Songa Product and Chemical Tankers IV AS v Gardsea Shipping Inc [2026] EWHC 1559 (Comm)

[2026] EWHC 1559 (Comm)Case No CL-2025-000008
IN THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES
COMMERCIAL COURT (KBD)
Venue Royal Courts of Justice, Rolls Building, Fetter Lane, London, EC4A 1NLDate 23 June 2026
Before
Paul Stanley KC(sitting as a Deputy High Court Judge)
Between
Songa Product and Chemical Tankers IV ASClaimantGardsea Shipping IncDefendant
Julian Kenny KC and Charles Connor (instructed by Mills & Co Solicitors Ltd) for ClaimantTom Corby for DefendantHearing Hearing dates: 9 June 2026
Approved JudgmentThis judgment was handed down remotely at 10.30am on 23 rd June 2026 by circulation to the parties or their representatives by e-mail and by release to the National Archives..............................

Paul Stanley KC:

[1]This is an appeal under section 69 of the Arbitration Act 1996, for which leave to appeal was granted by Butcher J on 13 November 2025. The issue is one of construction of a ship sale contract based on the Saleform 2012.[2]In 2022, the claimant (the “Sellers”) agreed to sell the MT Songa Coral to the defendant (the “Buyers”) for USD 25 million. The contract (the “MOA”) was based on the Saleform 2012. It required payment by release from a sum in escrow, in Norway, not later than three “Banking Days” after notice of readiness was given. It defined “Banking Days” as days when banks were open in various places, including the United States of America. As events transpired, the three Banking Days expired at the end of 8 September 2022. By midnight in Norway at the end of 8 September 2022, payment had not been made. The Sellers almost immediately gave notice to cancel the MOA. If payment was required by midnight in Norway, then they were entitled to do so. But it was not then yet midnight in every part of the world whose bank-opening dates were relevant to whether it was a “Banking Day”, so that in some of them (the last of which would be the State of Hawaii), it was still 8 September 2022. A distinguished LMAA arbitration tribunal has concluded that the notice was therefore too early, and that the first moment at which it could be said that the Buyers were in default under the MOA was midnight in the most western time zone of the United States, Hawaii-Aleutian time. The question of law before me is whether, as a matter of construction, that conclusion is correct.[3]I have concluded that the arbitrators’ conclusion is incorrect in law. The MOA’s definition of Banking Days served only to identify that 8 September 2022 was the third such day after notice of readiness was given. The question whether payment was made on that day is determined by whether it was made before midnight, local time, in the place where the obligation to release the balance of the purchase price was to be performed, which was Norway.

