unambiguous. It follows in my judgment that, where possible, the court should test any interpretation against the commercial consequences. That is part of the iterative exercise of interpretation. It is not merely a safety valve in cases of absurdity. So much is, in my judgment, also made clear by the decision of the Supreme Court in Rainy Sky SA v Kookmin Bank [2011] UKSC 50; [2011] 1 WLR 2900. In that case Lord Clarke said at [20]:

‘It is not in my judgment necessary to conclude that, unless the most natural meaning of the words produces a result so extreme that it was unintended, the court must give effect to that meaning.’”

He, then, added at [36]:

“I do not therefore agree with Mr Snowden that commercial considerations have no part to play in deciding whether a particular interpretation is or is not ambiguous. Moreover, to say that ambiguity or unambiguity is the governing factor may be to miss the point. As Lord Sumption observed in Sans Souci Ltd v VRL Services Ltd [2012] UKPC 6 at [14]:

‘It is generally unhelpful to look for an “ambiguity”, if by that is meant an expression capable of more than one meaning simply as a matter of language. True linguistic ambiguities are comparatively rare. The real issue is whether the meaning of the language is open to question. There are many reasons why it may be open to question, which are not limited to cases of ambiguity.’”

That said, Lewison LJ concluded at [37] with this observation:

“On the other hand where the iterative process of interpretation as described above produces a clear answer a court must be very wary of assuming that it knows what is or is not commercially sensible: Skanska Rashleigh Weatherfoil Ltd v Somerfield Stores Ltd [2006] EWCA Civ 1732 at [22].”

This is a further important consideration since, as Ms Prevezer KC rightly noted, Lord Neuberger emphasised in Arnold at [17] that “commercial common sense and surrounding circumstances … should not be invoked to undervalue the importance of the language”, before going on to say this:

“The exercise of interpreting a provision involves identifying what the parties meant through the eyes of a reasonable reader, and, save perhaps in a very unusual case, that meaning is most obviously to be gleaned from the language of the provision. Unlike commercial common sense and the surrounding circumstances, the parties have control over the language they use in a contract. And, again save perhaps in a very unusual case, the parties must have been specifically focussing on the issue covered by the provision when agreeing the wording of that provision.”

Care needs also to be taken to avoid invoking hindsight. As Lord Neuberger put it at [19]:

“The third point I should mention is that commercial common sense is not to be invoked retrospectively. The mere fact that a contractual arrangement, if interpreted according to its natural language, has worked out badly, or even disastrously, for one of the parties is not a reason for departing from the natural language. Commercial common sense is only relevant to the extent of how matters would or could have been perceived by the parties, or by reasonable people in the position of the parties, as at the date that the contract was made. … .”

Ms Prevezer KC relied upon Napier Park in another respect, however. This is in relation to Lord Neuberger going on at [32] to cite from In Re Sigma Finance in the Supreme Court at [37], where Lord Collins said this:

“Consequently this is not the type of case where the background or matrix of fact is or ought to be relevant, except in the most generalised way. … Where a security document secures a number of creditors who have advanced funds over a long period it would be quite wrong to take account of circumstances which are not known to all of them. In this type of case it is the wording of the instrument which is paramount. The instrument must be interpreted as a whole in the light of the commercial intention which may be inferred from the face of the instrument and from the nature of the debtor’s business. Detailed semantic analysis must give way to business common sense …”.

She also referred to Barnardo’s v Buckinghamshire [2018] UKSC 55 where Lord Hodge said this at [14]:

“A pension scheme, such as the one in issue on this appeal, has several distinctive characteristics which are relevant to the court’s selection of the appropriate interpretative tools. First, it is a formal legal document which has been prepared by skilled and specialist legal draftsmen. Secondly, unlike many commercial contracts, it is not the product of commercial negotiation between parties who may have conflicting interests and who may conclude their agreement under considerable pressure of time, leaving loose ends to be sorted out in future. Thirdly, it is an instrument which is designed to operate in the long term, defining people’s rights long after the economic and other circumstances, which existed at the time when it was signed, may have ceased to exist. Fourthly, the scheme confers important rights on parties, the members of the pension scheme, who were not parties to the instrument and who may have joined the scheme many years after it was initiated. Fifthly, members of a pension scheme may not have easy access to expert legal advice or be able readily to ascertain the circumstances which existed when the scheme was established.”

I take these observations into account. I recognise that, as Ms Prevezer KC submitted, there are some contracts which are public documents specifying rights and duties which may be passed on to others who were not party to the original bargain, and that in such cases there is all the more reason to be cautious in engaging in the iterative process of construction. This does not mean, however, that such a process is not engaged in at all. Furthermore, I agree with Mr Valentin KC when he submitted that in the case of such types of contract there is no good basis for suggesting that the Court, in conducting that process, should give textualism greater weight than might be afforded in other cases, not least because ultimately in Napier Park the Court of Appeal eschewed an overly textual approach. Nor, again I agree with Mr Valentin KC, is the analogy with the pension scheme in Barnardo’s particularly apt since the Securities do not have the same characteristics as a pension scheme, if only because, in contrast to a pension scheme trust deed, the Securities do not “exist primarily for the benefit of non-parties” (see [15]).

All of the above having been said, as will shortly appear, in truth, the Republic’s case requires the Court to have regard to certain other aspects of the principles applicable to contractual construction since ultimately Mr Valentin KC was constrained, if not expressly then impliedly in the focus which his closing submissions had, that the Republic must persuade the Court that the Adjustment Provision should be construed so as to correct a clear mistake made in its drafting through the application of what Mr Valentin KC characterised as the ‘Chartbrook Principle’, namely that a clear mistake in the drafting of a document may be corrected as a matter of construction, if it can be established that something has “gone wrong with the language”: see Chartbrook Ltd v Persimmon Homes Ltd [2009] 1 AC 1101.

In fact, the principle is of older vintage since, as Ms Prevezer KC pointed out, it was addressed by Brightman LJ (as he then was) in East v Pantiles (Plant Hire) Ltd: [1982] 1 WLUK 562, as follows:

“Two conditions must be satisfied: first there must be a clear mistake on the face of the instrument; secondly it must be clear what correction ought to be made in order to cure the mistake. If those conditions are satisfied, then the correction is made as a matter of construction. If they are not satisfied then either the claimant must pursue an action for rectification or he must leave it to a court of construction to reach what answer it can on the basis that the uncorrected wording represents the manner in which the parties decided to express their intention.”

As to the first of these two criteria, Lord Neuberger MR explained in Pink Floyd v EMI [2010] EWCA Civ 1429 at [20], as follows:

“Further, as Lord Hoffmann also made clear in Investors Compensation [1998] 1 WLR 896, there is a difference between cases of ambiguity, which may result in giving the words a meaning they can naturally bear, even if it is not their prima facie most natural meaning, and cases of mistake, which may result from concluding that the parties made a mistake and used the wrong words or syntax. However, he emphasised the court does ‘not readily accept that people have made mistakes in formal documents’ - Chartbrook [2009] 1 AC 1101, para 23. He also pointed out in paragraph 20, that, as the court, and therefore the notional reasonable person, cannot take into account the antecedent negotiations, the fact that the natural meaning of the words appears to produce ‘a bad bargain’ for one of the parties or an ‘unduly favourable’ result for another, is not enough to justify the conclusion that something has gone wrong. One is normally looking for an outcome which is ‘arbitrary’ or ‘irrational’, before a mistake argument will run.”

This means, as Aikenhead J put it in WW Gear Construction v McGee Group Ltd [2010] EWHC 1460 (TCC) at [12], a “draftsman’s blunder” where “something has gone wrong with the language”. As Richard Salter KC put it in Altera Voyageur Production Ltd v Premier Oil E&P UK Ltd [2020] EWHC 1891 (Comm) at [67]:

“The court can, of course, do just that where it is ‘clear’ that something has gone wrong in the language which the parties have used. However, although I accept that it is undoubtedly possible that something has gone wrong here, given the various pointers to which I have already referred, it is not by any means clear to me that it has in fact done so.”

