SATA Internaçional - Azores Airlines SA v Hi Fly Limited

The claimant, SATA, is a regional air carrier based in the Azores. It is owned by the government of the autonomous region of the Azores. In 2016 it leased an Airbus A300 (MSN 970, “the aircraft”). The lease was to expire on 15 March 2021. In 2018, the aircraft was sold to the first defendant, Hi Fly. The lease was novated, so that Hi Fly became the lessor. By 2019, SATA was in financial difficulty. The aircraft was not in use. SATA remained liable to pay rent and monthly maintenance reserves payments to Hi Fly, but had fallen behind with those payments, on which default interest was accruing. The lease had over a year left to run. When it ended, SATA would have to pay to put the aircraft into the condition that the contract required for redelivery. That was going to require expensive maintenance, especially to the engines.

Over about six months between October 2019 and early April 2020 a deal was hammered out between SATA, Hi Fly, and another aircraft leasing company, AELF, which is the additional party. It culminated in three agreements, all signed on 8 April 2020 and becoming effective that day. By one (the “sale agreement”) Hi Fly sold the aircraft to AELF. By another (the “novation agreement”) AELF took over as lessor. By the third (the “termination agreement”) AELF agreed to the early termination of the lease “as is where is”, for a lump sum payment. (In some cases the agreements involved affiliates, but this makes no relevant difference. I shall refer simply to the parties to this action.)

When the agreements were executed, SATA owed Hi Fly just under USD 3 million in unpaid rent, maintenance reserve payments, and default interest. (In the rest of this judgment I shall simply refer to “rent”. Unless I say otherwise, I mean rent, maintenance reserve payments, and default interest.) Of that, a bit under USD 1 million related to rent which had been outstanding since before 1 September 2019, and there is no dispute or doubt that SATA remained liable to Hi Fly for that. The balance related to rent that had become due after September 2019. The parties dispute the fate of those liabilities. SATA says that under the novation agreement they ceased to be due to Hi Fly and became due to AELF, and were then swept up in the termination agreement. AELF and Hi Fly accept that if they were transferred to AELF then they were settled by the termination agreement. But they maintain that they were not transferred. They say that they remained due to Hi Fly. If so (as everyone accepts) they were not addressed in the termination agreement, and remain due.

I must decide two questions. First, as a matter of interpretation, did the novation agreement transfer the right to recover rent relating to the period from 1 September 2019 to AELF, depriving Hi Fly of its right to recover it? Second, if it did, was that by mistake (either multilateral or unilateral) which would justify the court rectifying the novation agreement?

I have concluded that on its true construction the novation agreement did transfer the right to recover rent accruing after 1 September 2019 to AELF. I have also decided, however, that the novation agreement is to be rectified to preserve Hi Fly’s entitlement to those sums, because I find on a balance of probabilities that all the parties to the novation agreement positively believed that it would do so and shared that intention in the way required for rectification. My reasons for reaching those conclusions are set out below.

I heard evidence from five witnesses.

SATA called Mrs Ana Azevedo (a board member during 2019), Mr Chaves (its COO from early 2020), and Ms Teresa Gonçalves (its CFO from early 2020, later its CEO). Hi Fly and AELF, who presented their case together and made common cause, called Mr Sérgio Bagorro (Hi Fly’s CFO) and Mr Brian Hollnagel (a consultant to AELF, which is owned by his wife).

Hi Fly and AELF also served a witness statement from Mr Dan Roberts, who was a consultant who assisted Hi Fly with the documents. During the trial, SATA elected not to cross examine him. His written evidence was therefore admitted as unchallenged.

I did not get the impression that any witness was actively seeking to mislead me. But they were all focusing on events which took place 4–5 years ago, and often attempting to reconstruct their impressions at the time in hindsight and in the light of the issues that are now salient. Mr Chaves’ reconstruction depended on only a small selection from the relevant documents. That is not a criticism of those who prepared his witness statement, who no doubt wished to avoid polluting his recollection with too much documentary input. But the fact that he was not only joining the dots—as almost always happens in cases like this—but joining only some of them made the resulting picture somewhat distorted.

Mr Duffy KC, who appeared for Hi Fly and AELF, reminded me of judicial warnings that human memory is complex and not straightforwardly reliable, and that it can easily be distorted by wishful-thinking and the litigation process. He invoked Leggatt J’s now-famous observations in Gestmin SGPS SA v Credit Suisse (UK) Ltd [2013] EWHC 3560 (Comm), [2020] 1 CLC 428, and the many similar comments by judges in the commercial court, including in rectification cases such as Global Display Solutions Ltd v NCR Financial Solutions Group Ltd [2021] EWHC 1119 (Comm) at [403]–[404] (Jacobs J). Contemporaneous documents and inferences from them are critical for rigorous testing of oral evidence. In reconstructing events, he suggested, I would be helped more by the documents than by anything the witnesses could remember.

Contemporaneous documents are, of course, important. Some of the witnesses made errors of recollection which reference to those documents demonstrated, as one would expect. There is also, as Mr Duffy submitted, a risk that witnesses will confuse what they wish they had thought or done, or believe they should have thought or done, with actual reality. Particularly when being asked to recall intentions or beliefs, it is often difficult to distinguish between these things, and to avoid rationalising past actions and beliefs to harmonise with present ones. That process need not involve conscious embellishment or falsification, and often will not do so.

Documents, however, also have their complexities and dangers when they are deployed to establish subjective questions about what people thought, understood, or intended at the time. The litigation process can, here too, be distorting. By the time the trial begins, relevant documents (perhaps written months or weeks apart) have been brought together, and the irrelevant removed. A person who might receive tens or hundreds of emails every day is now being asked to look at just a few, which can be put alongside each other in moments. The witnesses, advocates, and the court focus sharply on well-defined issues that have become prominent, but may not have been at the time. We can ignore many other events and concerns which occupied people’s attention. In the slow quiet of the court we have the luxury of time to subject documents, which may have been written and read quickly and under pressure, to minute and leisurely exegesis.

All this may distort reality quite as much as uncritical reliance on oral evidence. In the real world people often read and write quickly. They forget things they read or wrote even a few days earlier. They express themselves imprecisely (sometimes by accident, sometimes by design). They do not explore issues that turn out to matter later; make unsupported assumptions; do not make connections that in hindsight seem obvious. They tolerate or overlook inconsistencies. They pursue plans which are not rational or wise with hindsight. The proper use of contemporaneous documents must be alive to these factors, especially when the aim is to unearth subjective beliefs.

The circumstances here illustrate some of these problems. SATA was in financial crisis. There were massive changes in critical personnel, with a wholesale replacement of the board in December 2019, and there must surely have been a great deal of drama and confusion associated with them. By the later stages of the story, the pandemic was in progress, and SATA was trying to deal with all that it involved for its business. Although resolution of what to do about the aircraft was important, it was evidently far from being the only issue, and the point I am now concerned with was far from its most important aspect. The negotiations took place over several months. For some of them I am working with translated versions of documents originally written in Portuguese (with the loss of nuance that involves). For others, the parties were themselves negotiating in different languages and some were not native speakers, with the same effect. Nobody at the time had a comprehensive “bird’s eye” view of the whole situation: SATA had only a very partial view of the discussions between AELF and Hi Fly, AELF had a very partial view of the discussions between Hi Fly and SATA, Hi Fly had a very partial view of the discussions between SATA and AELF. Many of the documents concerned negotiations in conditions of considerable uncertainty. It is a situation in which forensic reconstruction risks producing fantasy, not history. The documents, no less than the oral evidence, demand to be considered with that in mind.

Paragraphs 16 to 106 set out my findings of primary fact. I have necessarily sometimes anticipated findings that I make about subjective beliefs at a particular point in time. It will help the reader to understand that the trial’s main focus was on whether, by the time the novation agreement was concluded, Mr Chaves positively believed that it would leave Hi Fly entitled to recover all outstanding arrears or knew that Hi Fly believed that. The alternatives are either that he positively believed that it would transfer all rent arrears to AELF (as he now says, and as SATA primarily contends), or that he was confused or doubtful and had no positive view on the matter (which is a logical alternative to either of the parties’ primary case). At this point in my judgment, I examine Mr Chaves’ apparent beliefs at close range, event by event. Later on, at paragraphs 152–177, I examine them from a greater height and as a whole to explain my ultimate factual conclusions on the key issues.

The aircraft was bought by Hi Fly, and the lease novated, in 2018. The lease was due to expire in March 2021. Hi Fly had a long and varied commercial relationship with SATA. It treated SATA with commercial flexibility that a more hard-nosed business partner might not have shown. For instance, although SATA was in default for long periods, Hi Fly did not seek to default the lease, and SATA never seemed to consider that a real risk.

We can pick up the story in around mid-2019. By then, SATA was in financial difficulties. The aircraft was not in use. Arrears of rent and maintenance reserve payments had built up. For some time SATA and Hi Fly had sporadically discussed bringing the lease to an early end. By September 2019, Hi Fly’s President, Mr Paulo Mirpuri, had made proposals for early termination. They involved SATA clearing the arrears that it already had to Hi Fly, then said to amount to around USD 2.8m, putting the aircraft in redelivery condition, and paying half the rent and maintenance reserve payments that would become due for the remainder of the term as a lump sum.

Mr Mirpuri made it clear that progress depended on the arrears being cleared. SATA understood that.

On 8 October 2019, Mrs Azevedo (then SATA’s CFO) prepared a memorandum for the board, setting out what she understood to be the financial implications of various options. Her starting point was that if SATA retained and stored the aircraft until the lease expired, and then paid what would be needed to put it into the required state for redelivery, the cost would be USD 7.3 million (in rent accruing until the lease expired) plus USD 17 million (to put the aircraft in redelivery condition). Mrs Azevedo at this point assumed that accumulated maintenance reserves of USD 11 million would be available to defray this cost. On that basis, the cost of keeping the aircraft was just under USD 13 million (taking into account the security deposit that would be returned at the end of the lease, though Mrs Azevedo for some reason used the wrong number). In contrast, Hi Fly was demanding that the rent be brought up to date, that 50 percent of the future rents and maintenance reserves be paid, that the security deposit be forfeited, and that SATA pay for the work required on the aircraft without recourse to the maintenance reserves. Clearly, on those figures, Hi Fly’s proposal was not attractive.

Meanwhile, AELF and Hi Fly had begun to discuss AELF buying the aircraft and taking over the lease. On 5 October 2019, the relevant parties executed a proposal letter containing an agreement in principle to this effect.

This sale was based on an “Economic Closing Date” of 1 September 2019, at a price of USD 25 million. What this meant was that although ownership would not be transferred until some unknown future time, it would then be transferred “as if” it had taken place on 1 September 2019. This would be done by reducing the purchase price by any rent or maintenance reserves that Hi Fly received between 1 September 2019 and the transfer date. The agreement said that “[b]ecause Transfer shall occur after the Economic Closing Date … the purchase price of the aircraft shall be reduced by an amount equal to the monthly rent and maintenance payments actually received by the Seller during such period”.

Mr Bagorro, when cross-examined, wished to equate “actually received” with “actually due”, suggesting that from an accountant’s perspective there is no difference. I did not find this evidence convincing. An accountant well understands that there is a difference between payments which ought to be made, and payments which have been made. In any case, Mr Bagorro was not the main negotiator of this document, and I did not hear evidence from Mr Mirpuri, who was. I find that Hi Fly’s expectation at this point was that it would have to account for any sums paid by SATA between 1 September 2019 and the aircraft’s transfer, but not for sums falling due for that period which were not received. At the time, however, that would not have been a major point, because Hi Fly were still making it a condition of novation that SATA would clear the arrears. If that happened then the sums falling due for that period would have been “actually received”.