The factual background

[4]The strictly relevant factual background (that is to say the facts necessary for the determination of the issue before me) consists solely of the MOA and the Escrow Agreement. But it is easier to understand the issue if one understands the background out of which the issues arise.[5]The MOA dated 8 July 2022, was based on, but in some respects varied, the Saleform 2012. The key provision for payment was clause 3, which provided as follows (I have emphasised those provisions which were added to the standard form):
“The Buyers shall minimum one (1) Banking Day prior to the expected date of delivery of the Vessel hereunder preposition the balance of the Purchase Price and any other amounts payable by the Buyers to the Sellers under this Agreement on the escrow account with the Escrow Agent. The balance of the Purchase Price remains at the order of the Buyers only and shall only be release[d] to the Sellers upon written instruction from Buyers concurrently with Delivery of the Vessel. On delivery of the Vessel, but not later than three (3) Banking Days after the date that Notice of Readiness has been given in accordance with Clause 5 (Time and place of delivery and notices): (i) the Deposit shall be released to the Sellers; and (ii) the balance of the Purchase Price and all other sums payable on delivery by the Buyers to the Sellers under this Agreement shall be released from the escrow account [those words replaced the Saleform’s text ‘paid in full free of bank charges’] to the Sellers’ Account and shall constitute full completion of the Buyers’ payment obligations under this Agreement but without prejudice to provided herein.”
[6]The term “Banking Days” was defined under the MOA as follows:
“ ‘Banking Days’ are days on which banks are open both in the country of the currency stipulated for the Purchase Price in Clause 1 (Purchase Price) and in the place of closing stipulated in Clause 8 (Documentation) and United States of America, Canada, United Kingdom, Switzerland, Turkey, UEA [sic], Greece, Norway (add additional jurisdictions as appropriate).”
[7]As of 8 September 2022 I believe that the various countries mentioned had the following relative times relative to UTC, moving from east to west: UAE + 4 hours Turkey, Greece + 3 hours UK + 1 hour Switzerland, Norway + 2 hours Canada (6 time zones) –2.5 hours to –7 hours United States of America (6 time zones) –3 hours to –10 hours[8]Most of the changes to clause 3 reflected the parties’ agreement that sums due under the MOA (both the deposit and the balance of the purchase price, which was USD 25 million) should be held in escrow. Lawyers in Oslo (also, as it happened, the Sellers’ lawyers) were appointed as escrow agents under a tripartite agreement dated 19 July 2022. That agreement identified the “escrow account” as an account at Nordea Bank in Norway to be operated by the escrow agent.[9]Under the MOA, the Sellers’ Bank was given as Nordea Bank, Norway Branch, and a specific account was identified.[10]Had everything gone according to plan, the sequence of events would have been as follows:(a) the 10 percent deposit would have been paid to the escrow account at Nordea Bank;(b) the 90 percent balance of the purchase price would have been in the escrow account one Banking Day before the expected date of delivery, held to the Buyers’ order;(c) notice of readiness would have been given; and(d) no later than three Banking Days later, the deposit would have been released (not necessarily to the Sellers’ account) and the balance would have been released from escrow to the Sellers’ account, also at Nordea Bank.[11]This did not happen. The Tribunal’s award did not go deeply into the reasons for that, and for present purposes they do not matter. It is common ground that I should assume, for present purposes, that notice of readiness was given on 2 September 2022. That was a Friday, so 3 and 4 September were not Banking Days. Nor was 5 September, because it was Labor Day in the United States and Canada, and banks were closed. 6 and 7 September were Banking Days, and the Tribunal held (in a part of its Award from which there is no appeal) that 8 September was also a Banking Day, even though some banks in the Swiss Canton of Geneva were closed for a local holiday.[12]No payment had reached the escrow account by midnight, in Norway, at the end of 8 September 2022. At 00.09, Norway time, on 9 September, the Sellers served a purported notice of cancellation. Midnight had not then struck in the UK, Canada, or the United States. Midnight arrived in Hawaii well into the working day in Norway on 9 September, by which time the payment into the escrow account had as I understand it been made. So the question arises: were the Buyers in breach of clause 3 at 00.09 Norway time on 9 September because the entire sum due had not been released to the Sellers’ account? If they were, then the Sellers were on the face of it entitled to cancel in accordance with clause 13 of the MOA, which provided:
“Should the Purchase Price not be paid in accordance with Clause 3 (Payment), the Sellers have the right to cancel this Agreement, in which case the Deposit together with interest earned, if any, shall be released to the Sellers. If the Deposit does not cover their loss, the Sellers shall be entitled to claim further compensation for their losses and for all expenses incurred together with interest.”
[13]The sole question before me is whether the obligation to ensure the release of the funds from the escrow account was one that the Buyers were obliged to perform before midnight at the end of 8 September in Norway. There may be various other issues. Those issues were not before the arbitrators for the purposes of the preliminary issue they decided, and are not before me.[14]The MOA provided for LMAA arbitration and English law. A tribunal consisting of three arbitrators was constituted: Mr Simon Gault, Sir Jeremy Cooke and Mr Jonathan Elvey. They considered two preliminary issues. The first, as I have said, concerned the question whether 8 September 2022 was or was not a “Banking Day”. They held it was, and their conclusion is not challenged. The second preliminary issue was whether the time for performance under clause 3 expired(a) at midnight on the end of 8 September 2022 in the most western part of the United States, that is in Hawaii, or(b) at midnight on the end of 8 September 2022 in Norway. The tribunal formulated that issue as follows (disregarding a variation that has now been abandoned by the Sellers):
“… [D]id the Buyers have until midnight on 8 September 2022 in the last of the jurisdictions listed in the definition of ‘Banking Days’ (i.e. until midnight Hawaii-Aleutian Standard Time) before being in breach in failing to pay? Or were the Buyers in breach of clause 3 of the MOA by failing to comply with their payment obligations by … 2400 Norwegian time on 8 September 2022 despite the fact that midnight had not yet passed in the USA, Canada or the United Kingdom?”
On that point, in a Partial Final Award dated 9 December 2024, the arbitrators held, in favour of the Buyers, that the correct interpretation was the first. They set out that conclusion in paragraph BB of the dispositif of their Partial Final Award:
“WE FIND HOLD AND DECLARE that the Buyers had until midnight [on 8 September 2022] in the most westerly of the jurisdictions listed in that definition (i.e. until midnight Hawaii-Aleutian Standard Time) before being in breach in failing to pay the relevant sums pursuant to Clause 3 of the MOA.”