The same point was made by Nugee LJ in Monsolar IQ Ltd v Woden Park Ltd [2021] EWCA Civ 961, [2022] P & CR 10 at [25] when he said this:

“There was very little dispute about the law.  The principles applicable to the construction of written instruments in general, and contracts in particular, have been considered by the Supreme Court in a series of well-known cases, which it is not necessary to go over again.  Those authorities are largely concerned with the position where a contractual provision is open to two possible interpretations.  In the present case we are not concerned with such an exercise as it is common ground that the Formula, read with the relevant definitions, is clear and unambiguous and not open to two different interpretations. Rather we are concerned with the Chartbrook principle, under which the literal meaning of a provision can be corrected if it is clear both that a mistake has been made, and what the provision was intended to say.  This is in principle a different exercise from that of choosing between rival interpretations: see for example the recent decision of this Court in Britvic plc v Britvic Pensions Ltd [2021] EWCA Civ 867.”

Nugee LJ expanded on this at [31]:

“There is therefore a distinction between a case which concerns a provision which seems merely imprudent and one which appears irrational. The position was neatly summarised by Briggs LJ in Sugarman v CJS Investments Ltd [2014] EWCA Civ 1239 (‘Sugarman’) at [43]-[44] where he referred to the fine dividing line between a case where the result appears ‘commercially unattractive and even unreasonable’ and a case which appears ‘nonsensical or absurd’.”

He continued at [32]:

“As Mr Watkin submitted in his brief but cogent oral submissions, there is a consistent line of authority illustrating the sort of case that falls on the far side of the line. The language used by judges to describe such cases naturally varies but the concept is consistent. In City Alliance Ltd v Oxford Forecasting Services Ltd [2000] EWCA Civ 510 Chadwick LJ at [13] referred to the Court being satisfied that the words actually used ‘produce a result which is so commercially nonsensical that the parties could not have intended it’. In Chartbrook itself Lord Hoffmann referred variously to ‘an interpretation … sufficiently irrational to justify a conclusion that there has been a linguistic mistake’ (at [15]); to a ‘commercially absurd’ interpretation (ibid); to one that ‘makes no commercial sense’ (at [16]); to his not being able to believe that ‘any rational parties’ who wished to provide for a catastrophic fall in the market would have adopted the precise sum which the literal interpretation produced (at [19]); and to the interpretation adopted by the trial judge and majority of the Court of Appeal as not just producing provisions that were favourable to Chartbrook but as making the structure and language of the relevant provisions appear ‘arbitrary and irrational’ when the concepts could be combined in a rational way (at [20]). In Sugarman Briggs LJ at [43] referred to a case where the apparently unambiguous meaning of the words used ‘produces such a nonsensical result’ that it cannot be treated as expressing the meaning of the document.”

As to the second of the criteria, whilst, as Lord Hoffmann explained in Chartbrook at [25], “there is not, so to speak, a limit to the amount of red ink or verbal rearrangement or correction which the court is allowed”, nonetheless, as Lord Hodge put it in Arnold at [78], if the Court is satisfied that there was a clear mistake in the parties’ use of language, unless it is also clear what correction ought to be made, there cannot be a correction since the Court “must be satisfied as to both the mistake and the nature of the correction”.

It follows that, even if the Court were to conclude that the plain words of a provision could not reflect what the parties intended, it cannot correct by construction unless there is only one clear answer. This was the position in Trillium (Prime) Property GP Ltd v Elmfield Road Ltd [2018] EWCA Civ 1556, in which Lewison LJ declined to correct a rent review clause on three bases: first, because (as he put it at [15]) “if anything has gone wrong with the rent review provisions, it is a failure to think through the consequences of what the parties agreed, rather than any deficiencies in drafting”; secondly, because (again as he put it at [15]) “the alleged error may not lie in the rent review provisions at all”; and, thirdly, because (as he put it at [16]) “there is more than one possible solution to the alleged drafting error”, with the consequence that “I do not think that it can be said that it is clear what a reasonable person would have understood the parties to have meant”.

It is with these principles concerning construction in mind that I now endeavour to carry out the iterative or unitary process which is required (and, to the extent necessary, also consider the Chartbrook Principle), starting by saying something about factual matrix.

In their respective closing submissions, particularly the written submissions, both Ms Prevezer KC and Mr Valentin KC made reference to a considerable body of background material. Mr Valentin KC reminded the Court in this context that Lord Hoffmann made it clear in Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896 at pages 912H-913B that the admissible background to which the Court may have regard includes “absolutely anything which would have affected the way in which the language of the document would have been understood by a reasonable man”, subject only to the requirement that it should have been reasonably available to the parties, and to the exception that the parties’ previous negotiations and their declarations of subjective intent are excluded from the admissible background.

Ultimately, it was not entirely clear what each side was suggesting represents appropriate factual matrix. However, for my part, I regard relevant matters as including, first, the GDP-related concepts previously outlined together with, secondly, the fact that the Securities were issued against the background of a major economic crisis in Argentina which had been caused by an inability to keep up with debt repayments since I accept, in essence, that the fundamental purpose of the Securities was to ensure the sustainability of payments for the Republic to stabilise its financial position and prevent another default.

I am clear, thirdly, that the intention behind the Securities was that payments would only be made when the Argentine economy was growing at a sufficiently healthy rate, and that they would not be made if the Republic’s economy was not growing. This was made clear in the presentation previously mentioned given by Dr Nielsen at the Annual Meetings of the Boards of Governors of the World Bank Group and IMF which took place in Dubai on 22 September 2003. As that made clear, one of the “Key Restructuring Principles and Objectives” was to “reach a sustainable solution for its debt”, the first item identified by Dr Nielsen under “Basic features of a proposal” being that it “Needs to be based upon Argentina’s payment capacity over medium & long term”.

It is borne out also by two other presentations which were given by Dr Nielsen. The first was to the Emerging Market Traders Association (‘EMTA’) in New York on 4 December 2003, in which Dr Nielsen stated that “a country’s capacity to pay is the result of the interaction between primary surplus, growth, exchange and interest rates” and that “Of these variables, the most important one is growth”, before then adding that “Any debt restructuring can only be considered successful if the restructured liabilities can be matched by the repayment capacity of the debtor”.

The other was a presentation which he gave entitled “An update on Argentina” at the Merrill Lynch Latin American Investor Conference on 27 March 2004 and in which he stated as follows:

“We are now working on ways to solve this apparent conflict between sustainability and acceptability by offering debt enhancements, such as GDP indexed bonds, for example. An offer that includes GDP indexed bonds recognizes our responsibility to secure a sustainable debt restructuring with the potential upside benefits that a long-term recovery of the Argentine economy offers. They also evidence the good faith which we always pledged to follow.”

I am, however, unpersuaded that it is appropriate to look to other aspects relied upon by Mr Valentin KC, specifically material, including Mr Katz’s evidence as set out earlier, which he suggested shows that the parties proceeded on an assumption of long-term 3% growth trend in real GDP without regard to any particular base year, based on the historical performance of the Argentine economy.

In this respect, Mr Valentin KC, for example, highlighted a presentation published on MECON’s Website entitled “Consultative Working Group’s Meeting” in October 2003, which contained a slide headed “Payment Capacity Analysis: Macroeconomic Assumptions” and which identified long term real GDP growth as 3.0%.

He also drew attention to a presentation entitled “Argentina Bondholders Committee Restructuring Guidelines” dated 3 December 2003 likewise identified, among the “Key Medium/Long Term Assumptions”, “GDP Growth Target (2004-2030)” as 3.0%, and a MECON Press Release entitled “Eleven Basic Points on the Reduction of the Country’s Public Debt” dated 1 June 2004 stated as follows:

“A growth-linked component (‘GDP Unit’) is added to the aforementioned bonds, which will increase the payments derived from the bond menu presented each year if growth actually achieved by the country during the previous year exceeds that projected in the sustainability model; in this case, Argentina will share part of this surplus with the creditors.