Mr Hollnagel, who conducted the negotiations for AELF, did not conflate “actually received” with “actually due”. He accepted that there was a difference. He said that he did not think that the difference would matter in practice because he expected rent to be current at the date of the transfer, and did not learn until later in October or November that SATA was in significant arrears. I accept that evidence. As I set out below, the initial drafts of a novation agreement were prepared on the basis that there would be no rent arrears at the transfer date, and it was not until 31 October 2019—after analysing a document he obtained from SATA a few days earlier—that Mr Hollnagel first raised the arrears issue with Hi Fly. Mr Hollnagel’s recollection is also confirmed by a lawyer’s email of 25 February 2020 which said that “at the time the LOI was signed, AELF had no idea that the lessee was in default (or by how much)”.

SATA was not involved in the agreement about the sale, and therefore had no way of knowing what had been agreed between AELF and Hi Fly (or indeed that anything had been agreed).

Later in October,SATA was informed of the possible transfer of the aircraft to AELF. On 23 October 2019 Mrs Azevedo and SATA’s then-CEO Mr Antonio Texeira met Mr Hollnagel in Ponta Delgada. It is common ground that at that meeting SATA allowed Mr Hollnagel to see and take a photograph of Mrs Azevedo’s memorandum. The memorandum was written in Portuguese, which Mr Hollnagel did not speak, but it is clear from notes he made later that he subsequently subjected it to reasonably detailed thought, presumably with the aid of a translation.

The purpose of the meeting was to discuss the possibility of early termination. It seems to have been largely an exercise in fact-finding by Mr Hollnagel. Mrs Azevedo says they did not discuss any economic closing date or what would happen to rent falling due while the termination was being negotiated. That was not challenged. Mr Hollnagel says that he made it clear that he could not discuss the existing lease arrangements with Hi Fly. I am sure that was clear to SATA. The purpose of the meeting was to set the ball rolling on possible negotiations for termination after AELF acquired the aircraft. The topic of how arrears accruing after 1 September 2019 would have been addressed did not arise—not least because Mr Hollnagel was not aware of them before the meeting.

First drafts of the novation agreement, and of the sale agreement, were exchanged between AELF and Hi Fly on 31 October 2019. They were consistent with the scheme in the letter of 5 October 2019:

On 31 October, Mr Hollnagel wrote to SATA. The thrust of his letter was to persuade SATA that the figures Mrs Azevedo had prepared were over-optimistic. It would cost more than SATA thought to put the engines into redelivery condition (about an extra USD 1 million per engine). Also, AELF did not agree that SATA would be able to use the maintenance reserves for this purpose. Whether or not this was the first time SATA was alerted to this issue, it is common ground that after taking legal advice SATA accepted that this was indeed a risk.

Mr Hollnagel proposed a termination payment of USD 14.98 million odd, assuming simultaneous novation and termination on about 16 December 2019. His letter said nothing at all about any difference between any “economic” closing date and actual closing, and nothing about any rent arrears at 16 December 2019. Nor was the proposal precisely justified by reference to SATA’s liability for future rent. It was presented simply as the “cost we believe new lessor will incur to induct the Engines into maintenance; positioning of the Aircraft for maintenance; transport of engines, etc; and other expenses, not outlined in your analysis, and is net of a credit to the Lessee in an amount equal to the Security Deposit”.

On the same day that Mr Hollnagel sent his proposal to SATA, he sent an email to Mr Mirpuri in which he noted that “a document that [SATA] has provided referenced that they were $3M in arrears, are they currently in default under the lease?” Mr Mirpuri replied the next day, somewhat evasively, to the effect that SATA had “always paid their bills” but are “government owned and sometimes there are delays waiting for [g]overnment funds”, and were a “bit relaxed on their ontime payments”. Mr Hollnagel was able to read between the lines of this ostensibly reassuring assessment to see that Mr Mirpuri accepted that there were arrears.

On 13 November 2019, a draft novation agreement was sent to SATA. Clause 7.3 was in substantially the form that it ended up taking in the executed agreement (see paragraph 113), except that some words remained to be added to the end of clause 7.3(a), which happened later as I describe in paragraph 70. Although Mrs Azevedo was copied, there is no evidence that she focused on the document at that point, and it is common ground that SATA did not engage in any meaningful way with the proposed terms through the rest of 2019.

SATA’s internal thinking satisfied Mrs Azevedo that Mr Hollnagel’s offer (which, since it involved AELF retaining the security deposit, had a cost of just over USD 16 million) was likely to be better than allowing the lease to run its term. But SATA did not have cash available to pay at once, and would therefore need to negotiate for instalment payments. This is what it planned to do. On 20 November 2019, Mrs Azevedo warned Mr Hollnagel that payment would depend on a capital injection from SATA’s shareholder, which would not happen until January 2020.

Mr Akoni, who appeared for SATA, argued that this analysis showed that Mrs Azevedo was assuming that nothing would be paid to Hi Fly for any rent accruing due after October. That would give the document more weight than it will bear. Mrs Azevedo started from a rough estimate of the amount due to Hi Fly of USD 3 million, which was the figure used in October and not updated or changed, and of the rent that would fall due for the rest of the lease period. She did not seek to update her figures to take account of additional invoices since October or any payments since then. But the small differences this might have made would have been practically speaking irrelevant. There was nothing to suggest, at this stage, that Hi Fly would not insist on being paid, in full, up to the novation date. That remained its position until late in January 2020. So if Mrs Azevedo did not account for this cost, it cannot have been because she thought it was one that could be avoided.

A further meeting took place on 22 November 2019 in Ponta Delgada. SATA prepared a note of that meeting. Mr Hollnagel attended for AELF. Various senior managers, including Mrs Azevedo, attended for SATA. The meeting seems to have had two main topics, at least so far as is relevant here. First, SATA sought to persuade Mr Hollnagel that he was over-estimating the costs of putting the aircraft into a returnable condition. Secondly, it sought to explore whether payments could be spread during 2020.

Among the matters the note records (as an argument “put forward by the lessor”, i.e., AELF as prospective lessor) was that “[AELF] will receive a plane with a debt of USD 3M; under these conditions its management will be forced to file a lawsuit against SATA, implying a delivery of the aircraft with all contractual costs in the order of USD 23.5M.” Mrs Azevedo’s evidence was that she understood Mr Hollnagel to be threatening that if SATA would not pay what he asked for termination, AELF could take over the lease with the outstanding debt, and then declare a default, with the immediate requirement for redelivery and compensation for lost rental that would entail. Mr Hollnagel’s manuscript notes, which he said were prepared in advance of the meeting to equip him for it, suggest that he did indeed intend to contrast SATA’s position if it accepted his offer (and paid USD 15 million) with its position if it rejected it, and AELF defaulted the lease. The point he presumably would have wanted to get across was that if that happened SATA would not have the luxury of 18 months to find the money. In cross-examination, he said that the default he had in mind was not the existing arrears, but a further default in payment of rent after transfer to AELF, which seemed inevitable if SATA was as cash-strapped as it seemed. I accept that is what he now believes, but I am sure that Mrs Azevedo’s note more accurately captures what he said, or at least what she understood him to be saying. The reference to AELF “receiving a plane with a debt of USD 3M” (emphasis added) can mean nothing else. There is other contemporaneous evidence, which I refer to in paragraph 37, that shows that this is how SATA understood matters.

The position, then, had begun to become more complicated than it had seemed before. Mr Hollnagel now knew—contrary to his original expectation—that there were existing arrears. Hi Fly’s position was that the outstanding debt, including sums due for periods from October 2019 which had recently been invoiced, needed to be settled urgently, and that this was a condition of transfer. Mr Hollnagel appeared to be contemplating that if agreement could not be reached, AELF might take over the plane with USD 3 million of arrears, and use that as a means of terminating the lease early whether SATA wanted to or not, with an immediate acceleration of SATA’s liability. Hi Fly’s and AELF’s positions were not, therefore, consistent. But that inconsistency was not explored.

Hi Fly continued to send invoices for rent and maintenance reserve payments, and to press for them to be settled. By 24 November 2019, SATA had realised that paying Hi Fly’s debt could draw the sting from Mr Hollnagel’s threat to take over the plane, with the debt, and default the lease. One internal email pointed out that “if we pay the USD$3M debt, we are up to date, there is no longer any default”, so that SATA could avoid the fate that Mr Hollnagel threatened if it paid Hi Fly up to date and then started paying AELF promptly.

Perhaps spurred on by that thought, SATA met Hi Fly on 27 November. Mrs Azevedo took a note. The discussions concerned the “need to settle SATA’s debt by December 15, 2019, before the plane’s passage to the new Lessor”. It recorded the debt at 27 November as being just under USD 2.13 million (i.e., for reasons that were not explored in the evidence, somewhat lower than the figure given on 8 October), and contemplated some financial operation which would enable the settlement of the debt by that stage.

The documents do not show whether the “need” for this to happen reflected Hi Fly’s insistence on settlement as a precondition of transfer, or SATA’s concern that the aircraft needed to pass to AELF “default free” as a prophylactic against Mr Hollnagel’s threat to terminate for default, or both. But either way, if the debt was going to be fully paid before the plane was transferred to AELF, the question whether SATA would be liable to AELF or to Hi Fly for any part of it after termination would not arise. The note does not suggest that any distinction was being drawn between arrears relating to periods before September or October 2019 and later periods. In so far as Mrs Azevedo suggested that the arrears being discussed were simply the USD 3 million outstanding up to October, I think she is mistaken. I accept that SATA’s internal accounting systems were in some disarray so that it was hard to get accurate figures; but the figure being discussed was clearly intended to be current “on November 27, 2019”. The overall tenor of the discussion was plainly that (subject perhaps to minor uncertainties) SATA would be fully up to date by 15 December 2019. That is in effect what Mrs Azevedo told the regional government on 29 November 2019 (“a plan to pay the debt”, not part of it). It is also what she told Mr Hollnagel on 29 November (“we have a plan to pay the debt until 15 December 2019”). And it is how he understood what he was being told (“they are focused on getting current before novation”, as he put it on 2 December 2019). It is also how Mrs Azevedo explained things to Mr Hollnagel on 2 December (“We are doing all efforts in order you receive the aircraft with no debt behind”).

On 2 December 2019, Mrs Azevedo wrote to Mr Hollnagel telling him that the shareholder had “no final answer” on lease termination, and suggesting that it would be necessary to move forward with the novation agreement, which she expected would be signed on 16 December.

On the same day, Mr Roberts sent an email to Mr Mirpuri regarding the documents. He noted that SATA were “significantly in arrears” on lease payments and presumed that “our rights to collect this USD 3m should be preserved and handled as part of any ‘early return’ deal with SATA”.

On 12 December 2019, Mrs Azevedo made a proposal to Mr Hollnagel. SATA would pay the termination sum of just under USD 15 million that AELF had asked for, but wanted to pay it in three instalments between January 2020 and April 2020, with the aircraft to be delivered by 15 January 2020, “as is where is”, but in an airworthy condition. AELF accepted that proposal the next day.

On 14 December 2019, Ms Azevedo informed Hi Fly that SATA had been unable to raise the money required to pay the outstanding arrears. Hi Fly responded by pointing out that it would be crucial for the debt to be paid that week. After further discussions, Hi Fly agreed to tolerate continued delay provided €500,000 was paid by 28 December. SATA failed to make a payment by that date, but did pay around USD 500,000 on 3 January 2020. There was not yet any indication that Hi Fly’s indulgence would extend to allowing anything to remain outstanding at novation.

The members of SATA’s board, including Mrs Azevedo, stood down in December 2019, and were replaced from 6 January 2020. From that date, Mr Chaves was COO, and Ms Gonçalves was CFO. Both were new to SATA. Of the two, Mr Chaves had relevant aviation industry experience, but Ms Gonçalves did not. Their evidence attests to a situation of crisis and uncertainty. SATA was heavily in debt. There were serious cash flow problems. Ms Gonçalves made it clear in her evidence, and I accept, that she was managing the resources of a company that was on the brink of insolvency, if not actually insolvent, and trying to find ways to prioritise paying employees and critical ongoing suppliers.

One of the topics for the new management team was to decide whether to proceed with the agreement in principle that had been reached with AELF. Mrs Azevedo had provided Mr Chaves with a file of relevant documents (mostly her own notes). Mr Chaves concluded that the deal made financial sense, but SATA did not have the cash, or shareholder support, to fulfil it. In those circumstances, SATA needed to renegotiate the termination payment to allow for payment in instalments over a longer period.