The tribunal’s reasons

[15]The arbitrators described both rival interpretations as “plausible”. They preferred the Buyers’ construction, because they considered that this was the effect of the definition of “Banking Days”. They said this was plain as a matter of the construction of that definition, and that they were “unable to see how the words used, when reading the definition of ‘Banking Days’ into clause 3, could properly be said to give rise to any other meaning”. They accepted that this construction had “little commercial logic”, but thought that “odd results arose from either construction”. They recognised that in reaching this conclusion they were displacing the prima facie presumption (which would be that an obligation that is to be performed in a particular place must be performed by midnight there). But they thought the language did displace that presumption, and detected no sufficient reason to imply any term to displace the displacement. They found nothing useful in the other provisions of the contract in which “Banking Days” were referred to, and nothing in commercial considerations to undermine their construction.[16]For the Buyers, Mr Corby emphasised that this was a conclusion reached by an expert and distinguished tribunal. So it certainly is. It is, however, ultimately and explicitly not based on any established market understanding, or any assessment that commercial logic or “commercial or trade experience” of the practice of ship sales dictated a particular result: see Cottonex Anstalt v Patriot Spinning Mills Ltd [2014] EWHC 236 (Comm), [2014] 1 Lloyd’s Rep 615 at [51]. It is explicitly a conclusion of law based on the arbitrators’ assessment of the meaning of the words that have been used. In those circumstances, I must consider the arguments in support of it and those against it dispassionately on their merits and reach my own conclusion: “in the final analysis the tribunal’s decision is either right or wrong as a matter of law” (Cottonex at [51] per Hamblen J).

The parties’ arguments

[17]There was no dispute before me about the principles of construction: counsel drew my attention to the convenient summary given by Popplewell LJ in Lukoil Asia Pacific Pte Ltd v Ocean Tankers Pte Ltd (The Ocean Neptune) [2018] EWHC 163 (Comm), [2018] 1 WLR 654 at [8]; see to similar effect the summary by Lord Burrows JSC in Providence Building Services Ltd v Hexagon Housing Association Ltd [2026] UKSC 1, [2026] 1 WLR 538 at [21]–[23]. Ultimately, both parties rested their argument on an interpretation of the language and purpose of the definition of “Banking Days”, what that definition as applied in clause 3 would convey to the reasonable business person, and the practical consequences in general terms of the respective possibilities. Both made limited reference to the factual matrix specific to this contract.[18]For the Sellers, Mr Kenny KC submitted that the arbitrators were wrong to treat the definition of “Banking Days” as comprehensive or determinative, or indeed as a definition—as such—of “Days” at all. He submitted that it was syntactically and commercially inappropriate to regard the words “in the country … and in the place of closing … and the United States” as intended to define the word “days”. The definition simply assumed that a day was a day, and provided a test to establish whether any given day was a Banking Day. As such, it had nothing to say about whether 8 September 2022 had, relevantly, ended when it ended in Norway—where the escrow account was to be released—or when it ended in Hawaii. That question was left open, and the gap was appropriately filled by the common law principle that a calendar day ends at midnight, and by the common law presumption that where time matters, it is usually appropriate to look to local time. That left the question of what locality mattered—but in this case at least it was clear that the relevant locality was Norway. This produced, he submitted, a workable and clear answer. The Buyers’ approach, on the contrary, did not make sense as a way of producing clarity, since it produced a “day” which began repeatedly and ended repeatedly, and required an arbitrary selection of one particular end-point.[19]For the Buyers, Mr Corby submitted that the arbitrators had been entirely correct. He emphasised that these were vastly experienced arbitrators, who had reached a clear conclusion about what the language meant. They were right: the definition of “Banking Days” was a definition of “Banking Days”, which as much defined the “Days” part of that expression as the “Banking” part. Each day, so defined, ended when midnight arrived in Hawaii. This, moreover, enhanced the prime requirement of certainty in commercial contracts. It avoided complex and tricky questions about where obligations fell to be performed. To elevate what was at most a presumption or canon of construction about “local time”, originating in the nineteenth century, into more than an indication, ran counter to the clear words of this contract, and was inappropriate.