The surplus that triggers the payment of this component is defined in the medium term as that which exceeds 3% annual growth.

These are not payments on estimates or prospects, but on growth actually achieved in future years (starting from the year 2005).”

In addition and consistent with the evidence which was given by Mr Katz (as well as Dr Nielsen at least to a degree), the Court’s attention was drawn to other roadshow presentations. One, entitled “Debt Exchange Offer” dated January 2005, noted that “Holders of this facility may receive up to 5% of available excess GDP over a pre-determined growth trend line if payment conditions are met” before listing “base case GDP” as “AR$287 mm in 2005 and a growth of approximately 3.6% in 2006, 3.4% in 2007, 3.3% in 2008, 3.3% in 2009, 3.3% in 2010, 3.3% in 2011, 3.3% in 2012, 3.2% in 2013, and 3.0% thereafter”.

Mr Valentin KC submitted, specifically, that, when the Securities were issued in 2005, Base Case GDP Growth was readily capable of being ascertained in advance for every Reference Year throughout their life by performing a simple arithmetical calculation using the figures for Base Case GDP contained in the table that forms part of the definition of Base Case GDP. He explained that the relevant percentages produced by that simple calculation followed a clear and predictable path (pre-set in advance and known to all): starting at 4.26% (in 2005), steadily declining each year to reach 3.29% (in 2009), 3.22% (in 2013), and then levelling off at 3% (from 2015 onwards until maturity).

Base Case GDP, Mr Valentin KC added, which is necessary for the assessment of the Level Condition, was similarly ascertainable in advance and clear: it was expressly identified for each Reference Year. Accordingly, Mr Valentin KC submitted, it is implicit in (and consistent with) the Terms and Conditions that, in the event of the adoption of a new Year of Base Prices, Base Case GDP Growth would continue to be capable of being ascertained in advance, and that it should not depart from the clear and predictable path previously applicable (i.e. levelling off at 3% from 2015 onwards until 2034).

There are a number of reasons why I am unpersuaded by Mr Valentin KC’s submission that there was an assumption that there would be long-term 3% growth which was common to both the Republic and investors.

First, there is nothing in the Securities or in the 2005 Prospectus which refers to a fixed percentage, still less anything with growth rates in something other than 1993 YOBP. It would be surprising, given this, if the Republic were right that the Adjustment Provision is to be understood as requiring the application of a fixed percentage. Had the intention been as the Republic suggests, then this is a significant omission. That applies, obviously, to the Securities. However, it applies also, with similar force, to the 2005 Prospectus since that (like the Securities) was a substantial document which, as was to be expected, set out in detail what was being offered. The notion that such a document would not explain that what was intended was that there should be a fixed percentage is implausible to say the least.

I might add that, although the 2010 Prospectus has a table at page S-49 which contains a column headed “Base Case Growth Rate (%)” with references to 3.29%, 3.26%, 3.22%, 3.02% and 3.0%, it is clear nonetheless from the previous column headed “Base Case (1993 pesos in millions)” that those growth rates are in 1993 YOBP. In any event, the 2010 Prospectus cannot assist as factual matrix given that the 2005 Securities were entered into five years earlier and it is not suggested that the 2010 Securities have a different meaning to those issued in 2005.

Secondly, in the Base Case GDP Table itself, the GDP levels are expressed in constant 1993 pesos (“Year of Base Prices” as defined under the Securities). There is no reference to growth rates in the table, and Mr Katz accepted that potential investors were not provided with any growth rate. It would, in such circumstances, be somewhat surprising if what mattered was not what was referred to (constant 1993 pesos) but what was not referred to (growth rates).

Moreover and thirdly, as Mr Katz also agreed, Base Case GDP was a series of GDP levels in 1993 YOBP with GDP growth being extrapolated as a percentage change from those levels; in other words, the growth rate was itself in 1993 YOBP. As Ms Prevezer KC reminded me in closing, that is consistent with the experts’ views that the growth rate is a function of the levels and not vice versa, Professor Hubbard noting (and Dr Borensztein agreeing) that “The growth rates will be derivative of the changes in the levels”.

Fourthly, in none of the presentations relied upon by the Republic, as previously described, was it stipulated that the Adjustment Provision should be understood as requiring the application of a fixed percentage. The most that can, perhaps, be said is that 3% was used for illustrative purposes. That, however, is not enough. In short, as Ms Prevezer KC submitted, whatever Mr Katz may have been doing internally, this was not communicated to creditors generally and as such does not amount to permissible factual matrix.

Lastly and fifthly, although Mr Katz gave evidence that he had mixed different YOBP for the purposes of his assessment of growth rates, his internal documents show those figures produced as 1993 YOBP figures. This is apparent, in particular and by way of example, from a so-called “Century” document which expressly states figures in 1993 YOBP, with the growth rates between each year then being derived by a formula in the table based on the change in levels, rather than the other way around.

I make it clear that I take the matters which I have identified as appropriate factual matrix into account in carrying out the iterative or unitary process which is required. This includes, again I make it clear, in considering whether the wording of the Adjustment Provision is ambiguous as Mr Valentin KC would suggest. I do not, therefore, merely focus on the words used but consider them in their proper context and by considering the commercial consequences of the competing interpretations.

In this respect it was Mr Valentin KC’s submission that the mathematical effect of the Claimants’ interpretation is that the Adjustment Provision becomes redundant because, on their interpretation, following a rebasing to a new year of base prices, the Level Condition and the Performance Condition continue to be satisfied if (and only if) they are satisfied at the 1993 YOBP. It follows, he went on to submit, that there is no need at all to compare Actual Real GDP in the new Year of Base Prices with adjusted Base Case GDP (despite what the definition of Actual Real GDP Growth requires), and for the purposes of determining whether the payment triggers have been met, the Adjustment Provision is superfluous.

In support of this proposition, Mr Valentin KC relied upon an annex to his written closing submissions, Part 1 of which, he submitted, demonstrated that, on the Claimants’ interpretation from the perspective of the 2004 Year of Base Prices, there is a stark drop of the Base Case GDP Growth figure for Reference Year 2013 and then no indication as to what Base Case GDP in 2004 base prices and Base Case GDP Growth measured against GDP in those prices will be for future years. Part 2 of Annex 1, then, he submitted, depicts the mathematical consequences and demonstrates how the Adjustment Provision is written out of the Securities altogether (as are the defined term “Year of Base Prices” and the adjustment language in “Actual Real GDP Growth”).

In supplemental written submissions, produced after the trial, Mr Valentin KC revisited this issue, describing the “short point” as being that the evidence demonstrates that the mathematical consequence of the Claimants’ interpretation is that the Adjustment Provision (as well as the equivalent adjustment language in the definition of “Actual Real GDP Growth” and the entire contractual concept of “Year of Base Prices”) serves no purpose. That, Mr Valentin KC submitted in paragraph 5 of the supplemental submissions, is because:

“… it leaves the position with respect to the Payment Conditions and the Payment Amount as if:

(1)

Actual Real GDP continued to be measured by INDEC (despite the rebasing to 2004 Year of Base Prices) in 1993 Year of Base Prices (contrary to the definition of ‘Actual Real GDP’ by reference to ‘Year of Base Prices’);

(2)

the Payment Conditions (which determine whether or not any Payment Amount is due) would be met if and only if they would be met in 1993 Year of Base Prices; and

(3)

‘Excess GDP’ (the measure of the Republic’s economic outperformance, which is required to be calculated for the Payment Amount) can be measured solely by reference to GDP in 1993 Year of Base Prices (albeit multiplied by a deflator, to take account of the effect of inflation).”