In reaching his conclusion, Mr Chaves relied on various spreadsheets. They are not entirely self-explanatory, and were not fully explored in the evidence. One of the calculations they make is of the costs that SATA would incur if there was an “immediate phase out” (i.e. termination) of the lease. That involved starting with the termination cost (then assumed to be around USD 15 million) and adding USD 3 million debt to Hi Fly, giving a total cost of USD 18 million. Those costs were then compared to the costs that would be incurred in other scenarios.

On 29 January 2020, Ms Gonçalves sent an email to Mr Chaves, which reported a discussion she had had with Mr Bagorro. In pertinent part her email (as translated) read:

“We are really hard pressed [sc. “under pressure”] to decide the theme of [the aircraft].

[Hi Fly] called the financial area to say that it really has to sign a novation contract with the other company next week as it runs the risk of not signing and everything going down the drain because the company that will sign the contract may be without funding.

Initially, the signing of the contract also required [SATA] to pay, by Feb 15, the amount owed to [Hi Fly] (approximately Eur 1.9M). Currently they say that later the payment of the outstanding amount will be discussed but that they need to have the novation contract signed by us tomorrow. …

We need to know what to do. But signing the contract I don’t think it would be an option if we wanted to negotiate, would it?”

This marked a significant change. Up to this point, Hi Fly’s position had been that the outstanding debt to it must be cleared, so that rent was up to date, before novation. Now, to save the sale, Hi Fly was prepared to postpone “discussion” of the outstanding amounts until later. There is no indication that Hi Fly (or Ms Gonçalves) distinguished between arrears that had accrued up to October (or any other date) and those that accrued thereafter. Although the figure given may have come from Hi Fly, and although it was not expressly stated to relate to the aircraft (there were other sums owed to Hi Fly on some wet leases of other aircraft), the context in which it was mentioned strongly suggested that it at least included the debt due in respect of the aircraft.

Mr Bagorro followed this conversation up with an email to Mr Plácido (SATA’s financial manager) the same day. He pressed SATA to sign a novation agreement. He said that this needed to be signed urgently in order not to “jeopardize the whole process”, and that “later, we can calmly implement a payment plan for the current debt”. That evening Ms Gonçalves forwarded the email to Mr Chaves and Mr Rodrigues, SATA’s CEO, for information, saying “We really have to make a decision regarding the A330”. In the subsequent days it appears that SATA had this in mind. For, in a spreadsheet dated 2 February 2020, alongside the entry for the debt to Hi Fly of USD 2.2 million, someone has typed “Defining future payment plan”.

Late on the evening of 29 January, Mr Chaves responded to Ms Gonçalves’ information about Hi Fly’s concerns. He said that, having looked at the analyses that had been done, SATA would have to choose between “accepting the agreement made by the previous administration” or attempting to renegotiate. He noted a risk that if they renegotiated “we will have to pay 23Mi USD instead of the 15Mi negotiated”.

Unbeknownst to SATA, it was around this time (at the end of January) that the draft sale agreement between Hi Fly and AELF was amended so that the purchase price would be reduced not just by rent actually received between economic closing and transfer, but by the sums “due and payable under the Lease” during that period. As Mr Roberts noted in an email he sent to AELF’s lawyer on 31 January, “Hi Fly is now responsible for collecting unpaid [Maintenance Reserves] and Rents from SATA since September 2019”. I cannot be quite sure when this was finally agreed by Mr Mirpuri, but he had certainly done so by 4 March 2020, when he sent an email to Mr Roberts confirming that Hi Fly would agree “to deduct [the amount due from SATA] from the purchase price and recover direct from [SATA] upon closing or later on”.

On 7 February a meeting took place at Hi Fly’s offices. The meeting had two distinct parts. The first part included Hi Fly, SATA, and AELF. The second part was simply between SATA, represented by Mr Chaves and Ms Gonçalves, and AELF, represented by Mr Hollnagel. It is an important meeting, because Mr Chaves singles it out as the moment at which his view of the way the debt to Hi Fly was going to be handled changed, and the reason for that change of view.

It is common ground that the first part of the meeting involved purely technical discussions, and it is therefore only the second part of the meeting that concerns me.

Mr Chaves’ evidence was that Mr Hollnagel had said that he had “authority” to discuss all financial issues, including amounts owed to Hi Fly. I do not accept, in any literal sense, that that is correct. I do not think that Mr Hollnagel said, or Mr Chaves understood him, to have any authority to compromise Hi Fly’s own debts as such. But I do not think Mr Chaves meant to say that. What I think he says he understood was that Mr Hollnagel was able to discuss all aspects of the financial position once AELF took over the aircraft, including arrears if they were transferred. That is not controversial.

It is common ground that the main topic of this meeting was whether AELF would accept a termination payment in instalments over a longer period of time, and if so how much it should be. It is also common ground that in discussing this issue the parties compared SATA’s economic position on the alternative possibilities that (a) AELF agreed to terminate the lease and accept the aircraft “as is where is” upon novation or (b) the lease was left to run to expiry.

Mr Hollnagel’s witness statement described his thought processes as follows:

“My view was that if AELF was to finance SATA’s buyout proposal, it would have to be on terms similar to the Lease, because if the parties failed to reach an agreement to terminate early, AELF would still have all of its rights to accelerate payment under the novated Lease in the event that SATA defaulted. … Therefore my response to SATA’s request was that if it wanted to restructure with a longer payment period than AELF would have had under the remaining term of the Lease, then I wanted AELF to be compensated for not accelerating the former obligations and accommodating a restructuring of our previous agreement.”

Mr Hollnagel’s manuscript notes (possibly made in advance of the discussion as a plan for how he would proceed) confirm that there was discussion of the “default” scenario, and the effect it would have. They also suggest that his assessment of the amount that would then have to be paid was USD 23 million. Whether that figure was one that Mr Hollnagel had independently arrived at, or whether it was something that SATA volunteered, is not clear to me: it was in fact roughly equivalent to SATA’s own estimate of the total cost of holding the aircraft to lease expiry (and had been mentioned as such by Mr Chaves in the email I referred to at paragraph 50).

Mr Hollnagel may not have thought the matter through very clearly: his point was to brandish a persuasive stick, not to engage in jurisprudential discussion. But whatever he intended, it is plausible that Mr Chaves understood threats of “default” to concern, or include, existing arrears. That is how they appeared from Mrs Azevedo’s note of the November meeting. I have little doubt therefore that as Mr Chaves understood it, Mr Hollnagel’s suggestion during this meeting carried with it the implication that he would or might “make use” of the outstanding liability, then owed to Hi Fly, if the novation took place without SATA agreeing to termination. That being so, it does not seem implausible that Mr Chaves might have thought that Mr Hollnagel would be in a position, in practice, to take decisions about what would happen to that debt after novation. If AELF expected to be able to enforce the debt after novation then it would also be able to compromise it.

This was, Mr Chaves said, a change of position compared to what he had expected. As he put it in cross-examination “before we entered the room there was one reality, and then when we started discussing in the room, the reality changed”. Nobody suggests, however, that this was expressly spelled out, or that there was any express discussion of the topic.

Mr Chaves’ impression would not be inconsistent with Mr Hollnagel’s evidence that he was only discussing the period of time during which AELF owned the aircraft. What that meant was open to interpretation. Certainly, all that was being discussed was the termination payment. But whether that termination payment would sweep in outstanding rent would depend on whether that indebtedness stayed with Hi Fly, or transferred to AELF. By discussing “default” in terms that at least suggested that AELF expected that it would or could acquire the plane with debt it could enforce, Mr Hollnagel at least raised the possibility that the termination payment could compromise these liabilities too—not “as liabilities owed to Hi Fly” but “as liabilities now owed to AELF as a result of novation”. All that, however, was a matter of inference and speculation, rather than express discussion.

The upshot of the meeting was that AELF agreed to accept a total payment of USD 23 million in various instalments over two years in return for termination and acceptance of the aircraft “as is where is”. There is no evidence that the arrears featured specifically in the way Mr Hollnagel presented this offer. Ms Gonçalves, on the contrary, says that it was presented simply as a figure which represented what AELF demanded for agreeing to accept payment from a counterparty with obviously high insolvency risk spread over two years. Mr Hollnagel’s witness statement includes a cryptic reference to his having stated that AELF expected “to be compensated for not accelerating the former obligations and accommodating a restructuring of our previous agreement”. Although the first part of that statement may well refer to the default threat (and is, to that extent, consistent with Mr Chaves’ evidence), Mr Hollnagel is clear that the actual presentation of the offer did not involve any sort of break down or any discussion of the arrears. Nor did it form part of his own thinking. That was largely directed at the financial risk of financing SATA over 24 months, and how far he would be able to push SATA.

Mr Akoni submitted that I should conclude that, at least implicitly, the discussion of the sum to be paid reflected the arrears because SATA was naturally comparing the figure it would have to pay if the lease was terminated with the figure it would pay if the lease ran to term. As I understood his submission, it was that if one assumed that acceptance of AELF’s offer would mean that SATA had to find USD 23 million (to pay AELF), plus USD 3 million (the amount Mr Chaves assumed would be needed to pay Hi Fly), the effective price of the offer was so far above what it was estimated (by Mr Hollnagel, though SATA’s figures were similar) to pay over the course of the lease and on redelivery, that the offer would have made no commercial sense. As Mr Chaves put it, the comparison he was making was to an “all in cost” of termination totalling USD 18 million (which was the assumed effect of Mrs Azevedo’s agreement), an “all in cost” now being proposed of USD 23 million, and an “all in cost” of running the lease to expiry of USD 23 million. USD 23 million was therefore, for Mr Chaves, SATA’s “walk away point” to deal with the lease in its entirety, including the arrears: “We could not go over 23”.

Mr Chaves stoutly adhered to this point during cross-examination. I accept that it would not have been practical to pay AELF more than USD 23 million, which was SATA’s internal assessment of the costs of allowing the lease to run to term. I do not, however, accept that the sort of arithmetic that Mr Akoni invited me to engage in reflects how the discussion proceeded. Just a few days before the meeting—and at a time when he believed that come what may Hi Fly’s debt would have to be paid—Mr Chaves has compared the figure of USD 23 million not to USD 18 million, but to USD 15 million. If that was an “apples to apples” comparison, one would assume that neither figure included the Hi Fly debt. Precisely where the USD 23 million figure came from is, in any event, unclear to me. If one started from Mrs Azevedo’s 8 October memorandum, and adjusted for (a) the fact that the lease now had 15 (not 18) months to run but (b) the costs of putting the aircraft into redelivery condition were now estimated at USD 19 million and maintenance reserves were unlikely to be available to meet them, the cost of allowing the lease to run to term would have been USD 24 million odd, without the Hi Fly debt. So it is not likely that the USD 23 million figure that Mr Chaves had in mind was a figure including the Hi Fly arrears.

I do not, therefore, think that it is possible to read into the fact that Mr Chaves’ walk-away point was USD 23 million that he must have thought he was also settling the Hi Fly debt (nor that he did not). The thinking was simply not that refined or precise, and the figure proposed was simply the price demanded by Mr Hollnagel—equipped as he was with all too much information about SATA’s internal thinking—for an agreement that SATA was not in a strong position to quibble over.

It would be unwise for me to reach a definitive conclusion about Mr Chaves’ state of mind based simply on the evidence surrounding one event. My provisional view, however, looking simply at the accounts of this meeting (and the notes of it) is that it is not impossible that Mr Chaves might have gained the impression that AELF was in a position, if it chose to be, to acquire title to the overdue debt (of unspecified amount) which could be used to trigger an immediate event of default on transfer, and that if so he might have assumed that all or some of that debt could be part of the agreement he was making. All of that, however, would have been a matter of inference for him, not of direct or explicit agreement. Whether he actually thought that the arrears would be wrapped up in the termination agreement cannot be decided simply from what happened at the meeting. Significantly, neither Mr Hollnagel nor Ms Gonçalves seems to have drawn the conclusion that Mr Chaves said he did.