Discussion: the arbitrators’ core argument

[20]My starting point is that “Banking Days” under the MOA refers to calendar days. This was, at least ostensibly, common ground. It is certainly the correct starting point: see Euronav NV v Repsol Trading Ltd (The Maria) [2021] EWHC 2565 (Comm), [2022] 1 Lloyd’s Rep 247, at [25]. If notice of readiness is given on Monday 1 July, and the banks at every relevant place are open all that week, then payment must be made no later than Thursday 4 July. But it does not matter whether the notice was given at noon or just before midnight, and it does not matter whether the payment is made at 1 am or 11 pm.[21]In my view, the purpose of the “definition” of “Banking Days” is not to define what is meant by “day”, but to identify which calendar days count for the purpose of calculating intervals, and which do not. The definition takes as its starting point—undefined, because it is not thought to call for definition—the idea of a “day”, meaning simply a calendar day, identifiable by date, “the mean solar day, a period of twenty-four hours” (Prowse v McIntyre (1961) 111 CLR 264 per Windeyer J as quoted in The Maria at [34]). It then provides that such a “day” qualifies as a Banking Day if, but only if, it is a day when in each of the identified places, banks are open. The fact that the banks are not simultaneously open at the same time is irrelevant, so long as they are open on that day. One simply asks of, say, 8 September 2022: were the banks open in all the places mentioned? If they were, then 8 September 2022 is a “Banking Day”, and if they are not then it is not. “Banking Days” are not something different from calendar days, but some calendar days.[22]I do not reach that conclusion, principally at least, on the syntactic argument made by Mr Kenny KC (which attributed weight to where the words “on which banks are open” appear in the clause), but because it is plain that this is the purpose served by the definition. It does not answer the question “what is a day?” but the different question “which days are Banking Days”, and in my view does nothing more than that.[23]On the other hand, on the view advocated by Mr Corby in which the definition does not simply identify calendar days that count as “Banking Days”, but in some way defines a special sort of period called a “Banking Day”, the definition, although it starts from “calendar day”, ultimately produces a “day” which lasts for 37 or 38 hours (depending on the time of year), starting (he explained) at midnight in the UAE and ending at midnight in Hawaii. Over the course of such a “day”, the various contributing jurisdictions pass through three different calendar dates, so that a single “day” as so defined began at midnight in the UAE at the start of 8 September 2022 there (which was mid-afternoon on 7 September 2022 in Hawaii), and ended 37 hours later at midnight in Hawaii at the end of 8 September 2022 (by which time it was afternoon on 9 September 2022 in the UAE). I do not think that it would occur to many people to think of such a period as a “day” at all, and it assuredly not a calendar day. Moreover, while this “day” was running its course a second and overlapping “day” will have already begun. The arbitrators were unfazed by these factors. But, while I do not doubt that parties could decide to regulate the passage of time under their contract by overlapping intervals of 37 or 38 hours, and could even if they chose to do so call such a period a “day”, it strikes me as extremely improbable, and inconsistent with the starting point—common to both parties—that the word “day” in the definition of “Banking Days” means calendar day.[24]Understood as I think it should be, then, the definition of “Banking Days” does not purport to tell one what is meant by “day”, nor broaden that word’s ordinary meaning, nor fix its start or its end. It takes all that for granted: it is all work done, without elaboration or qualification, by the word “days”. The “definition” of “Banking Days” does not tell us what a day is, or when it starts or finishes. It simply tells us which calendar days (whenever they start or finish, for like any other calendar day they start and finish at different absolute times in different parts of the world, from Kiribati to American Samoa) count as Banking Days: that some calendar days are Banking Days, and other calendar days are not.