This, Mr Valentin KC submitted, illustrates that there is, at a minimum, an ambiguity in the Adjustment Provision and, more than this, that the Claimants’ approach to the construction of the Adjustment Provision cannot be right. In truth, as perhaps demonstrated by the fact that the Republic lodged further (and detailed) written submissions addressing the issue post-trial, this is an issue which lies at the heart of the Republic’s construction case. It is, however, not an issue about which I consider the Republic to be right as a matter of substance. Indeed, as I shall explain, given the view which I have reached concerning the substance, I am not even sure that Mr Valentin KC was right when he submitted that, at a minimum, the issue means that there is ambiguity concerning the wording of the Adjustment Provision.

First, whilst in his supplemental written submissions Mr Valentin KC picked up on Ms Prevezer KC’s description of the Adjustment Provision, on the Claimants’ construction, as being to “restore the like-for-like comparison” so that the “rebasing is neutral”, characterising the Claimants’ approach as being to ensure that rebasing is to be treated as irrelevant, this was not a stance which the securities experts, Dr Buiter and Dr Borensztein, adopted. On the contrary, Dr Borensztein expressly accepted in the course of cross-examination that the Adjustment Provision has an effect which is necessary for the operation of the Securities. Specifically, where there is a rebasing, the measurement of Actual Real GDP changes (in accordance with its definition) and the Adjustment Provision is thus required to adjust Base Case GDP so that there is a like-for-like comparison.

Secondly and probably in recognition of this, it is to be observed that in the supplemental written submissions dealing with the redundancy issue, Mr Valentin KC framed the point in another way. He said this in paragraph 6:

“Put differently, all three of these consequences of the Claimants’ interpretation would have arisen, if the Securities had simply provided that, even after a rebasing: Actual Real GDP will always be calculated in 1993 Year of Base Prices (irrespective of the actual Year of Base Prices INDEC uses to measure GDP); and no adjustment is to be made to Base Case GDP … .”

In other words, as Ms Prevezer KC observed in responsive written submissions, the Republic’s position was that the effect of the Claimants’ construction could have been achieved more directly by providing expressly that certain matters be calculated only by reference to 1993 YOBP. This is not quite the same as saying, as Mr Valentin KC had during the course of his oral closing submissions (and had done to date) that the Annual Adjustment Construction “essentially writes out” the Adjustment Provision.

Nor is this what was being said in paragraph 8 of the post-trial written submissions. There, Mr Valentin KC said this:

“The Claimants’ approach is to say that the role of the Adjustment Provision is to ensure that rebasing is to be treated as irrelevant, despite the fact that the definitions of ‘Actual Real GDP’ by reference to ‘Year of Base Prices’, and ‘Actual Real GDP Growth’ (with its proviso language), and the very fact of the inclusion of the Adjustment Provision in the definition of Base Case GDP, make it obvious that rebasing is highly relevant: after rebasing, it is the new Year of Base Prices that is to be used to measure Actual Real GDP, Actual Real GDP Growth and Base Case GDP. It simply makes no sense that the effect of applying the Adjustment Provision, and the equivalent provisos in the definitions of Actual Real GDP, Year of Base Prices and Actual Real GDP Growth, is the same as not applying these provisions at all. The Claimants’ construction deprives (i) the definition of ‘Actual Real GDP’ by reference to the further defined term ‘Year of Base Prices’, (ii) the proviso to the definition of ‘Actual Real GDP Growth’, and (iii) the Adjustment Provision in the definition of Base Case GDP of any practical effect, and the construction is therefore futile … .”

Again, what Mr Valentin KC was here submitting was not that, on the Claimants’ construction, the Adjustment Provision is redundant. On the contrary, what is stated recognises that the Adjustment Provision does do something, namely offset for certain purposes the effect of the change to Actual Real GDP arising on a rebasing. Mr Valentin KC’s submission was, rather, that the drafting could have achieved the same result by different means. As Lewison LJ, however, observed in Napier Park, “The fact that something should have been clearer doesn’t mean that it doesn’t have the meaning that it has”. That is the position here.

There are, however, as Ms Prevezer KC submitted, other difficulties with the Republic’s approach. Thus, there is no dispute that Actual Real GDP changes on a rebasing and becomes a figure in the new YOBP. It follows that what she described as the left-hand side of the equation (Actual Real GDP and Actual Real GDP Growth) does change on a rebasing.

Nor is it in dispute that Excess GDP and the Payment Amount are changed by a rebasing on the Claimants’ construction of the Adjustment Provision. Mr Valentin KC himself acknowledged this to be the case in the brackets at the end of paragraph 5(3), which qualify the assertion that Excess GDP can (on the Claimants’ construction) “be measured solely by reference to GDP in 1993 YOBP” by making the point that this is only the case if the figure is multiplied by “a deflator, to take account of the effect of inflation”. Since, however, there is no mention of such a “deflator” being used in this way anywhere in the Securities, still less an explanation as to how the “deflator” is to be calculated, the detailed calculations on which the Republic relies, based on Dr Buiter’s mathematics, do not matter and, as such, this point goes nowhere.

Turning, therefore, to the Adjustment Provision, I have reached the clear conclusion that the Claimants are right and that the Republic is wrong: in other words, that the Annual Adjustment Construction, rather than the One-Off Overlap Construction (or, for that matter as I shall come on to explain, the Hubbard Deflator Construction), is the appropriate construction.

I say this for what, ultimately, is a very simple reason. This is that, as Ms Prevezer KC submitted and as demonstrated by the fact that it requires the Adjustment Provision to be read as though it contained additional wording which it does not have, the One-Off Overlap Construction just does not reflect the wording used in the Adjustment Provision. In contrast, the Annual Adjustment Construction is faithful to the wording used.

Specifically, and with apologies for setting out the wording again, albeit with emphasis which to date has been absent, it is the use of the words “each” and “such” in the Adjustment Provision which should be noted, as follows:

“…if the Year of Base Prices employed by INDEC for determining Actual Real GDP shall at any time be a calendar year other than the year 1993, then the Base Case GDP for each Reference Year shall be adjusted to reflect any such change in the Year of Base prices by multiplying the Base Case GDP for such Reference Year (as set forth in the chart above) by a fraction, the numerator of which shall be the Actual Real GDP for such Reference Year measured in constant prices of the Year of Base Prices, and the denominator of which shall be the Actual Real GDP for such Reference Year measured in constant 1993 prices.”

I agree with Ms Prevezer KC when she submitted that the words “each” and “such” make it clear that the envisaged adjustment is made for each Reference Year and, furthermore, that the numerator and denominator in the adjustment fraction are based on the Actual Real GDP for that (“such”) Reference Year in the new YOBP, and such Reference Year in constant 1993 prices.

That is what the words say in terms: that there is an adjustment for “each” Reference Year and that the adjustment for any given Reference Year (hence the reference to “such Reference Year”) is performed by a fraction using Real GDP for that particular Reference Year in each of the new YOBP and 1993 prices.

That is why, as it seems to me, there are two references to “such Reference Year” following the first reference. To approach matters differently entails ignoring the fact that this wording is repeated.

To repeat, the One-Off Overlap Construction requires the Adjustment Provision to be read as though it contains the further words highlighted by the underlining below:

“… provided that, if the Year of Base Prices employed by INDEC for determining Actual Real GDP shall at any time be a calendar year other than the year 1993, then the Base Case GDP for each Reference Year shall be adjusted to reflect any such change in the Year of Base Prices by multiplying the Base Case GDP for such Reference Year (as set forth in chart above, or as previously adjusted) by a fraction, calculated for the last Reference Year for which official INDEC data is available, the numerator of which shall be the Actual Real GDP for such Reference Year measured in constant prices of the Year of Base Prices, and the denominator of which shall be the Actual Real GDP for such Reference Year measured in constant 1993 prices (or, if INDEC has effected more than one change, the previous Year of Base Prices).”