AELF’s offer was confirmed in an email from Mr Hollnagel dated 11 February 2020, and accepted by SATA in an email from Ms Gonçalves dated 13 February 2020. Neither email discussed the arrears. Work then began between the lawyers to finalise the termination agreement and ultimately other documents. None of the principal decision-makers was involved in that detail.

Various documents record discussions between Mr Roberts and Mr Mirpuri which show that their internal thinking was that since the purchase price paid to Hi Fly would be reduced by rent arrears outstanding from SATA, it would be for Hi Fly to recover the debt from SATA. Mr Mirpuri’s view was that priority should be given to dealing with the novation and sale, and deal with recovery of the outstanding amounts later. Since that is not in dispute, I need not describe the detail.

On 28 February 2020, Mr Bagorro sent an email to Ms Gonçalves and Mr Chaves. He complained about the delay in finalizing the novation agreement, and said “In practice, [SATA] is not paying Hi Fly for the lease of the plane, but the Novation process is not helping either, as we are being doubly penalized”. Ms Gonçalves’ evidence was that she read this email essentially as a complaint about delay in that process. I accept that, which is consistent with the main express purpose of the email (“to talk to you about the Novation”), with what she did shortly after receiving the email, which was to chase a colleague for a status report on the novation, and report that back to Mr Bagorro. Mr Chaves responded in the same way, apologising for the delays in the novation and promising to chase the engineering department to close the outstanding issues.

On 12 March 2020, Mr Roberts sent an email to AELF’s lawyer, Mr Fry, asking for him to confirm with SATA that some wording he proposed could be added to clause 7.3(a), adding the words “and such amounts will remain payable by the Lessee irrespective of any subsequent amendment or waiver occurring in relation to the Lease after the Effective Time”. This clause in its unamended form had been in the draft sent to SATA in November. Mr Roberts’ expressed purpose was to “clarify that SATA still owes Hi Fly the arrears of rent and maintenance reserve irrespective of the upcoming lease termination you are working on”. As is common ground, Mr Roberts was making a mistake here, because (as the novation was drafted) the amendment related only to arrears in respect of rent due before 1 September 2019, and not to all arrears at closing. More extensive surgery would have been required to put the result Mr Roberts intended beyond argument.

SATA did not accept that change. The next day, Mr Roberts sent an email to Ms Viveiros, a lawyer at SATA, copying Mr Chaves and Ms Gonçalves, among many others. He set out clause 7.3(a) of the Novation Agreement, including his additional words, and referred to the omission of his revision from the most recently received draft. He then said:

“Please could you let us know if this requested wording was perhaps just missed (as it was sent separately) or if there is a reason why it was deleted?

It is important for us to clarify that SATA continues to [owe] Hi Fly the arrears of rent and maintenance reserves irrespective of the upcoming lease early termination you are working on with AELF.”

Ms Viveiros responded later that day:

“Thanks for your comment; We deleted the final paragraph because it’s redundant with the clause itself. Nevertheless, we accept the introduction of this mention, in order to avoid to have this as an open item.”

Objectively speaking, this exchange cannot show that SATA positively believed that all arrears would remain due, because the clause to which it was attached referred only to arrears relating to the period before the “Economic Closing”, i.e. 1 September 2019, and the objective reader would have understood Mr Roberts’ reference to arrears as confined to this period. But it was flatly and prominently inconsistent with any positive agreement that the entire arrears would be transferred to AELF and swept up in the termination agreement.

I can safely infer that no such message had reached Ms Viveiros, whose response shows that she at least was reading the email with care (as one would expect). Had she understood herself to be implementing a deal which involved the complete transfer of all Hi Fly’s indebtedness, she would hardly have been able to regard Mr Roberts’ request as a simple point that could be immediately conceded.

What of Mr Chaves? He was, he says (and I accept) not involved in drafting the novation agreement. He thought he would have read the email, but “very quick”. I also accept that evidence. It seems likely that Mr Chaves, reading in en masse copy a document specifically relating to a drafting point sent to a lawyer, would have scanned rather than studied it. This was also in the early stages of Covid and Mr Chaves’ plate was very full. He also said that he did not think that he had asked anyone to deal with it specifically. So far as the documentary record shows, he did not react to it.

I cannot accept, however, that even a “very quick” reading of this email would not have rung alarm bells for Mr Chaves if, as he says, his understanding at the time was that all arrears would be dealt with by the termination agreement. Confronted with this point, he said that he thought there might be “something prior to that debt that was not discussed by either AELF or Hi Fly with ourselves” and that it was possible that the email related to “rents that could be out prior 2019 or even sooner from the early beginning of the rents” of which he was not aware. He gave similar answers in relation to other occasions in which he was faced with the same or similar points. I do not think they can be an accurate reflection of his thinking at the time, for reasons I explain in paragraph 160 below.

On 20 March 2020, Mr Bagorro sent an email to Ms Gonçalves, with Mr Chaves in copy. His email stated as follows:

“Given the huge delay in the MSN970 … Novation process, I believe we have reached an unsustainable point with SATA’s debt. At this moment we have:

1 – MSN970 Dry Lease Debt: total of 2,754,214 Usd (already includes rent from March 15th to April 14th and March Maintenance Reserves).

2 – Wet Lease debt for services provided in December of 84,646 Eur, with 3 reconciliations still to be completed of a relevant amount.

The situation is quite serious and SATA has caused Hi Fly not only an enormous embarrassment with the future Lessor of the A330, but also a real and very high loss due to the delay in payments. We need you to resolve this situation immediately, so I come to request an urgent Conference Call with you and the Regional Government Representative on Monday. Even in telework [sc, during Covid], I assume that we can all contact each other but I have to have the situation resolved asap. …

I remind you that the payment of the Dry Lease debt is independent of the signature of the Novation, therefore please do not make the payment dependent on the signature of the Novation Agreement. Debt is a separate issue and has to be resolved separately.”

Mr Chaves could not recall whether he had read this email, and referred to the fact that this came at a particularly fraught time in the Covid pandemic. I think it is likely that he did read it, not because it is headed “urgent”, but because my impression of Mr Chaves is of a conscientious man who probably would, as he said he hoped he would, read an email from a significant long-term business partner relating to a topic of some obvious importance. Ms Gonçalves certainly read it because she responded to it internally (including to Mr Chaves) later that day, saying that “this thing with HiFly is getting dramatic”. And Mr Chaves certainly read her email, because he replied to it. It is unlikely that he did not, before doing so, read the short email to which she was referring. His response to her was to explain that he had signed the novation and termination agreement.

Ms Gonçalves said that when she read the email, she paid it little attention. Hi Fly, she says, was not her priority. The money available to SATA would not allow payment before the novation, and Hi Fly was “saying to me that it would deal with the debt later on”. She noticed there was a problem with the novation (which was the “drama” she referred to), but was content simply to leave things to play out and deal with any debt claims later.

Although Mr Duffy challenged some of this evidence as “implausible”, I broadly accept it. I think that it is apparent from her response and internal action that Ms Gonçalves, and Mr Chaves, did indeed see the relevant “action point” for her as being to unblock the stalled novation agreement. She is likely to have mentally dismissed urgent demands for the immediate settlement of the debt as beside the point, given SATA’s financial position and the crisis in which it was engulfed by Covid. She was more than content to leave any discussion of the debt until later, and regarded urgent demands to assemble regional government representatives to discuss a long-outstanding debt on an aircraft that was about to be returned as melodramatic fantasy. But that is itself telling, for Ms Gonçalves herself made it clear that in her understanding Hi Fly was “saying to me that it would deal with the debt later on”. She did not say that she believed the novation would extinguish all, or even any, of the debt. Ms Gonçalves never suggested this to Mr Bagorro, or internally. Nor did Mr Chaves.

As I have said, when Ms Gonçalves she received that email her focus was not on the debt—a problem, as she saw it, for another day—but on the novation agreement. It was about that topic that she responded to Mr Bagorro the following day, saying that SATA had signed the contracts and therefore “done its part” so far as novation was concerned. Her reply said nothing at all, one way or another, about the outstanding debt. Mr Chaves’ reaction was similar.

On 23 March, Mr Bagorro responded to the news that the novation agreement had already been signed by thanking Ms Gonçalves, and saying that he would press the AELF side. He did not, however, drop the topic of the debt. Instead, in an email copied to Mr Chaves, he wrote:

“If so, with Novation signed on March 18, the amount owed to Hi Fly for this Dry lease would be 2,425,410 Usd instead of 2,754 k$. It is about this value that I would like to speak to you.”

Two things are objectively clear from this email. First, Hi Fly considered that, after signature of the novation, there would remain a substantial sum due to Hi Fly in respect of the aircraft (there was no other “dry lease”). That is apparent simply from reading the email in isolation. Second, Hi Fly considered that the liability to it continued to accrue right up to the moment of execution, since this is the only explanation of the small difference in the figures (which would have accounted for the time between the novation and 14 April 2020, which if the novation had been completed on 18 March would be due to AELF, not Hi Fly).

Ms Gonçalves certainly read this email, because just over an hour after it was received, she sent an internal reply to Mr Chaves and Mr Rodrigues in the following terms:

“HiFly will want to get some… we don’t have any money… ☹️ [sad face emoji]”

Clearly, then, Ms Gonçalves realised, whether or not she had grasped the detail, that Hi Fly expected that even if the novation was completed they would be looking to pursue arrears. Her reaction to this was not that SATA would not owe any money, but that it did not have any.

Ms Gonçalves’ explanation was, in effect, that at this point she had a lot of other things on her mind, and had not been able to reconcile or verify the Hi Fly debt, including the wet lease debt. I accept that she did not, at this point, seek to bottom out Mr Bagorro’s claims about how much was owed, and that it would not have been a priority for her to do so. But that does not meet the point. Whether or not Ms Gonçalves agreed to the calculation, what she was pointing out was that Hi Fly expected to be paid something substantial, and that was sufficiently important that she thought it worth mentioning to the CEO and COO. Given the recent emails she had had about this she cannot conceivably have imagined it was all “wet lease” debt. Her internal response is inconsistent with any belief on her part that the novation agreement would transfer that debt. I have no doubt that, had she thought that, Ms Gonçalves would have raised it with Mr Chaves, and that she would then have responded to Mr Bagorro. The debt to Hi Fly was not urgent, but it was not the sort of small change that Ms Gonçalves would have simply overlooked if her actual belief was that it was not due at all.

What of Mr Chaves? In cross-examination his evidence was as follows:

“Q. So you knew that Hi Fly, even though the agreement had been signed, still thought the debt had to be paid?

A. Correct.

Q. You also must have known that was the true position.

A.

No. Basically, I was seeing that they were chasing that number, and I was expecting that someone on our side would then reconcile the number.”

On this evidence, Mr Chaves must have believed that Hi Fly might be entitled at least to be paid something after novation. “Reconciliation” is not a term that could appropriately be used, or an activity that Mr Chaves would have considered necessary, if there could be nothing at all to pay. Mr Chaves did not respond to Ms Gonçalves’ email.

When further pressed on this topic, Mr Chaves again sought to distinguish between a “past debt, for a certain amount that would be due anyway because that could be prior to the agreement” and the debt being transferred, in his mind, to AELF. Mr Chaves thus repeated essentially the answer that he had given in relation to the exchange between Mr Roberts and Ms Viveiros. Once again, I find the explanation problematic, and explain my reasons for doing so in paragraph 160 below.

In the immediate aftermath of this discussion, there was a plan for a call to take place. But Mr Rodrigues did not see “any added value” in having a call with Hi Fly, saying (in an email to Ms Gonçalves) that it was not going to be acceptable to SATA’s shareholders for SATA to promise payments to Hi Fly, and that it was not in SATA’s interests to “be used as a backup for Hi Fly to break up with [AELF]” (as I would interpret that, “as a pretext” or “a reason”, i.e. that SATA needed not to rock the boat in a way that might jeopardize the novation). The call was cancelled.

The novation agreement and its associated documents were finally executed by all the parties on 8 April 2020, which was the “effective time” for their purposes. In so far as they are relevant, I discuss their terms below.