[25]Where does one then find an answer to when, in any case, the day ended? The short answer is: in the same place that one would find the answer to when any other calendar day begins and ends. It is a fact of every calendar day that it does not end at the same moment everywhere on earth, but it may be a requirement for commercial certainty that for any particular contractual act one should be able to say whether a day has started or has ended. So one must for that purpose assume both a time and a place. The time, absent some contrary provision, is midnight: Afovos Shipping Co SA v R Pagnan and Filii (The Afovos) [1983] 1 Lloyd’s Rep 335 (HL). And the usual and common sense view is that the place is to be found by considering where the act in question has been or is to be performed.[26]Mr Corby described this as a “19th century legal presumption”. But it seems to me to be no less compelling in the 21st century, and a matter of modern no less than ancient commercial common sense. The charming Victorian decision in Curtis v March (1858) 3 H & N 867, 157 ER 719 was not about days but about hours. The question was whether counsel had arrived late for court by arriving shortly before 10 am by the town clock in Dorchester, a few minutes after Watson B, running on Greenwich Mean Time, had promptly taken his seat on the bench and no less promptly disposed of the case with a directed verdict. Pollock CB observed that “A person hearing that the Court would sit at 10 o’clock would naturally understand that to mean 10 o’clock by the time of the place, unless the contrary was expressed.” That seems to me no less true, as a general observation of common practice, today, where it is conventional (for passengers) to use local time for such things as the departure and arrival times of planes. The continuing vitality of the principle is demonstrated by Henshaw J’s decision in The Maria [2021] EWHC 2565 (Comm), [2022] 1 Lloyd’s Rep 247. Having analysed existing authority on the topic of “time as a local concept” (at [29]–[40]) he said that “the starting point, at least, is that the date of an event should normally be determined using local time at the place where the event happened” [41]). This is a matter not of archaic legal presumption but of practical reality. If a business person says “I will make sure you have the payment in Norway by Tuesday”, they will not normally be understood to mean “by some time on Wednesday morning, which is still Tuesday somewhere else”: they will be understood to mean “by Tuesday in Norway”. If the contract had said “payment must be made in Norway by no later than 8 September 2022”, there would be no real doubt at all that when midnight struck in Oslo the payment was late, even though Honolulu’s banks were still open for business.[27]In saying this, I do not suggest that the presumption amounts to a rule of law, or that it is helpful—if it ever is—to require “clear words” to “rebut” it. That sort of algorithmic approach to interpretation is rarely apt. But it is a natural and a reasonable and (usually) workable starting point, in touch with ordinary ways of thinking about time, and it has that weight that those factors give it.[28]The MOA uses relative not absolute terms. It does not define a fixed calendar date, but one that is to be calculated by reference to an interval. The starting point of that interval is when notice of readiness is given. That may happen at different dates at different points on the globe: the notice given in London at 8.00 am on 2 September would be a notice given at 10 pm on 1 September in Honolulu. I have no doubt that commercial certainty requires that the notice be treated as having a particular date. Mr Corby agreed, I think. He said that the notice should, in the example I have posited, be treated as given on 2 September. I agree. But why? I pressed Mr Corby on why the logic of his view might not allow for the notice of readiness to be treated as having been given on 1 September. I think his answers were either that it should not be so treated because of where it was actually given, or that it should not be so treated because, on his analysis the start of 2 September had already passed (in the UAE). One can, however, readily change the facts to make that a pretty unsatisfactory answer: suppose a notice given in New York at 16.00 on 2 September. Is that to be treated as having been given on 3 September, which has by then started in the UAE? The obvious and sensible answer is that in either of the hypothetical cases I have given, the relevant date is 2 September because the relevant day is the calendar day on which the notice is given at the place where it is given. But if so, why should the practical logic that governs when the period begins not also govern whether a given act has been performed before it ends?