Mr Valentin KC, who accepted during the course of his opening submissions that “the words create a difficulty” yet in closing was minded to suggest that “the contractual scheme of the Securities … points unambiguously to the correctness of the Republic’s interpretation”, accepted also that it is only “if the concept is, and must be, to preserve the percentages, that you do a one-off adjustment, you have to construe those words in a way to achieve that” that the One-Off Overlap Construction comes into play at all.

Since I have already rejected the suggestion that the Adjustment Provision is to be understood as requiring the application of a fixed percentage, the Republic’s position is somewhat problematic. This is because I have decided, in essence, that the “concept”, to use Mr Valentin KC’s terminology, was not to “preserve the percentages”.

There are, however, also other reasons why I cannot accept that the Republic is right.

First, the additional words which are required for the One-Off Overlap Construction to work include, importantly, adding the “calculated for the last Reference Year for which official INDEC data is available” wording, yet this is a concept to which there is absolutely no reference in the Adjustment Provision. This is not a minor point; on the contrary, it is central to the construction which the Republic puts forward.

Secondly, adding this wording renders meaningless the use of the word “such” before the words “Reference Year (as set forth in chart …”) since, as can be seen, the One-Off Overlap Construction introduces a calculation based not on “such Reference Year (as set forth in chart …”) but, instead, on a different Reference Year altogether, namely “the last Reference Year for which official INDEC data is available”. It follows, as Ms Prevezer KC submitted, that, on the Republic’s construction, the word “such” bears two different and directly contradictory meanings in the space of one sentence: the first instance of “such” would refer to the relevant Reference Year in question (in the present case, 2013), whereas the second and third instances would refer instead to the Overlap Year (in the present case, on the Republic’s case, 2012).

Thirdly, the One-Off Overlap Construction, then, goes on to identify how that (unexpressed) Overlap Year is to be identified, but nowhere is the Overlap Year defined or identified in any way by the words of the Adjustment Provision.

Lastly, the One-Off Overlap Construction requires something else which is not mentioned in the provision. That is the insertion of a separate step for each rebasing. This, despite the fact that the wording says nothing at all about such a step being required. Indeed, it expressly requires something else in not merely the first rebasing but all subsequent rebasings: a direct comparison between Actual Real GDP in the new YOBP (i.e. the newest YOBP, the one adopted by the second or subsequent rebasing) with Actual Real GDP in 1993 YOBP for the relevant Reference Year. The One-Off Overlap Construction, in contrast, would entail, in any subsequent rebasing, a comparison between Actual Real GDP in the new(est) YOBP Actual Real GDP and Actual Real GDP in whatever was the YOBP which resulted from the previous rebasing.

This entails an entirely different comparison to be made to the comparison which is described in the Adjustment Provision. It is not the mere “lacuna” which Mr Valentin KC sought to portray it as being.

Furthermore, I agree with Ms Prevezer KC when she made the point that the “lacuna” only arises on the Republic’s construction since, with the Annual Adjustment Construction, no difficulty arises in relation to the multiple rebasings which would have been contemplated during the lifetime of the Securities. This is because that construction (unlike the Republic’s preferred construction) works for each and every rebasing in precisely the same way and without discrimination.

The fact that the One-Off Overlap Construction requires a different comparison each time that there is a rebasing is a strong indication that the One-Off Overlap Construction is not the right one since it is to be expected that whatever construction is applicable ought not to give rise to this situation, whether it is characterised as a mere “lacuna” or otherwise. The more so, given that the different comparison is not called for because of any words actually used in the Adjustment Provision.

For these various reasons, I am clear that the One-Off Overlap Construction does not work and that the Annual Adjustment Construction, which does work and which requires no departure from the wording of the Adjustment Provision, is to be preferred.

Nor, in my equally clear view, is it appropriate to correct the wording of the Adjustment Provision through application of the Chartbrook Principle since in the present case it is quite impossible to conclude that, in the drafting of the Adjustment Provision, something has “gone wrong with the language” or “draftsman’s blunder” and so that there has been the type of clear mistake which the Chartbrook Principle would require.

To repeat, Mr Valentin KC accepted that it is only if the concept is taken to be “to preserve the percentages, that you do a one-off adjustment” and “you have to construe those words in a way to achieve that”. However, I see no reason to suppose that the Republic is right in saying that the economic considerations dictate the premise for which Mr Valentin KC contended - in any event to the extent that the Court should regard itself as obliged to afford the Adjustment Provision a construction which is at odds with its express wording.

This is a matter in relation to which Mr Valentin KC and Ms Prevezer KC took diametrically opposed starting points, Ms Prevezer KC submitting that there are strong economic reasons why the Annual Adjustment Construction is right and Mr Valentin KC (no doubt with the Chartbrook Principle in mind) submitting that that construction is “economically absurd”. In truth and to adopt a description deployed by Mr Valentin KC in closing when seeking to criticise the Claimants’ case, both sides’ cases on economics entail a strong element of question-begging in that they each somewhat assume the very thing that each side wishes to establish in support of their respective favoured constructions.

That said, Ms Prevezer KC made three preliminary points, each of which seem to me to have substance and none of which was disputed by Mr Valentin KC. The first has already been touched upon and is that any interpretation of the Adjustment Provision needs to cater for potentially multiple rebasings over the 30-year life of the Securities. Secondly, in considering commerciality, the position needs to be viewed from both sides. Thirdly but linked to the last point, in considering the economic position, it is important to have in mind what Ms Prevezer KC described as the balance of economic risk and reward between the Republic and the holders of the Securities.

There are various reasons why I agree with Ms Prevezer KC in relation to her submissions concerning the economics.

First and perhaps most fundamentally, the effect of the Annual Adjustment Construction is that rebasing does not make a difference to whether or not the Performance Condition and the Level Condition are satisfied since the Annual Adjustment Construction ties the Payment Conditions to 1993 YOBP. In that sense, whatever the bargain was (and, in particular, however good or bad it may have been), the Annual Adjustment Construction preserves that bargain as regards the Payment Conditions. As a result, a creditor or investor who makes the decision at any point to accept or purchase the Securities, based on the known position in terms of base case in 1993 YOBP and published Actual Real GDP in 1993 YOBP, knows that a rebasing will not affect their assessment of whether payments are likely in the future.

With the One-Off Overlap Construction, on the other hand, the position is different in that a rebasing may mean that there is a liability to make a payment more or less likely in a manner which is unknown to the Republic. As Ms Prevezer KC put it, whereas the One-Off Overlap Construction entails the goalposts being moved by a rebasing, with all the uncertainty that that brings, the Annual Adjustment Construction does not do this: the goalposts remain in the right place and at the right distance from each other.

Although Mr Valentin KC submitted that it is hard to understand how there could be any practical use or value to investors in tying a bond that is supposed to pay out depending on the performance of the real economy to what he characterised as an outdated and superseded measure of the performance of the economy, I agree with Ms Prevezer KC that there are obvious reasons why investors may have considered it to be a good thing. Again, 1993 YOBP was a known measurement in 2005, having been in operation since 1999. As Dr Borensztein agreed, on the Claimants’ interpretation, the parties have certainty in advance as to what level and growth of Actual Real GDP is required to meet the Payment Conditions for the relevant Reference Year. Conversely, the parties had no idea what any rebased measurement would look like.

Secondly, there is the concept of non-linearity to consider. Although it is common ground that the relationships between measures of GDP levels and GDP growth trajectories in different years of base prices are not linear in the sense that the ratio between Real GDP in the two different YOBP in any given Reference Year will not be the same as in any other given Reference Year, it was Mr Valentin KC’s submission that the non-linearity is not concerned with changes to the real world economy caused by the rebasing but with differences in the way that the real world economy is measured before and after a rebasing.