After the novation had taken place, Hi Fly and SATA continued to discuss the outstanding rent. SATA’s subjective beliefs during this period are not, as such, directly relevant for any issue I have to decide. But they are indirectly relevant, in that it is provisionally reasonable to suppose that SATA’s beliefs would have stayed the same, unless something had happened to change them. Nobody suggests that anything had happened to change them.

The relevant communications can mostly be summarised quite briefly. Hi Fly continued to chase for payment of the rent arrears after signature of the novation. In May, it delivered pro forma invoices covering rent and maintenance reserves due up to the date of novation. It is objectively clear, as Mr Chaves accepted in cross-examination, that these were only consistent with Hi Fly considering that it was owed money up to that date.

At no point during the discussion of these invoices, or the outstanding sums that Hi Fly wanted to be paid, did SATA ever say anything to Hi Fly which suggested that they were not due. Based simply on the correspondence, the objective impression was to the contrary. In July 2020 Ms Gonçalves positively suggested that her financial director should speak to Hi Fly’s financial director “and thus close the outstanding amounts”. She said it was “important that we agree the amounts so that we can, when the state aid is made available to us, proceed with the resolution of the outstanding amounts”. On 16 July, Mr Plácido sent his counterpart at Hi Fly what he described as a “summary of SATA’s debt to Hi Fly”, which included (either as approved invoices or pro-formas) the debt due up to novation.

Ms Gonçalves suggested that this was simply a process of agreeing what had been invoiced, not accepting that it was due. Although of course it must have involved reconciling the records, it seems clear to me that it went further than that: the stated objective was to enable the debt to be paid. At no point was there the faintest suggestion that the novation agreement rendered any substantial amount of these past costs irrecoverable.

Mr Bagorro said that there were also telephone calls between him and Ms Gonçalves, to the same effect. I do not think it appropriate to make any finding as to the detail of any unrecorded and undocumented calls. It is assuredly, however, common ground that Ms Gonçalves never said anything to call the debt fundamentally into question.

More telling yet are the internal communications. When she received the pro forma invoices for rent in May, Ms Gonçalves asked Mr Plácido for an explanation: “Does the invoice below mean that we have to pay rent for the A330? Are [Hi Fly] still billing us for the A330?” Mr Plácido replied that Hi Fly were invoicing for rent up to 8 April, when novation occurred. He added “I don’t know if another kind of agreement was made on the A330 rents, but if it wasn’t, it seems to me to make sense”. Ms Gonçalves replied simply “Okay, thank you Luis”. It is hard to see that she would have done that if she had believed there had been any other agreement. Ms Gonçalves’ evidence was that she “had doubts” and therefore consulted Mr Chaves. But it certainly seems unlikely that she would have replied in this way if she had positively believed that Mr Plácido’s explanation was wrong. I cannot therefore accept her evidence that “the way I saw it, or understood it, we did not have to pay anything”.

Ms Gonçalves did, however, email Mr Chaves to ask him whether there was “something I’m missing”, asking “Do we continue to pay rent to HiFly?”. Mr Chaves’ response was: “We shouldn’t, unless it’s from past costs?”

I asked Mr Chaves what he meant by “past costs”. He gave a number of responses. He first suggested he had in mind the costs relating to the wet leases. But that cannot explain his answer, for the question he was asked was specifically about the A330, and his answer referred to “past” costs not “wet lease costs”. It therefore cannot be what he meant; nor can it have been how Ms Gonçalves understood his answer, for she would assuredly have pointed out that it was irrelevant to the A330.

His next attempt to explain his thinking was that “past” meant “before the novation period”. That would certainly make sense, and would accord with what Hi Fly seemed to intend. Mr Chaves would then be saying that he expected Hi Fly to be paid for rent accruing up to the date of novation, but not afterwards.

It would, however, be obviously inconsistent with the idea that any rent prior to novation had ceased to be payable to Hi Fly. I asked him specifically to focus on rent due in, say, January 2020, and whether he thought that would be a “past cost” or not. His reply was as follows:

“Well, my thought was that, as I said since the beginning, I thought that when we went to AELF and we closed the deal, there would be, I would not say a freezing period, but a period where everything will be paid within a package. That is my point. I did not know what was before that. So I cannot specifically say. The only thing I can say is that my mind was frozen on the 3 million that we started since the beginning. So for me it could be something before that. That was my point.”

Mr Duffy described this as a “word salad”. That is an apt description. Once again, Mr Chaves attempted to explain how some “past costs” might fall outside the transfer to AELF (and termination), while others remained within it. Again, I address this evidence—and the reasons why I cannot accept it as a reliable reconstruction of events—in paragraph 160 below.

The natural interpretation of Mr Chaves’ comment is along the same lines as Mr Plácido’s response to the same question, and with the first answer Mr Chaves gave to my question. “Past costs” meant simply costs prior to the date of novation. In effect, having been asked “do we continue to pay rent to Hi Fly now that the A330 lease has been novated and terminated” his answer was: “no, unless it is rent that was due before novation.” I also accept, however, that as the question-mark at the end of his own reply suggests, Mr Chaves himself may have had some doubt about the answer.

There were also, in July 2020, various internal discussions and analytical work attempting to reconcile the amounts that SATA thought Hi Fly had invoiced with what Hi Fly was saying was due. There is no indication in these discussions that it occurred to anyone (including Mr Chaves and Ms Gonçalves) that the novation agreement and subsequent termination was a relevant factor to be considered as reducing the debt. Nor does that seem to have occurred to Ms Gonçalves when she complained of an administrative error in issuing an invoice for EUR 5,000 “when we owe almost Eur 3 million”.

Mr Akoni suggested that I should understand the entirety of this period as one in which Ms Gonçalves was playing for time while she tried to sort out the most serious of SATA’s financial problems and somehow keep it afloat. I agree that that was her objective. But it does not follow that her failure to make any suggestion that the debt to Hi Fly had been transferred, much less the failure to identify any coherent internal communication to that effect, is not informative. The fact that Ms Gonçalves set about agreeing the amount of the debt, without internally or externally suggesting that the novation agreement needed to be taken into account in doing so, with the express purpose of being in a position to “close the outstanding amounts” when an infusion of EU-approved state aid became available, tells strongly against her thinking that Hi Fly’s demands were at odds with how she understood the novation agreement to work.

Following some further discussions in September 2020, SATA’s restructuring lawyer Mr Kikoyo took over negotiations. During discussions with Hi Fly, in October 2020 he raised the argument that the novation agreement had transferred arrears accruing after September 2019 to AELF. After setting off sums due the other way, this would have left money due to SATA, and asked for reimbursement.

In the event, SATA did not meet its payment obligations to AELF, which were eventually compromised in a court-approved settlement in December 2020. For present purposes, however, nothing turns on that.

There was no dispute before me about the principles of construction established by cases such as Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896 (HL); Chartbrook Ltd v Persimmon Homes Ltd [2009] UKHL 38, [2009] AC 1101; Arnold v Britton [2015] UKSC 36, [2015] AC 1619; Rainy Sky SA v Kookmin Bank [2011] UKSC 50, [2011] 1 WLR 2900; and Wood v Capita Insurance Services Ltd [2017] UKSC 24, [2017] AC 1173. I was referred to the useful summary originally written by HHJ Pelling KC and approved by Vos C in Lamesa Investments Ltd v Cynergy Bank Ltd [2020] EWCA Civ 821, [2021] 2 All ER (Comm) 573 at [18].

I shall not attempt another summary or restatement. For present purposes, the key uncontroversial points are as follows:

The key provisions of the novation agreement for present purposes are clauses 2.1, 2.3, and 7.3.

The definitions section defined “Economic Closing Date” as 1 September 2019 and “Effective Time” as the date when the Effective Time Certificate was executed, which was 8 April 2020.

Clause 2.1 provided:

“As of and with immediate effect from the Effective Time, and subject to the provisions of clause 2.3 (Rights and Remedies arising prior to the Effective Time; Indemnities) and clause 7 (Deposits, Maintenance Reserves and other payments):

(a)

[Hi Fly] releases [SATA] from all of [SATA]’s obligations, duties and liabilities under the Lease, and [SATA] agrees that it has no further rights against the [Hi Fly] under the Lease;

(b)

[SATA] releases [Hi Fly] from all of [Hi Fly’s] obligations, duties and liabilities under the Lease, and [Hi Fly] agrees that it has no further rights against [SATA] under the Lease …”

Clause 2.3 provided:

“Without prejudice to the rights of [AELF] under the Novated Lease and under clause 7.4, [SATA] and [Hi Fly] agree that [Hi Fly] shall have the same rights and remedies against each other as each would have had under the Lease in respect of any losses, liabilities or claims suffered or incurred or payments due to each other in respect of or attributable to the period prior to the Effective Time (or at or after the Effective Time with respect to any period prior to the Effective Time) as if [Hi Fly] had remained the Lessor under the Lease, and [SATA] shall not make any claim or exercise any right (including any set off or counterclaim) in respect of such losses, liabilities or claims against [AELF]”

Clause 7.3 provided:

“[Hi Fly], [SATA] and [AELF] agree as follows:

(a)

in respect of any Rent, Maintenance Reserves or default interest relating to the period prior to (but excluding) the Economic Closing Date, such amounts will be paid by [SATA] to [Hi Fly] (and if any such amounts are received by [AELF] after the Effective Time, then it will promptly transfer such amounts to [Hi Fly]) and such amounts will remain payable by [SATA] irrespective of any subsequent amendment or waiver occurring in relation to the Lease after the Effective Time; and

(b)

in respect of any Rent, Maintenance Reserves or default interest relating to the period after (and including) the Economic Closing Date, such amounts will be paid by [SATA] to [AELF] (and if any such amounts are received by [Hi Fly] after the Effective Time, then it will promptly transfer such amounts to [AELF])”

The basic scheme is not in dispute. Clause 2.1 “wipes out” Hi Fly’s rights, but is subject to clause 2.3 and clause 7. Clause 2.3 “writes back” any right that Hi Fly had at the date the novation takes effect, “as if” the lease had never been terminated. Clause 7.3 preserves that in respect of rent due for periods before 1 September 2019, which is to be “paid” and “remain payable” to Hi Fly, but requires SATA to pay AELF rent due on or after that date, and Hi Fly to transfer it to AELF if it is paid to Hi Fly. Mr Akoni submitted that this was entirely clear: rent due after 1 September 2019 must be paid to AELF (only), so that clause 7.3 must qualify clause 2.3 in that respect. Mr Duffy submitted that the clause was concerned only with the “mechanism of payment”, but did not alter what clause 2.3 said about Hi Fly’s entitlement to be paid.

It was common ground that in construing the agreement I could take account of certain basic facts that everyone knew. The aircraft was being sold to AELF (through various intermediate sales, presumably for tax efficiency) simultaneously with the novation. The novation recited that. The lease, upon transfer, was to be terminated. All three parties knew that. At the moment of transfer, SATA was in arrears. Again, all three parties knew that from November onwards. There had been a previous novation agreement for the aircraft, which all three parties knew about.

Mr Duffy submitted that there are some obvious errors in the agreement which might incline one to read it with the expectation that it is sloppily drafted and so should be flexibly construed. I do not agree. This was an agreement thrashed out between lawyers (or expert consultants). It was not prepared in any particular haste. It contains at least one demonstrable mistake (the mistaken survival of a reference to clause 7.4), but it is a formal and considered document, and the objective interpreter would proceed on the basis that its words have been chosen with care. (The fact that Mr Roberts had in fact made what was from his perspective a serious mistake is not something that is admissibly relevant.)

Mr Duffy submitted that construction in this case could be anchored by two fundamental points of commercial common sense. First, he said, it should be assumed that a novation would leave the lessee liable to pay rent and maintenance reserve payments to the old lessor up to the point at which ownership transferred and to the new lessor from that moment. Secondly, he said, it should be assumed that an accrued liability for rent would not “disappear”, and that since under the sale agreement Hi Fly bore the cost of all outstanding payments, it was a matter of commercial logic that it should be entitled to recover them.