[29]The straightforward answer to these problems is quite simple: the local time presumption applies. If a notice is given in London on 2 September, it is given on 2 September: we simply do not ask what date it was anywhere else. One then works forward, ascertaining of each successive calendar day (3 September, 4 September, 5 September …) whether it is a Banking Day. That requires knowing whether banks were open that day (i.e. on that date) in a variety of places, but does not require any attention to when the day began or ended. If banks were open, one adds one to the tally of “Banking Days”. If they were not, one moves to the next day. At the end of the process, one will have identified the latest calendar date on which payment can be made. The definition of “Banking Days” has now done all its work, and has no more to offer. When it comes to deciding whether the payment has been made (or, in this case, released from escrow), one asks whether that happened before the end of the day at the place where it was to happen.[30]If the contract said simply “no later than three days after notice of readiness”, this would be obvious: there would be no reason to alter the natural assumption that the shutter falls at the end of the day at the place where the required act is to be performed, where that is defined, or the last place where it can be performed, if it is not defined. How, really, does the introduction of the concept of the “Banking Day” make any difference? All it does is to make it clear that some days count, and other days do not for the purposes of calculating the interval. It does not comprehensively change what is meant by “day”; it assumes what is meant by “day”, which means calendar day; it marks out certain days for special treatment; and it does so simply for the purpose of calculating an interval which will define the “day” on which the obligation is to be performed.[31]I cannot therefore agree with the arbitrators’ view that the contract’s “definition” of Banking Days tells one what a day is, or when it starts or ends. It does not. It simply tells one which calendar days are to count as Banking Days. The fact that the definition proceeds by reference to the state of commercial affairs in various parts of the world is not relevantly informative. It is just a convenient way of identifying certain days which, as calendar days identified in the usual way by their date, are to be ignored in calculating time. To make sense of the MOA by “reading in” the “definition” of “Banking Days” into clause 3 requires no more than to say that payment should be made by the release of the escrow account (in Norway) not later than 3 days after notice of readiness has been given, disregarding any day on which the banks have not been open in any of the specified jurisdictions. That still means, naturally, on that day in Norway not on that day in Hawaii.[32]I therefore do not accept that the arbitrators were correct in their conclusion that there is a “definition” of “day” which can be “read into” clause 3 so as to displace what they rightly accepted would be the usual meaning and effect of that clause. In my view, they were led to treat that definition as doing more work than it linguistically or (more importantly) practically is intended to do—as providing a special definition of “day” when all it really does is mark certain days (i.e., in practical terms, calendar dates) as relevant or irrelevant for the purpose of calculating an interval of time. Having established that the last day for the release of payment under clause 3 was 8 September 2022, and that the place for the release of that payment was Norway, the proper conclusion as a matter of law was that the obligation to release the payment fell to be performed before midnight in Norway.[33]That conclusion is not inconsistent with the views expressed in M Strong and P Herring, Sale of Ships: The Norwegian Saleform (3rd ed, Sweet & Maxwell 2016). They note that there is “no authority as to which time zones should be taken into consideration for the purposes of determining when ‘midnight’ has passed”. They say that one view would look to the place of payment, but that it is “arguable that New York time may be relevant, at least for US dollar payments”. They prudently advise caution. They do not express the view that the question is resolved by the definition of “Banking Days”. In short, they have the author’s luxury of declining to grasp a nettle that I must grasp. Nor do the published notes to Saleform 2012 say anything specific on this topic, though I accept they indicate—as one would in any event assume—a concern with clarity and certainty.