As Mr Valentin KC pointed out, Dr Borensztein explained in his evidence that “the fact that the ratio between actual GDP in the new base and in the old is non-linear and varies over time and will be different every following year, it’s really not relevant for the purposes of the security”. Dr Borensztein went on to say this:

“the ratio of estimates on the new base and the old base is not a relevant economic variable. It is not related to the revenue, the strength of the fiscal revenues, that the Republic is getting … But that difference between the new base and the old base doesn’t give you any information at all over how the Republic’s fiscal revenues are doing, and this what I want to link to the payment.”

Accordingly, Mr Valentin KC submitted, the relationship or link which matters is between the payments, on the one hand, and the growth in the “real world economy” and the Republic’s ability to pay, on the other. This, he observed, is consistent with Dr Buiter having agreed (in line with what I have myself already agreed) that the purpose of the Securities was to pay out when the real economy is growing at a healthy rate, and so with the need to adjust Base Case GDP in a way that makes it comparable to the best measure of the real economy. On that basis, Mr Valentin KC submitted, the non-linear point is irrelevant as an economic justification for the Annual Adjustment Construction.

The difficulty with Mr Valentin KC’s submission, however, is that it, in effect, focuses upon something which the Adjustment Provision does not itself focus upon, namely what Mr Valentin KC described as the “real world economy”, when the Adjustment Provision is, in its own terms, focused upon a method of measurement of that economy which entails looking at GDP performance relative to a contractual benchmark, namely the Base Case, rather than economic performance in the abstract. As Dr Buiter explained, when giving evidence:

“This can only happen if the reason the growth rate of real GDP is negative and still in excess of the growth rate of base case GDP is that the rebasing, the data revision, the adoption of an alternative measure, has turned a materially positive growth rate at the old year of base prices into a materially negative growth rate at the new year of base prices. So it is not what I would call a recession by any normal economic use of the word, it is a major data revision.”

He added:

“Again, if the reason they have -- you call it ‘recession’ -- negative real GDP growth is just the rebasing rather than, I would say, a proper cyclical contraction, which is a reduction in the utilisation rates of labour and capital across in the economy, which would show up in the old year of base prices for sure as well as the new year of base prices, basically, rebasing exercises are not there to capture recessions that are untracked by the old base.”

Furthermore, thirdly, a proportional adjustment reflects the wording of the Adjustment Provision in ensuring that the effect of the rebasing on both sides of the propositions is properly reflected, whereas the Republic’s approach does not. The Annual Adjustment Construction ensures that in each Reference Year an adjustment is made which reflects the difference between the two measurements (1993 YOBP and New YOBP) in that particular year.

In this way, as Ms Prevezer KC submitted, it takes account of the non-linear relationship of two GDP series in different YOBPs. Only an annual adjustment which is based on the relationship between Real GDP measured in 1993 YOBP and the new YOBP for the Reference Year in question can do this: ensuring that the change in Actual Real GDP from using one system of measurement (YOBP) rather than the other is reflected in Base Case GDP for that year in the same way as it is reflected in Actual Real GDP.

As Ms Prevezer KC went on to explain, this is done by applying the relevant ratio for the particular year to both sides of the propositions (Base Case GDP and Actual Real GDP), which will be different in other years due to the non-linear relationship between GDP in different years of base prices. Adopting a Fixed Fraction approach, in contrast, involves adjusting Base Case for each future Reference Year following the rebasing by a single ratio which is based on the relationship between Real GDP in 1993 YOBP and in the new YOBP in a single historic year, even though there is no dispute that the non-linear relationship means that the ratio between the two GDP series will be different in each following year.

Ms Prevezer KC illustrated the point by explaining that, on the Republic’s approach, whilst in the Overlap Year there is a match in that, for example, if Actual Real GDP is 20% higher in the New YOBP, then, Base Case GDP will also be 20% higher, that would not be the position in later years. This is because, although the left-hand side of the propositions (Actual Real GDP in the new YOBP) is affected by the rebasing to a different extent in each Reference Year depending on the effect of the new measurement in that particular Reference Year, the right-hand side of the propositions is affected by a fixed amount in every subsequent Reference Year (based on the effect of the rebasing in a single Overlap Year). This means that in later years the change in YOBP will have a differing effect on one side of the propositions when compared with the other.

This, in circumstances where it was common ground between the GDP securities experts that, if Actual Real GDP is rebased (so as to mean that it is measured in a different YOBP), then, Base Case GDP needs also to be adjusted since, if that is not done, the benchmark to be met will be in one YOBP whilst Actual Real GDP will be in another. As Dr Borensztein put it, Base Case GDP “should be consistent with data calculated in the new Year of Base Prices” and, as Dr Buiter explained, given the reason for a rebasing is to reflect structural changes in the real economy, those structural changes are bound to impact potential output.

Fourthly, I agree with Ms Prevezer KC when she submitted that it does not make any sense to treat the Adjustment Provision as applying only to a single year: specifically, an adjustment for the single year to both sides of the equation in an equal/proportionate way which would see use of the new YOBP meaning that Actual Real GDP is measured 2% higher than it would have been measured using 1993 YOBP, with a 2% adjustment to Base Case GDP also needing to be repeated. There is, however, nothing in the Adjustment Provision to suggest that there should be a different treatment in relation to later years. Nor, more importantly in the present context, is it obvious why it would make sense for this to be required to happen from an economic perspective.

Fifthly, the One-Off Overlap Construction makes the choice of the year used to generate the fixed fraction highly significant. This is because the single overlap year will inevitably produce a fraction which is not applicable to or representative of other years, the difference producing (or potentially producing) materially different thresholds under the Level Condition, and not only for one year but in respect of every subsequent year until such time as there is another rebasing. Again, it is difficult to see why this would be regarded as a commercially sensible arrangement to have entered into. It is not clear, in particular, why everything should essentially pivot on one year. If a fixed factor were used, it would make much more sense that the risk should be spread in the sense that, as Ms Prevezer KC described it, there is, then, a weighted average of several overlap years.

Furthermore, it should be borne in mind that the use of a fixed fraction impacts not only on the Level Condition but also on the Performance Condition since, whilst it is not in dispute that the numerical percentage figure required to meet the Performance Condition (3.22% in 1993 YOBP) would remain unchanged in the new YOBP, if the new YOBP series has a higher growth trajectory (with higher average growth rates), the Performance Condition becomes easier to meet. Conversely, if it has a lower growth trajectory, with correspondingly lower average growth rates, it becomes harder to meet.

Sixthly, the risks to which I have referred are all the greater for the holders of the Securities in view of the fact that, on the Republic’s approach, they have no say in the choice of the Overlap Year. It is the Republic alone which makes that choice by deciding when rebasing is to take place and when to cease publishing data in 1993 prices or any other YOBP.

This was what Ms Prevezer KC characterised as a moral hazard objection, which she illustrated by reference to what happened in respect of the 2014 rebasing. As Ms Prevezer KC explained, by early 2014 the Republic had published GDP data for the first three quarters of 2013 in 1993 YOBP and had also been working on rebased data in 2004 YOBP. On that basis, assuming the One-Off Overlap Construction to be applicable, it was open to the Republic to look at the draft data for the last quarter and decide to do one of four things: hold back on publishing the rebased data, in which case Reference Year 2013 would simply be judged by 1993 YOBP; publish 1993 YOBP data and 2004 YOBP, in which case 2013 would be the Overlap Year; publish only 2004 YOBP data, in which case 2012 rather than 2013 would be the Overlap Year (which is what actually happened); or continue publishing 1993 YOBP and 2004 YOBP data in parallel for one or more subsequent years in which case the Overlap Year could become 2014, 2015 or later still (determined by when the Republic decided to stop the 1993 series). Which of these choices the Republic made would result in the selection of the Overlap Year. I agree with Ms Prevezer KC when she submitted, in the circumstances, that there is no reason why investors would have wanted to give the Republic that power, even allowing for the fact that (as Mr Valentin KC acknowledged) there would be an implied contractual restriction on the Republic’s power to exercise any discretion in a way which is irrational, arbitrary or capricious (see Braganza v BP Shipping Ltd and another [2015] 1 WLR 1661).