I do not accept that either of those starting points is cogent. The first might be valid in many cases, but it is inconsistent with the express terms of the agreement. The introduction of an “Economic Closing Date” distinct from the “Effective Time” is evidently intended to acknowledge a potential distinction between them. That distinction is rational, in principle. It may be unusual but it is not absurd for parties agreeing that a legal transfer will be structured so as to backdate its economic consequences to some earlier time “as if” it had happened then. If nothing else is said one might expect previous rights to be preserved. But this is an assumption that can readily give way in the face of counter-indications in the agreement, and there are such counter-indications here.

The second of Hi Fly’s commercial points would involve bringing to the novation agreement’s construction a fact that was not known to SATA, namely that Hi Fly was reducing the purchase price by the amount of the outstanding payments at the effective date. But SATA did not know that to be the case, and had no reason to assume that it must be so. All SATA knew, and all that is admissible for construction, was that the aircraft was being sold. The terms could have been different. Hi Fly and AELF might have agreed that AELF would take over the aircraft with certain arrears, taking the risk of non-collection. Indeed, there was a time when that seems to have been contemplated as at least a possibility. If that was so, the transfer of the right to recover those arrears from Hi Fly to AELF would have been consistent with commercial logic. Their “disappearance” would then not be the result of the novation agreement, but of AELF and SATA’s agreed termination, in which they would be (quite reasonably) swept up.

Nor am I assisted by any consideration of what would be “usual” in a novation. In so far as there is evidence of this, the evidence of Mr Hollnagel is that the most usual position would be that the issue would never arise, because a new lessor would normally refuse to take over a plane which was in arrears.

In my view, the language and structure of the novation agreement are clear, and admit only one possible interpretation.

The parties debated whether the reference in clause 2.3 to “clause 7.4” was (as a matter of construction) an express reference to clause 7.3. There is not any clause 7.4. The objective observer would therefore know that this was an error. They would not, however, be able to decide whether the error should be cured by substituting “7.3” for “7.4” in the text, or simply by striking out the reference to “7.4” as a redundant hangover from some earlier draft, which it may have been. So, although the error is manifest, the requisite correction is not clear, and it cannot be made simply as a matter of construction: see Bellini NE Ltd v Brit UW Ltd [2024] EWCA Civ 435, at [18]–[19] (Vos C). But this makes no real difference because clause 2.1 is (by its terms) to be read subject to both clause 2.3 and clause 7, and the agreement must be construed as a whole. Even if the express reference to clause “7.4” in clause 2.3 is ignored, clause 7.3 cannot be. Since it deals specifically with a particular defined category of payment, I would resolve any apparent contradiction between clause 2.3 and clause 7.3 in favour of clause 7.3, so that the overall scheme is one of progressive qualifications: a general release (2.1), qualified by a preservation of Hi Fly’s rights accrued at or in respect of the period before the Effective Time (2.3), in turn qualified by a specific regime for rent and maintenance reserve payments, tied to the Economic Closing Date (7.3).

There is no verbal difficulty in understanding clause 7.3, and no lack of clarity in its terms. After novation, unpaid rent due in respect of periods before 1 September 2019 is to be paid by SATA to Hi Fly, and if paid to AELF is to be transferred to Hi Fly. Unpaid rent due in respect of periods after that date is to be paid by SATA to AELF, and if paid to Hi Fly is to be transferred to AELF.

Hi Fly and AELF submitted that this was concerned simply with the “mechanism of payment”: arrears of rent relating to periods after 1 September 2019 must be paid to AELF, but would remain due to Hi Fly, but become payable to AELF. I cannot accept that the clause is capable of being read in that way. So far as SATA is concerned, it would be senseless: why pay different categories of rent to different parties if all of it is due to only one? Why make AELF an agent for Hi Fly for the partial collection of some rent? It would be a finicky mechanism that served no useful objective. Nor can a “payment mechanism” explain Hi Fly’s obligation to pay on rent relating to the period from 1 September 2019, if received, to AELF, if that rent was actually owed to Hi Fly. That obligation makes sense only on the assumption that it is AELF that is entitled to obtain and retain rent relating to the period after “Economic Closing Date”—a conclusion to which that term, standing alone, in any event strongly points. The final words of clause 7.3 (a), making it clear that the clause deals with what is “payable” to Hi Fly, lends support to the idea that this was not the distinction.

I therefore conclude that the novation agreement, properly construed as a written document in the light of all the material that is admissible for that purpose, carved rent due for the period up to “Economic Closing Date” out of the general release, but left the right to recover rent due after that date within it (and therefore released by Hi Fly, and transferred to AELF).

The parties were largely agreed on the relevant principles governing rectification. Much is now clear from Leggatt LJ’s tour de force in FSHC Group Holdings Ltd v Glas [2019] EWCA Civ 1361, [2020] Ch 365. It is, however, important to identify precisely what Hi Fly and AELF need to establish if they are to be entitled to rectification.

Rectification is an equitable remedy by which the court orders a document to be altered to “put right” (rectify) some error in it. As has often been said, the court mends documents, not bargains. In other words, rectification is concerned with things that go wrong in the way the parties have embodied an agreement in writing, not with its substance.

When dealing with a written contract, there are three cases where rectification may be ordered:

The principled rationale for rectification in each of these categories is different. Category 1 rests on the principle that agreements should be honoured: the court rectifies a written instrument so that it corresponds to what the parties agreed: see FSHC at [142]. In contrast with that, but consistently with each other, Categories 2 and 3 both rest on the perception that it may be inequitable to take unfair advantage of someone else’s known mistake. As Leggatt LJ put it at [105]:

“In the case of common mistake it is inequitable for a party to the contract to seek to apply the contract inconsistently with what that party knew to be the common intention of the parties when the written contract was executed. The doctrine of unilateral mistake extends this principle to the situation where a party seeks to apply the contract inconsistently with what that party knew the other party believed to be the common intention of the parties when the written contract was executed.”

This is a different rationale from that which applies in Category 1 cases:

“We do not, however, accept that the same reasoning [as applies in Category 1 cases] can be applied to a situation in which parties have not made any prior contract but had a common continuing intention in respect of a particular matter in the document sought to be rectified. … [T]hat was not historically the principle on which equity interfered with written contracts which mistakenly failed to reflect the common intention of the parties; nor in our view does it provide a proper basis for such interference. Rather, rectification to give effect to a ‘common continuing intention’ not amounting to a legally enforceable contract is justified, and is only capable of being justified, as an instance of the second form of rectification, based on an equitable principle of good faith.”

In Category 2 and 3 cases, rectification goes beyond simply bringing a document into conformance with a prior binding agreement. It imposes terms which were never contractually agreed. It still, however, draws sustenance not from any analysis of the substantive fairness of the bargain, but from standards of fair dealing in the process by which an agreement has been reached, recorded, and enforced.

FSHC established that the “agreed common assumption” required for Category 2 cases requires an analysis of the parties’ subjective intentions, not their objective prior agreement, and it is in precisely this respect that Category 2 cases differ from Category 1, and from construction. So rectification differs from construction in the range of material that can be considered (for instance, evidence about negotiations remain inadmissible for construction, but admissible in a rectification claim). More fundamentally, however, it sometimes also differs in the question being asked. The inquiry in Category 1 cases, as in construction, is about what the parties’ words and conduct would objectively convey about their joint intention to a reasonable observer. The inquiry in Category 2 and Category 3 cases is about the parties’ subjective intentions, and their subjective mutual understanding of those intentions.

What then of the requirement for “accord”? The Court of Appeal was quite clear in FSHC that it is insufficient to show, for a Category 2 case, a mere coincidence of mistaken intention. Instead, Leggatt LJ pointed out at [76]:

“Joscelyne v Nissen [1970] 2 QB 86 (CA) clearly held that it is essential for rectification of a written contract to show an agreement, not in the sense of a prior concluded contract but ‘in the more general sense of an outwardly expressed accord of minds’, and this requirement has been affirmed by the Court of Appeal on many subsequent occasions.”

Moreover, such a requirement was, in his view, “sound in principle”, since rectification in Category 2 cases “is only justified if [the revised terms] are founded on mutual agreement”.

This affects what we should understand by “accord” in Category 2 cases. One possible meaning would be that it required an objective analysis of the parties’ dealings to ascertain whether a reasonable observer would have concluded that they shared this mistaken assumption. But if it required that, then Category 2 cases would collapse back into Category 1: we would be seeking a “prior agreement” in the objective sense. The Court of Appeal in FSHC was clear that this would be an error, because Category 2 cases rest on different principles and require subjective evaluation: see [183].

What the court meant is clear from its summary of the relevant principles for Category 1 and Category 2 cases, at [176]:

“[B]efore a written contract may be rectified on the basis of a common mistake, it is necessary to show either (1) that the document fails to give effect to a prior concluded contract or (2) that, when they executed the document, the parties had a common intention in respect of a particular matter which, by mistake, the document did not accurately record. In the latter case it is necessary to show not only that each party to the contract had the same actual intention with regard to the relevant matter, but also that there was an “outward expression of accord” – meaning that, as a result of communication between them, the parties understood each other to share that intention.”

The “outward expression of accord” is therefore to be found when “as a result of communication between them, the parties understood each other to share that intention”. That does not mean that the reasonable objective observer of the communications would have concluded that they were ad idem on the point, for that would be a different test, and one that the Court expressly rejected. We should understand “accord” and “agreement” in a sense which encompasses a situation in which the parties share an assumption because they both (subjectively) make it and know from their mutual communications that each is making it. So understood, Category 2 and Category 3 cases are close to each other. They share the critical factor that the person now seeking to take advantage of the other’s mistake knew when the contract was concluded that the other party was making an assumption that would be falsified unless the contractual words are changed.

It follows, as FSHC shows, that there is not necessarily any difficulty in relying on “silence” (or an absence of communication) to demonstrate the necessary “accord”, as there may sometimes be when mutual dealings are to be analysed as “offer” and “acceptance”. What matters is that it was known to each, from their mutual dealings, that they shared a particular intention. There may be cases (like Frederick E Rose (London) Ltd v William H Pim Jnr & Co Ltd [1953] 2 QB 450, as it would now be understood) where we know that the parties shared the same assumption because they told each other so. And there may be cases (like FSHC) where we know that they shared the same assumption because it was obvious that they would have done unless something else had been said. In either case, there is “accord”. It is in this sense that the Court of Appeal in FSHC compared the situation in that case to the “dog that did not bark”.

It might be thought that this makes the “accord” requirement rather easy to satisfy. But it would be a serious mistake to think that the Court of Appeal in FSHC saw rectification as an easy hurdle to clear. On the contrary, at [174], Leggatt LJ, addressing an objection that a subjective approach would make rectification too difficult, said:

“In our view, that fact that a ‘subjective consensus’ (that is to say, a common intention in the sense we have described) is harder to prove than an ‘objective consensus’ is not an objection to adopting a subjective test for rectification but a positive merit of such a test. As a matter of policy, rectification should be difficult to prove. The reasons for adopting an objective test of agreement which makes it easier to establish a legally enforceable contract than would a subjective test are not reasons for making it easier to alter such a contract. We agree with the response of Professor Paul Davies that: ‘Formal, written contracts should be presumptively upheld and instances of rectification should be rare. Any other approach would undermine the importance commercial parties put on the final written agreement.’”

The point, as Mr Akoni rightly emphasised, is that the key requirement for rectification is proof of a common subjective mistake. Once that is proved, it may indeed often be relatively easy also to establish that the mistake was shared (mutually known from the parties’ dealings) and not merely coincidental, because when parties to a contract share a common intention they will usually have derived it from their mutual communications or shared it during them. But the court will not rectify a contract simply because one party could or should have realised that the other was mistaken, or because the objective observer reviewing their mutual dealings would have thought the error was common to both. The proven starting point must be that both had, in fact, a definite positive subjective intention about how their agreement would operate, and that they had the same definite positive intention. The “accord” requirement is added on top of the cardinal feature, which is the positive subjective mistake which must be distinctly proved.