Other points: the factual background

[34]Thus far, I have been expressing my views on the main point that drove the arbitrators’ conclusion, that is on the meaning and effect of the definition of “Banking Days”, especially it would apply in clause 3. However, it is necessary to cross-check those views by reference to the contract as a whole, to its surrounding circumstances, and to the considerations that both parties put forward as to the commercial consequences of each interpretation.[35]In terms of factual matrix, I can be very brief. It may be true, as Mr Corby submitted, that the place of performance of every obligation under clause 3 might not be clear beyond rational argument. (For instance, the obligation to release the deposit to the Sellers, might entail payment at their order outside Norway.) But even under this modified agreement, I have no doubt that the core obligation (the release of the balance of the purchase price from escrow) was to be performed in Norway, and as Mr Kenny pointed out that is equally clear under the unmodified Saleform 2012, which provides for payment to a particular account. Nor do I think that any significance attaches to the fact that the parties (after the conclusion of the MOA) expressly referred to “Norwegian banking days” in the escrow agreement. That does no more than make it clear that whether banks are open in Norway is the sole criterion to be applied in reckoning the days that will count for one particular purpose.[36]It is of more interest to consider the practical reality—which I think can fairly be taken into account as factual matrix—that for a dollar payment an American bank is likely to be involved. But that seems to me to tell one very little. The three Banking Days allowed for payment may indeed be squeezed a bit if the deadline expires before American banks close on the third day. But, as Henshaw J observed in The Maria, “[t]hat is simply a feature … of a notice period framed in terms of calendar days rather than ‘elapsed time’”. Parties who know that they must pay dollars in country X within three Banking Days of notice of readiness being given in country Y can be expected to make sure they will be in a position to do so, or to stipulate for additional time if that is foreseeably required. (In this case, as it happens, they had thought about this and anticipated that the money would be in Norway in advance. And in no case will notice of readiness come like a bolt from the blue, since the Saleform 2012 envisages estimated dates for delivery to be communicated in advance.)

Other points: the contract as a whole

[37]So far as the scheme of the contract as a whole was concerned, I was taken without much enthusiasm by either counsel, to other provisions of the MOA which use “Banking Days”. The arbitrators did not think these shed any useful light on the issue. Neither do I. None mandates or precludes either interpretation. The most that can be said about them is that some of them may illustrate that it is not always clear beyond argument what “locality” is relevant when considering “local time”, which might suggest that a “non-localised” deadline could sometimes be useful. I address that argument later. As to the specific instances: i) Clause 5 (c) provides that if the Sellers anticipate that the vessel will not be ready for delivery at the cancelling date they may notify the Buyers in writing proposing a new cancelling date. The Buyers then have three Banking Days to decide whether to cancel or to accept the option, and if they do nothing in this time the proposed date becomes the new cancelling date. It seems clear that the starting point for this will be the date the Buyers receive the notification, and the ending point will be the date on which they give their counter-notification. The MOA provides where notices are to be given, and that will be the relevant locality. ii) By clause 6(a)(ii) there are circumstances in which a deduction from the Purchase Price may be made to cover anticipated repairs. The deduction is to be based on the “average of quotes for the repair work obtained from two reputable independent shipyards at or in the vicinity of the port of delivery, one to be obtained by each of the Parties within two (2) Banking Days from the date of the imposition of the [class] condition/recommendation, unless the Parties agree otherwise”. Again, I see no great difficulty: the relevant locality will be either (as seems to me most likely, given the language) the port of delivery—where the quotes are to be “obtained”—or, as Mr Kenny suggested, the places at which the parties notify each other, as they implicitly need to, of those quotes. iii) Clause 8 provides for the delivery of documentation. By the parties’ agreement (modifying the Saleform 2012 to reflect a “virtual” closing), hard copies of that documentation are to be sent to each other “within five (5) Banking Days following delivery of the Vessel”. I can see that there may be room for argument about whether this clause is complied with by sending the documents within five days, if they are received later, or only by sending them so that they are received within five days. But that uncertainty is not resolved on either party’s interpretation of Banking Days, and whatever the right answer to that is will also dictate the relevant locality to determine whether the deadline has been met. iv) Clause 18 was a clause added to address Covid-19. It addresses what might happen if crew disembarkation becomes impossible, and does so using a mechanism that sometimes refers to things happening within “Banking Days” and sometimes to them happening within “days”. Neither party could make much of this, and neither party pointed to any particular difficulty with operating it on either interpretation. v) For completeness I would add that the term is also used in clause 2, by reference to the payment of the Deposit. Neither party addressed me on that clause. In that case, as in clause 3, it is clear where the obligation is to be performed, since it is to be by payment to the account of the escrow agent.