Mr Valentin KC dismissed the moral hazard objection on the basis that it is unreal, as he put it, to suppose that a sovereign state would seek to understate growth in order to avoid payment under the Securities. He relied in this context upon Dr Buiter having accepted that any sovereign country which is going out to the market and wishes to signal that it is a healthy economy will be incentivised to accurately report GDP, rather than understate it. Mr Valentin KC also highlighted the fact that Dr Buiter additionally agreed that “the ability to wait one or two years to implement a rebasing would not be a powerful tool to manipulate statistics as it would not be known in advance whether a future year would be more or less favourable, and in fact the opportunity to pick a particularly favourable changeover year may not present itself at all”.

I agree with Ms Prevezer KC, however, that, whilst a government is unlikely ever to want to suppress actual growth in the sense of taking steps to suppress the underlying performance of the economy and, furthermore, that a government may have incentives not to underreport growth, the fact is that the hazard is not concerned with any such policy on the part of the Republic but with the selection of the Overlap Year.

In addition, I agree also with Ms Prevezer KC when she observed that it is one thing to select the Overlap Year in such a way as to mean that there is underreporting of growth, thereby avoiding having to pay but running the risk that there will be adverse comment among the electorate, but quite another to report accurately and thereby incur a substantial liability. The more so, since, as Ms Prevezer KC submitted, “voters experience the real world economy and not the reported figure”. The disincentive to underreport is, in short, not as strong as might be thought.

There is, lastly, a point concerning the Payment Amount to bear in mind also. It is not in dispute that, prior to any rebasing, there is a proportional relationship between changes in Actual Nominal GDP and changes in the Payment Amount, in the sense that any change in Actual Nominal GDP will be matched by a proportional change in the Payment Amount. Dr Buiter and Dr Borensztein agreed that that proportional relationship is maintained under the Annual Adjustment Construction but not under the One-Off Overlap Construction. As Dr Buiter, in particular, put it, the reason for the proportionality in changes between Actual Nominal GDP and the Payment Amount is that the immediate ability to make payment under the Securities depends on the generation of actual money which generates tax revenue, with any changes in that ability impacting on the Payment Amount. That is why the Payment Amount is linked to Nominal GDP, rather than to Real GDP, and it matters because, as Ms Prevezer KC submitted, the One-Off Overlap Construction can create a scenario in which, even though both the level and growth rate of Actual Real GDP and the level of Actual Nominal GDP all go up in a particular Reference Year, the Payment Amount goes down or vice versa. The Annual Adjustment Construction, in contrast, maintains, post-rebasing, the proportional link between Actual Nominal GDP and the Payment Amount which exists prior to a rebasing.

For all these various reasons, both based on the wording used in the Adjustment Provision and based on the commercial considerations which I have sought to explain, I cannot accept that Mr Valentin KC can have been right when he submitted that the effect of the Claimants’ interpretation of the Adjustment Provision is “commercially nonsensical and economically absurd”. Specifically, I do not agree with him that the One-Off Overlap Construction is the only interpretation which, as he put it, gives any meaning to the Adjustment Provision, gives effect to the contractual language as a whole, is supported by the overall contractual scheme, makes any economic sense and is consistent with the admissible background.

It follows that nor do I agree with Mr Valentin KC that it is necessary for the Court to correct the mistake in the Adjustment Provision by giving effect to the Republic’s One-Off Overlap Construction. It is unnecessary, in the circumstances, to go on to consider whether the second of the two conditions required by the Chartbrook Principle is also satisfied, namely that it is “clear what correction ought to be made to cure the mistake”, as Brightman LJ put it in East v Pantiles. As to this, it was Mr Valentin KC’s submission that there is no difficulty as far as the Republic is concerned since it is sufficient that “the gist … of what needs to be inserted” and “the nature of the correction that’s required” are known; in other words, that a reasonable person would have understood the parties to have mean that for the purposes of the Adjustment Provision there should be a constant factor.

The difficulty with this, however, is that, as demonstrated by the analysis thus far and even assuming that the Annual Adjustment Construction is not accepted, it can hardly be suggested that the One-Off Overlap Construction is the only construction which the parties could have intended should apply and there is also the fact that the Republic has an alternative case, namely the Hubbard Deflator Construction, which does not entail the use of a constant factor at all. It follows that it is impossible to conclude that a reasonable person would have understood the parties to have intended that the Adjustment Provision should be understood, and only understood, in accordance with the One-Off Overlap Construction. As such, the case is similar to Trillium (Prime) Property, in which it will be recalled that Lewison LJ considered that there was “more than one possible solution to the alleged drafting error” and so that it was not “clear what a reasonable person would have understood the parties to have meant”.

This brings me, lastly, to a further matter raised by the Republic by way of objection to the Annual Adjustment Construction. This is that, in the absence of an express requirement in the Securities that the Republic should have to procure that INDEC shall continue to measure and publish in 1993 prices until the expiration date regardless of any change to the base year used to calculate and publish GDP for all other official purposes, the Court should conclude that there is no such requirement. Accordingly, Mr Valentin KC submitted that the Court should conclude that it cannot have been intended that the Adjustment Provision should have the meaning for which the Claimants contend since, had they done, the Securities would have stipulated that there should be continued publication of GDP in outdated 1993 prices.

Mr Valentin KC submitted, furthermore, that there is no basis on which to conclude that there is an implied term to such effect since such an implied term is not necessary to give business efficacy to the contract given that there is, so Mr Valentin KC submitted, an obvious and commercially sensible interpretation of the Securities which does not require it. Ultimately, however, in closing and when asked by the Court, Mr Valentin KC acknowledged that, if the Annual Adjustment Construction is accepted, then, it would be appropriate to imply the term. It is clear to me, indeed, that it would be: the fact that there is no express term dealing with publication is not a reason to conclude that the Annual Adjustment Construction is wrong in view of the conclusion which I have reached thus far.

There remains a related argument. This is Mr Valentin KC’s contention that it was open to the Republic simply to cease publishing GDP in 1993 YOBP. Mr Valentin KC submitted, indeed, that there is “no country in the world that has ever calculated and published GDP concurrently in an outdated and updated GDP base year series after a rebasing”, adding also that, even if it is “theoretically possible as a matter of Argentine law for the President to instruct INDEC to produce a GDP-related report using a historical base year it is wholly unclear whether and how, in practice, such a task could be accomplished”.

There is, however, no substance in these objections, however unusual (perhaps even unique) publication of such data as a parallel series might be. This is because, quite apart from the implied term which Mr Valentin KC accepted would exist if the Court were to agree with the Claimants on their Annual Adjustment Construction, Mr Davies and Professor Hubbard agreed that it is practically possible and there is no prohibition on so doing. Mr Rodríguez also agreed under cross-examination that it would have been possible to continue publishing data in 1993 YOBP. Professor Hubbard, at least when pressed during the course of cross-examination, agreed also that continued publication would be worthwhile if required by the contract, as the following exchange demonstrates:

“MR JUSTICE PICKEN: Assume I decide, which I’m not saying I will, the contract does require 1993 base still to be used. Then I think what is being put to you is, given that that's what the contract requires, it is economically worthwhile for the Republic to do that because, otherwise, investors won't be getting what they're contractually entitled to.

A.

I would agree with that, but I would just use the phrase ‘specific performance’ as opposed to economic utility. If it is your decision that they have to do it, then they have to do it.”