The requirement for a positive intention is important, and is illustrated by Ralph v Ralph [2021] EWCA Civ 1106, [2022] 2 All ER 325. It is plainly not enough that the parties have not considered a point, or that one of them has and the other has not. It is not even enough that they would or might, if they had considered it, have reached a common view. What needs to be shown is that both of them had the same assumption about how that point was to be addressed in their agreement. Thus, in Ralph, the fact that the protagonists had not agreed to share a property as tenants in common in equal shares did not suffice to rectify a declaration that they did (at [40]–[41]):

“I can see no finding by the trial judge that Dean and David agreed ‘joint legal title only’, even if that may, by deduction, be the correct legal analysis of what occurred. Morris J had already said … that the trial judge had ‘found that, not only was there no agreement as to sharing of beneficial ownership, but he also held that there was “no discussion” about how the Property should be held’. In those circumstances, it was impossible to find a sufficient, or any, continuing common intention that there should be no declaration of trust.

… The law does not make contracts for people unless they have, in the way explained in FSHC, agreed to them or shown a continuing common intention as to the term or terms in issue. Here Dean and David, on the evidence found by the trial judge, simply gave not thought to the matter at all.”

It is sometimes said that rectification must be proved by “convincing evidence”, and that the evidence must be strong enough to overcome the presumption that the written agreement records the parties intentions: see (for example) FSHC at [46]. However, references to “convincing evidence” should not be taken to mean that the relevant subjective intention is to be proved on anything other than the ordinary balance of probabilities. As Leggatt J pointed out in Tartsinis v Navona Management Co [2015] EWHC 57 (Comm) at [85] “It is not, in truth, the standard of proof which is high, thereby differing from the normal civil standard, but that sufficiently strong proof is needed to counteract the evidence of the parties’ intention displayed by the instrument itself”. The strength to be accorded to the presumption that the executed agreement offers a reliable guide to subjective intention must vary from case to case (see Tartsinis at [86]). An executed agreement is a definitive guide to “objective intention”. But if, for instance, the relevant individuals whose subjective intentions are being investigated did not read or write the executed agreement, or did not do so with any care (as appears to be the case here), the executed agreement may provide rather weak evidence of their actual subjective beliefs. The key safeguard against over-zealous rectification is not so much the means of proof, as the probandum. What must be proved, on a balance of probabilities, is a positive shared subjective intention on the relevant point.

In a two party case, the “mutual mistake” must be that of both parties. In a three party case, at least where (as here) the rights to be altered are those of all the parties, it must be the mistake of all the parties. This was not controversial.

Where one is dealing with legal persons the familiar problems of attribution that arise when considering a legal person’s “state of mind” must be confronted. The parties here agree that the proper inquiry is into the state of mind of the natural person or persons who have taken the decision to enter into the written agreement. In Murray Holdings Ltd v Oscatello Investments Ltd [2018] EWHC 162 (Ch) at [198], after analysing the relevant authorities, Mann J explained that one was “looking for the person who is in reality the decision maker in the transaction”. That would usually be the person with “authority to bind the company”, but need not be if the “reality on the facts” is that some other person was the decision maker. The intention and understanding of those who are not the relevant decision makers may be evidentially relevant because it may in fact be likely that they communicated their intentions to the decision maker.

Since Hi Fly and AELF put their case alternatively on the basis of unilateral mistake, I should say something more about that, though for reasons that will become apparent my decision will not turn on it. The classic exposition is that of Buckley LJ in Thomas Bates & Sons Ltd v Wyndhams (Lingerie) Ltd [1981] 1 WLR 505 (CA), at 515–6:

“For this doctrine to apply I think it must be shown: first, that one party A erroneously believed that the document sought to be rectified contained a particular term or provision, or possibly did not contain a particular term or provision which, mistakenly, it did contain; secondly, that the other party B was aware of the omission or the inclusion and that it was due to a mistake on the part of A; thirdly, that B has omitted to draw the mistake to the notice of A. And I think there must be a fourth element involved, namely, that the mistake must be one calculated to benefit B. If these requirements are satisfied, the court may regard it as inequitable to allow B to resist rectification to give effect to A's intention on the ground that the mistake was not, at the time of execution of the document, a mutual mistake.”

The party against whom rectification is sought must be aware of the other party’s mistake (that is, either actually know it of or wilfully shut its eyes to it): see Global Display Solutions Ltd v NCR Financial Solutions Group Ltd [2021] EWHC 1119 (Comm) at [447] (Jacobs J). But what must be known? Is it enough that B knows that A thinks the agreement contains terms to a particular effect? Or is it necessary that B knows that and also knows that the agreement does not do so? If the former is sufficient, rectification for unilateral mistake extends to situations in which B knows of A’s intention, but may have no positive intention of its own, and neither know nor care whether the agreement will in fact do what A thinks it will do. If the latter is required, it is available only in cases where B, when it enters into the contract, positively believes that its true effect will come as a nasty shock to A. Put differently, must B know “of A’s mistake” in the sense of knowing the thing A believes that turns out to be mistaken? Or must B know “of A’s mistake” in the stronger sense of knowing that it is a mistake when the contract is concluded?

When “mistake” is used in the context of bilateral mistake, it is used in a wide sense. The parties “share a common mistake” when they both have the same belief about what their agreement will do, although neither (obviously) knows that it is a mistake. What is “inequitable”—and what the court will therefore prevent—is one of the parties subsequently insisting on holding the other to what is (as it turns out) a mistake. The inequity arises if either party seeks subsequently to take advantage of the mistake. Mr Akoni submitted that this was insufficient for a unilateral mistake. Although dishonesty is not as such required where there is actual knowledge (see Global Display Solutions at [448]), something close to it should be. Neither counsel was able to identify any decision directly on point, and the question did not receive extensive debate in argument.

I would provisionally conclude that it is not necessary to show that the party resisting rectification realised that the other party was mistaken, and sufficient if (at the time of contracting) that party actually knew that the (as it turns out, mistaken) party had the positive belief that it did. As Leggatt LJ put it in FSHC: “The doctrine of unilateral mistake extends this principle to the situation where a party seeks to apply the contract inconsistently with what that party knew the other party believed to be the common intention of the parties when the written contract was executed.” This does not depend on showing that the non-mistaken party has specifically considered whether the mistaken party was correct or not. But it remains a testing standard. It is not sufficient that the unmistaken party knows that the mistaken party hopes or guesses that something will happen. The unmistaken party must know that the mistaken party believes, and is entering into the contract on the assumption, that the parties share a common intention that the agreement will have a particular effect.

There is one other potential issue that was not explored in argument. In the case of a non-shared mistake in a tri-partite agreement, where party A is not mistaken but parties B and C are (and share a common mistake), is it necessary that party A knows that both B and C are making that mistake? Or is it enough if party A knows B is mistaken, but does not know that C shares that mistaken belief?

Mr Akoni did not seek to argue that the rectification claim should fail for want of specific evidence about SATA’s knowledge of AELF’s intentions. In principle, it seems to me, that is correct. It would be appropriate to allow rectification in the circumstances described above. Even though it would not, then, be inequitable for A to “take advantage” of C’s mistake (of which A is ignorant), it plainly would be inequitable for A to take advantage of B’s mistake, and A’s ignorance of C’s error provides no positive counter-equity in A’s favour. Nor is there any reason not to rectify the agreement against C, since as between C and B there is an equity that arises from the shared mistake. (It would be quite different, of course, in a three-party case where only one of the parties is mistaken: there it would plainly be essential to prove that both the other parties knew of or shared the mistake, otherwise the rights of an entirely innocent contracting party would be altered.)

It follows that, to obtain an order to rectify the novation agreement on the basis of mutual mistake, Hi Fly and AELF must establish the following on a balance of probabilities:

I shall start with Hi Fly, in relation to which the evidence is clear and the conclusion undisputed. Hi Fly believed that it would remain entitled to be paid, by SATA, all rent and maintenance reserve payments relating to the period before 8 April 2020, and default interest on them.

SATA accepts that the same conclusion applies to AELF. Mr Hollnagel’s evidence is that he believed that “everything that SATA owed prior to that actual closing date was between it and Hi Fly”. Contemporaneous documents support the proposition that that was indeed his view, since he wrote to Mr Roberts on 5 March that Hi Fly “should have received the rent for the period of time that has past, or should receive it in the future, thereby enabling them to maintain the original economics of the transaction” even though the price was to be reduced by sums due.

SATA’s case is that Mr Chaves positively believed that the novation agreement would mean that SATA was no longer liable to pay Hi Fly any arrears of rent. As Mr Akoni put it orally in closing “His intention was to discharge SATA's liability for contractual sums up to the effective time by agreeing to pay AELF USD 23 million as a termination fee”, so that “he did not intend that Hi Fly would be entitled to recover, or to be paid any of the contractual sums up to the effective time”. In other words, SATA’s case is not that Mr Chaves correctly understood the novation agreement as I have construed it. Its case is that Mr Chaves was making a different mistake of his own: he thought that the agreement would transfer all arrears to AELF, so that they could then be settled.

Up to 29 January 2020, there was no reason why anyone at SATA, including Mr Chaves, would have given much thought to what the novation agreement would do to any rent arrears. Until then, Hi Fly’s consistent position was that all arrears must be cleared as a precondition of novation. Some slight uncertainty may have been injected into this by Mr Hollnagel’s threats to default the lease, which made sense only if (contrary to the prevailing expectation) AELF “took over” the debt. But that is a thread that seems, at that point, not to have been pulled, and by December it was again clear. Although Mr Duffy invoked this point as one in Hi Fly’s favour, I consider it, in itself, largely neutral. Until Hi Fly dropped the condition, the question of what “would happen” to arrears was academic. But I agree with Mr Duffy to this extent: if Hi Fly’s position had always been that novation could take place only if Hi Fly was paid, any change to that position would have been something of note and moment.

Nor do I consider that the fact that Hi Fly continued to invoice for rent helps its case. Whether or not Mr Chaves even knew that was happening, it would have been expected and neutral. Rent was plainly due to Hi Fly until the novation was signed. There had never been any agreement to suspend it while negotiations were ongoing. As things stood, Hi Fly was the right party to invoice for it.

On 29 January, however, the previous expectation that Hi Fly would insist on full payment of arrears as a precondition of novation changed. From that point onwards, the statements from Hi Fly consistently showed that although it was prepared to postpone discussion of the arrears until after novation, it still expected to be paid. The change related only to when, not whether, SATA would settle the arrears.

That was, as SATA’s witnesses accept, the position on 29 January 2020. The correspondence around that date cannot be read in any other way, for it was clear that Hi Fly was at that point postponing until after novation any discussion of how the debt would be settled, not forgiving the debt. SATA’s witnesses accept that, since Mr Chaves’ evidence was that this was his expectation when he arrived for the meeting on 7 February 2020, and only changed with the “changed reality” which he said dawned on him at that meeting.

As to the meeting itself, I have accepted Mr Chaves’ account of the discussions about default, and that they made it at least plausible that AELF might be able to acquire the debt (whether or not it was expecting to do so). But there is a striking absence of any recorded perception that the meeting was a watershed moment in this respect. The topic was not explicitly discussed. Mr Hollnagel did not deploy any ability to compromise the debt as sweetening the offer he was now making. The “changed reality” that Mr Chaves now says he felt did not result in any fresh instructions to those who were preparing the deal documentation at SATA, though Mr Chaves can have had no reason to think that the existing documentation would reflect a new position. No attempt was made to contact Hi Fly to make sure they also agreed. If this was a moment when the scales fell from Mr Chaves’ eyes, they fell silently and unremarked.

As to what followed, I attach little weight to the exchange of emails on 28 February 2020, and in particular to Mr Bagorro’s glancing reference to “double penalization” in the context of a request to speed up the novation. That was a vague statement, which could reasonably have been understood by someone who thought the novation would shift the debt to AELF as a complaint that by slowing that process AELF was preventing Hi Fly from getting paid for the arrears both directly (by not paying Hi Fly) and indirectly (by standing in the way of a novation through which, it might be assumed, Hi Fly would be paid out).