Other points: commercial logic and consequences

[38]I turn, finally, to the effect of the rival constructions in commercial practice. For his part, Mr Corby stressed the importance of certainty in ship sales. I agree that is always an important consideration in commercial contracts, though it should not be fetishized because it is never the only thing that matters. Commercial enterprises look for reasonable certainty, but they do so with a degree of practical common sense and in the knowledge that an ability to respond reasonably to varying and unpredictable circumstances may also be important. They also, I think, rightly assume that contracts will be interpreted in a way that generally gives an ordinary meaning to ordinary words such as “day”, perhaps especially when the ordinary meaning is one that embodies, and makes manageable, a good deal of underlying complexity.[39]It is trivially true that since a given calendar day ends at different times in different places, then in a contract with any significant international dimension, there may be room for debate if the precise moment is not in some way specified. It is also trivially true that if the relevant time is “local time” then it may be necessary to determine the relevant locality, and that where that is not specified or obvious there may be room for debate about it. Such considerations sometimes lead parties to be precise about the time-zone they have in mind, or to be precise about the locality they have in mind. But the difficulties should not be exaggerated: most of the time the philosophical complexities of “days” and legal questions about place of performance pass unnoticed, because they have obvious everyday answers. And, pertinently here, there is no room for doubt about them so far as the core obligations under clause 3 of the MOA are concerned: the release of money from a Norwegian escrow account to the Sellers’ Norwegian account was something that was going to happen in Norway.[40]I do not, therefore, think that it would be right to approach the MOA with a strong predisposition to think that the parties must have been looking for a way to draw a bright line across every possible grey area, so far as time is concerned. But, more importantly, if they had wanted to do that, the definition of “Banking Days” would be a strange way to achieve that end. Overlapping “days” of 37 or 38 hours, for large parts of which the date is different in different parts of the world, is not likely to produce certainty in general. Consider the case I posited in paragraph 28 above. Because one (obviously) needs a single starting date, the Buyers are constrained either to say that the starting date does depend on the local date at the place the relevant act occurs, whatever the dates elsewhere, or that it is always the date at the most eastern relevant jurisdiction, in this case the UAE. If the former is right, then the uncertainty has only been half addressed: determining “local time” still matters. If the latter is right, then it becomes in every case necessary to consider the precise time at which an event occurs where it occurs, in order to work out the date in the UAE—just the sort of “fractional date” analysis that the common law generally avoids: see The Afovos [1983] 1 Lloyd’s Rep 335 (HL) at 340.[41]The Buyers seek to resolve such objections by positing a rule that if one is looking forwards (to the end of a day) one must consider the most western place (Hawaii) and if one is looking backward (to the start of a day, as in various cases the contract does when it considers whether documents are dated no more than a certain number of Banking Days prior to a particular date) one looks to the most eastern one (UAE). The starting point, in either case, they leave to be determined by the facts. As a device for achieving certainty in a commercial contract, I find such a scheme unappealing.[42]I think Mr Kenny is also right to say that it is arbitrary. Mr Corby justifies it, for clause 3, by saying that a payment cannot be regarded as having been made “later” than three Banking Days if it is made while it is still a Banking Day in one of the jurisdictions. But why? Might it not equally be said that it was later than three Banking Days, since it is made at a time when it is later than three Banking Days in one of the identified jurisdictions? A definitive answer to the question would, in practice, need to be given by specifying one time zone which is to be determinative for all purposes, not by specifying twelve.[43]If I was satisfied—as the arbitrators were—that the language of the MOA plainly dictated that Hawaii-Aleutian time determined the moment at which a date passed under the contract, I would also agree with them that the absence of commercial logic to such an approach would not be an insuperable objection. Provisions which are designed to achieve rigid certainty may sometimes pay a price in commercial logic. But I do not accept the submission (which was not, as I read the award, the basis for the arbitrators’ decision) that considerations of commercial certainty militate in favour of the construction they adopted. Rather the reverse: although I think that neither interpretation guarantees crystal clarity in every case, the Sellers’ interpretation seems to me to involve markedly less mental gymnastics, and to cleave more satisfactorily to ordinary conceptions of how time works.

Conclusion

[44]For these reasons, despite the experience and excellence of the arbitrators, and Mr Corby’s powerful defence of their reasoning, I am satisfied that on this occasion their conclusion was legally incorrect. On its true construction, in the events that happened, the MOA required the balance of the purchase price to be released from escrow in Norway on 8 September 2022, and the Buyers were in default if that had not occurred by midnight Norway time at the end of that day. My preliminary view is that the appropriate order will be to vary paragraph BB of the award by substituting an appropriate declaration; but I shall hear counsel on the form of my order.