In summary, the continued publication point is a non-issue. Nor, in truth and for reasons which I have sought to explain, are any of the objections which the Republic has raised. On the contrary, not only does the language used in the Adjustment Provision point (if not wholly unambiguously, then, nonetheless, pretty convincingly) towards the correctness of the Annual Adjustment Construction, but there is an economic logic (as explored in some detail above) to that construction being more appropriate than the One-Off Overlap Construction. In any event, it is quite impossible to conclude that the latter is the only rational construction.

I come on, next, to address the Republic’s alternative case, namely the Hubbard Deflator Construction. I propose to do so relatively briefly since it is clear to me that it is a hopeless case.

It was Mr Valentin KC’s submission that, if the Court were to accept the Claimants’ contention that the Adjustment Provision requires a separate fraction to be applied for each Reference Year, then, the variable adjustment fraction proposed by the Claimants (as part of the Annual Adjustment Construction) is, nonetheless, still misconceived because it applies the wrong denominator, namely the Actual Real GDP for the relevant Reference Year measured in constant 1993 prices.

The Republic’s alternative case - albeit a case which was only put forward very late in the day - is encapsulated in the following reworded version of the Adjustment Provision (with underlining for emphasis):

“... provided that, if the Year of Base Prices employed by INDEC for determining Actual Real UDF shall at any time be a calendar year other than the year 1993, then the Base Case GDP for each Reference Year shall be adjusted to reflect any such change in the Year of Base Prices by multiplying the Base Case GDP for such Reference Year (as set forth in chart above) by a fraction, the numerator of which shall be the Actual Real GDP for such Reference Year measured in constant prices of the Year of Base Prices, and the denominator of which shall be the Actual Real GDP for such Reference Year measured in constant 1993 prices [in the new Year of Base Prices as published by INDEC, adjusted for inflation from pesos in the new Year of Base Prices to 1993 pesos using the INDEC-published deflator from 1993 to the new Year of Base Prices.].”

Accordingly, the Republic’s position is that the starting point for the denominator of the fraction should be the Actual Real GDP for the relevant Reference Year (as published by INDEC). This should, then, be adjusted to account for the effects of inflation since 1993 (using the INDEC-published deflator). The resulting figure is the Actual Real GDP of the Republic for the relevant Reference Year “measured in constant 1993 prices”.

Applying this “Inflation-Adjustment Method” for Reference Year 2013, there is no dispute that the Performance Condition would remain at 3.22%.

It was Mr Valentin KC’s submission that the Hubbard Deflator Construction is, as he put it, “faithful to the contractual text” and that it also has a number of advantages, namely that it neither assumes nor requires the continued availability of Actual Real GDP calculated and published by INDEC using 1993 as the base year, it is consistent with the economic logic behind the creation of the Securities (and preserves the pre-set growth percentages found in Base Case GDP Growth) and it can be used to produce GDP data in constant 1993 prices on a going-forward basis (including after any subsequent rebasing). More generally, Mr Valentin KC submitted that the Hubbard Deflator Construction represents a practical solution to problems that are only created if the Claimants’ primary interpretation of the Adjustment Provision is held to be correct.

I do not agree with Mr Valentin KC about this. My main reason for this is straightforward: whatever Mr Valentin KC might say about the Hubbard Deflator Construction being “faithful to the contractual text”, that is just not the case as demonstrated by the revisions to the Adjustment Provision which the Republic’s alternative case requires.

Specifically, the Republic seeks to read words into the Adjustment Provision which change the denominator of the Adjustment Fraction into an inflation adjusted figure. This, in circumstances where the denominator in the Adjustment Fraction is expressly defined in the Adjustment Provision as being “the Actual Real GDP for such Reference Year measured in constant 1993 prices”, and “Actual Real GDP” is defined as being “for any Reference Year, the gross domestic product of Argentina for such Reference Year measured in constant prices for the Year of Base Prices, as published by INDEC”, with the “Year of Base Prices” defined as being 1993. Since it is common ground that a YOBP and a GDP series in that YOBP includes not merely the prices of that Base Year but also the scope/base (the relevant methodology and the configuration/weightings of the economy in that year), it is quite clear that what the Adjustment Provision is there describing is the published INDEC GDP figure in 1993 YOBP produced by INDEC using its 1993 YOBP methodology.

Mr Valentin KC disagreed. He submitted that measuring GDP data “in constant 1993 prices” (i.e. accounting for inflation since 1993) does not require the use of the 1993 base since that would entail the use of outdated and less accurate methodology and weightings. As such, he submitted, there is an important distinction between measuring GDP “in constant 1993 prices” and measuring GDP “in constant 1993 year of base prices” (i.e. measuring GDP data in 1993 prices using the 1993 base). The distinction, in Mr Valentin KC’s submission, makes it necessary to adopt an inflation-adjustment method in order to determine the best estimate of the Republic’s Actual Real GDP “measured in constant 1993 prices”.

Mr Valentin KC submitted, furthermore, that the Hubbard Deflator Construction properly reflects the fact that the “Actual Real GDP” definition refers to “constant prices for the Year of Base Prices”, and that the “Year of Base Prices” definition not merely refers to “the year 1993” but goes on to state “provided that if the calendar year employed by INDEC for purposes of determining Actual Real GDP shall at any time be a calendar year other than the year 1993, then the Year of Base Prices shall mean such other calendar year”. So, Mr Valentin KC submitted, Actual Real GDP means the Republic’s GDP for a given Reference Year measured in constant year of base prices for the base year adopted by INDEC (which, in the absence of a rebasing, would be 1993). As a result, given the rebasing which saw the change to 2004 prices using the 2004 base, it was Mr Valentin KC’s submission that Actual Real GDP means the Republic’s GDP for a given Reference Year as measured in constant 2004 year of base prices (i.e. 2004 prices using the 2004 base). Accordingly, given that the Adjustment Provision requires the measurement of Actual Real GDP in constant 1993 prices (in contrast with 1993 as the year of base prices), Mr Valentin KC submitted that the Inflation-Adjustment Method provides the mechanism by which Actual Real GDP (i.e., actual GDP calculated using the 2004 base) is “measured in constant 1993 prices”.

The difficulty with this, however, is that the Hubbard Deflator Construction requires the words “in constant 1993 prices” to be crossed out and replaced by wording which, although it would have been open to the parties to have adopted, is not wording which they, in fact, chose to use. It is not, in such circumstances, open to the Court to arrive at a construction which would run counter to the language used in the Adjustment Provision.

In any event, as Ms Prevezer KC pointed out, even if “constant 1993 prices” is construed to be something other than a reference to Real GDP in 1993 YOBP, it is not in dispute that the Hubbard Deflator Approach does not, in fact, hold the 1993 prices of goods and services (the price configuration) constant but, rather, replaces the price configuration of goods and services based on their prices in 1993 with the price configuration of goods and services based on their prices in 2004.

Nor does the Hubbard Deflator Construction work in relation to a second or subsequent rebasing, unless INDEC continues to publish GDP data in 1993 YOBP anyway because the Hubbard Deflator Construction expressly requires an INDEC published Implicit Price Deflator between 1993 and the new YOBP. This is only published if Real GDP for both of those years has been measured in a single YOBP and so for the rebasing at issue in this case, there is such a deflator because INDEC did produce Real GDP figures for both 1993 and 2004 in 1993 YOBP. Accordingly, if INDEC were to cease publishing GDP data in 1993 YOBP (as it in fact did and the Republic says it was entitled to do) for 2013 and subsequent years, then for (say) a rebasing in 2023 to (say) 2017 YOBP, there would be no such deflator between 2017 and 1993. Professor Hubbard agreed with this, explaining that there would have to be “add[ed] a step if you are using the methodology that I outlined. You would effectively need one per rebasing”. What that step might be is wholly unclear.

For these reasons, I cannot accept that the Hubbard Deflator Construction is the right construction of the Adjustment Provision. The Annual Adjustment Construction, accordingly, remains the construction which I consider to be appropriate.