The correspondence in March, however, seriously undermines Mr Chaves’ current recollection. It was transparently clear (and Mr Chaves knew) that Hi Fly expected to remain entitled to payment of rent arrears after novation. If Mr Chaves had believed that the novation would transfer all those arrears from Hi Fly to AELF, whereupon they would be dealt with as part of the termination, he would inevitably have reacted quite differently to how he did, both as a matter of commercial ethics, and as a matter of commercial prudence, whether or not he understood precisely what the debt amounted to. There were at least three occasions on which he would have done so: when confronted with Mr Roberts’ amendment to the novation agreement on 13 March 2020, when considering Mr Bagorro’s email of 20 March 2020, and when considering Mr Bagorro’s email of 23 March, and Ms Gonçalves’ reaction to it. He cannot have believed at the time that the debt referred to on these occasions was unrelated to the aircraft, or that it concerned wet leases. He cannot have believed anything other than that Hi Fly thought that it would remain entitled to (at least some) arrears incurred in relation to the aircraft.

Mr Chaves’ explanation of his reaction is that he considered that there could be “past” debts which would remain with Hi Fly. I have already said that I found those explanations difficult to understand or accept (see paragraphs 75, 88, and 101). Their basic outline was in each case that because Mr Chaves was focusing (or, as he put it on one occasion, “frozen”) on the figure of USD 3 million representing the outstanding rent in October 2019, as given in Mrs Azevedo’s memorandum, he thought that some debt arising in the distant past might remain due to Hi Fly. However:

I cannot, therefore, regard Mr Chaves’ answers where he referred to some category of “past” debt as remaining due to Hi Fly as coherent or cogent. More importantly, however, the explanation (even if it were correct) would still be fundamentally at odds with Mr Chaves believing that all debt was to be transferred—which must become a belief that some (unspecified) debt was to be transferred.

For those reasons, it is clear (well above the threshold of the balance of probabilities) that Mr Chaves cannot have had, at the time, the belief that he now says he had. He cannot have thought that the novation agreement would transfer all Hi Fly’s debt to AELF. I would reach that conclusion even if there had been no correspondence after novation; but in fact that correspondence (including, in particular, Mr Chaves’ response to the question whether Hi Fly should continue to be paid) is confirmatory.

That, however, does not suffice to decide the rectification claim in favour of Hi Fly and AELF. For it is not enough that Mr Chaves did not have the belief that he has now, wrongly in my view, persuaded himself he had. SATA need prove no positive intention. It is for those seeking rectification to show that he had a continuing positive belief that the totality of the arrears would remain payable to Hi Fly. If Mr Chaves was probably confused, or equivocal, about this—even if his confusion or equivocation was objectively unreasonable—Hi Fly will have failed to prove the requisite positive intention on SATA’s part.

This might be an easy question to resolve if the impression I had of Mr Chaves was of a dishonest or evasive witness, for I might readily assume that a person who was now dissembling consciously about his beliefs was doing so to cover up an inconvenient truth. But that was not my impression. I did not find Mr Chaves evasive, and there is no hint that his dealings with Hi Fly were other than in complete good faith. I found him confused and inconsistent, torn between a candid acceptance that his communications at the time did not show any sense that the novation was critically important to Hi Fly’s rights, and a present conviction that in some way it was. So I have to decide, on a balance of probability, whether this is a confusion and inconsistency that has emerged as a result of the litigation, or whether it reflects confusion and inconsistency at the time.

There are, it seems to me, three potential indications or sources of confusion at the time:

These are, indeed, complicating factors. But they are not strong ones.

The first rests on inferences not express agreement, and it does not appear to have been an inference that Ms Gonçalves or Mr Hollnagel drew from the discussion.

The second places heavy weight on a punctuation mark, in circumstances where the natural reading of Mr Chaves’ response is clear, where Mr Chaves’ current reconstruction of his thinking is impossible to accept. In the period after novation, there is no sign that anyone at SATA, including Mr Chaves, did anything in relation to the Hi Fly debt which would have suggested any significant doubt that if it related to pre-novation rent, it would be payable. Since the natural reading of Mr Chaves’ reply is that he thought pre-novation rent was payable to Hi Fly, the document in fact supports the proposition that that was his view, especially when the alternative explanations he has offered are not coherent.

The third would be a telling point if Mr Chaves had read the document concerned with care, but is not relevant when he denies that he would have done so. When he was cross-examined about his reaction, even with the document in front of him, the explanation he gave about “past” rent was not keyed to the difference between the Economic Closing Date and the Effective Time. But by his own account there was nothing ever said to him which would make sense of the explanation he gave: his sense of the position on 8 October 2019 cannot do so, for reasons I have explained.

The proposition that Mr Chaves shared Hi Fly’s view is supported by the overall probabilities of the situation. Although I do not think the point helps when it comes to contractual construction (given the novation agreement’s careful distinction between economic and effective closing) I accept one would intuitively expect significant pre-novation arrears due under an aircraft lease to remain payable to the previous lessor. Mr Roberts evidently regarded it as an uncontroversial assumption. It was Mr Plácido’s immediate reaction when he was asked about it, after novation, by Ms Gonçalves. It is obviously not an iron rule, but it provides a reasonable starting point, and, without reading the draft novation agreement, it would have been Mr Chaves’ starting point.

Until just about a week before his meeting on 7 February 2020 Mr Chaves expected Hi Fly to insist upon payment before novation. The only thing that had changed before his meeting with Mr Hollnagel was that Hi Fly had said it would discuss that later, and everyone at SATA (including Mr Chaves) understood that it would have to be addressed, with Hi Fly, in due course. So Mr Chaves went into that meeting, as he accepts, expecting that Hi Fly would need to be paid its arrears, not AELF. If the novation agreement had been signed on the morning of 6 February 2020 clause 7.3 would—on Mr Chaves’ admission—have been contrary to his expectations and intentions: it would have taken him by surprise.

For those expectations to change would, I think, have required more than inference from one aspect of Mr Hollnagel’s negotiating tactics. It would have called for express discussion. I would also expect it to have left a mark (at least internally) in SATA’s communications and documents. But I can detect nothing of the kind. Mr Chaves, as I have pointed out, gave no instructions to those drafting the agreements to ensure that any “changed reality” was reflected in them; he asked no questions to clarify any uncertainty with anybody (including Hi Fly); he gave no instructions to Ms Gonçalves either; he watched—albeit without much involvement—as the discussions with Hi Fly continued in every discernible respect along the previous lines. Apart from Mr Roberts’ email of 13 March 2020, and then only if it was closely read with an eye to the precise terms of clause 7.3(a) in a way that Mr Chaves did not, there was no hint of any distinction being drawn between different categories of arrears. If Mr Chaves was reading this correspondence rapidly and without great interest, as he probably was, his natural assumption would be that Hi Fly’s expectations related to all the arrears. If he was reading it closely and with attention, that was plainly the case.

This pattern, taken as a whole, seems to me not only inconsistent with SATA’s primary case, but really consistent only, once one understands that he went into the meeting sharing Hi Fly’s expectation, with Mr Chaves sharing the same view as Hi Fly and AELF admittedly did throughout the entire process. I take fully into account the dangers of reading between the lines, or supposing that Mr Chaves would have given meticulous attention to the issue when he had so much else to do. But I consider that the proper inference here is not that he was baffled or confused about the position, but that he did (correctly) attribute to Hi Fly the expectation they very obviously and naturally had, and shared it. Again, the post-novation dealings tell the same consistent story.

I am therefore satisfied on a balance of probabilities that in the period leading up to the novation, and at the time when it was executed, Mr Chaves in fact shared, as a positive intention, Hi Fly and AELF’s mistaken expectation that the novation agreement would leave Hi Fly entitled to recover arrears predating it. In reaching that conclusion, I do not find that Mr Chaves gave deliberately untruthful evidence. But his present reconstruction of his thinking is irreconcilable with the documentary record. It is therefore likely that the wish to have believed that the novation would address past debt, and knowledge that SATA’s lawyers now consider that it partly did so, has been father to the thought that this was his belief at the time. Although it is not difficult to believe that someone in Mr Chaves’ position who had read the draft novation agreement might think the position unclear, and surmise that perhaps Hi Fly’s expectation of payment related only to arrears predating September 2019, Mr Chaves had not done that. In other words, although confusion would have been understandable, I do not think there is sufficient reason to think that Mr Chaves was, at the time, in any doubt.

As to the need for some “accord”, SATA did not ultimately contend that if the parties’ common positive intention was that SATA would remain liable to Hi Fly, there is not sufficient material demonstrating a meeting, rather than a mere coincidence, of minds. That material consists principally of the correspondence between Hi Fly and AELF, and Hi Fly and SATA, which are undoubtedly quite sufficient for this purpose. There is relatively little to suggest any communications between AELF and SATA from which the shared assumption can be derived, but SATA did not pursue submissions in that respect. It would, I think, have to be said that in this respect it is rather the absence of anything to contradict the assumption which would justify the conclusion that SATA would have assumed that AELF, too, shared it. I consider that a reasonable assumption.

My conclusion about SATA’s belief makes it unnecessary to address the claim for unilateral rectification. But, on the findings of fact that I have made, it would lead to the same conclusion.

Mr Chaves knew that Hi Fly was expecting to be paid arrears. A person who had carefully read the novation agreement, or even Mr Roberts’ email of 13 March 2020, and had not carefully read the detail of Hi Fly’s emails (especially that of 23 March 2020) might have assumed that this expectation related only to rent accrued before 1 September 2019. But Mr Chaves never had clause 7.3 in mind. His only reason for thinking that anything might be transferred to AELF was the comments about default that Mr Hollnagel had made at the meeting on 7 February 2020. He also knew, however, that Hi Fly had not been at that meeting. Such a person would have had no reason not to take Hi Fly’s demands in their most natural sense, as relating to the entirety of the outstanding rent at novation. On that basis, Mr Chaves knew that this is what Hi Fly believed. Whether that belief was positively inconsistent with his own (as he now says) or simply an expression of clear belief about something he was himself uncertain about, that would suffice to justify rectification. This might reflect rather badly on the accuracy of Mr Chaves’ ethical compass. I should make it clear that I do not think Mr Chaves is the sort of person who would have sought to take advantage of a business partner’s mistake. Quite the reverse. But that favourable conclusion about Mr Chaves’ character supports my view that he did not do anything of the sort. He allowed Hi Fly to believe that the arrears would be addressed after novation because that is what he himself also believed would happen.

For those reasons, I hold that on its true construction the novation agreement transferred Hi Fly’s right to be paid rent and maintenance reserve payments accruing from 1 September 2019, and default interest on them, to AELF.

I also find, however, that Hi Fly and AELF are entitled to have the novation agreement rectified to reflect a shared common intention that all such sums relating to the period prior to the Effective Time are to remain due to Hi Fly. This is an unusual finding to make in a commercial context, and I do not make it lightly, since the basic principle that courts apply seriously intended and carefully drafted agreements “as they are written” does much to promote commercial certainty. There are three features of this case which take it outside the norm. First, the very point in issue was specifically the focus of Hi Fly’s attention, and the subject of regular communication around the time the novation agreement was being made. Second, and unusually, SATA has accepted that Hi Fly and AELF did have the same positive intention, and that the contract as drafted was (from their perspective) mistaken. Third, this is a case where the presumption that the executed instrument accurately records the parties’ agreement is comprehensively rebutted, despite the expert care that went into preparing it. It is common ground that it does not represent AELF or Hi Fly’s intention. And it is also common ground that it does not provide useful evidence of Mr Chaves’ intention, and did not inform his intention, since he never read it or paid it attention. Had he done so, and had he based his understanding of the deal upon the draft novation agreement, the case would have looked very different.

The simplest route to rectification is to substitute “Effective Time” for “Economic Closing Date” in clause 7.3, even if that makes the entire clause redundant. The alternative would be to delete clause 7.3 entirely. However the agreement is corrected, it follows that Hi Fly is also entitled to a money judgment. I shall if necessary hear counsel on the precise form of order, and on costs.