[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Commercial Court) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Commercial Court) Decisions >> Glencore International AG v Alpina Insurance Company Ltd. & Ors [2003] EWHC 2792 (Comm) (20 November 2003) URL: http://www.bailii.org/ew/cases/EWHC/Comm/2003/2792.html Cite as: [2003] EWHC 2792 (Comm), [2004] 1 LLR 111, [2004] 1 Lloyd's Rep 111, [2004] 1 All ER (Comm) 766

[New search] [Printable RTF version] [Help]

IN THE HIGH COURT OF JUSTICE
QUEENS BENCH DIVISION
COMMERCIAL COURT

		Royal Courts of Justice Strand, London, WC2A 2LL
		20th November 2003

see also: [2004] EWHC 66 (Comm)

Mr Justice Moore-Bick:

I. Introduction

1. The Metro litigation

1. This is what has come to be known as Phase 5 of the Metro litigation arising out of the collapse of Metro Trading International Inc. ("MTI") and Metro Oil Corporation ("MOC"), collectively known as the Metro Group, in February 1998. For some years prior to that MTI had operated a floating oil storage facility and bunker supply business using a number of vessels anchored in the waters off Fujairah. The centre of these operations was the Metrotank, a ULCC that had been specially modified for the purposes of fuel oil blending operations. MOC operated a refinery and oil storage facilities at Fujairah for the processing of crude oil and the handling of the resulting products.

2. A number of different oil companies entered into contracts with MTI for the storage of various grades of oil at Fujairah and in some cases they also sold oil held in store to MTI for use in its bunker supply business. Of these by far the largest depositor in terms of the quantities of oil delivered into storage was Glencore International A.G. ("Glencore"). Glencore and other companies also entered into agreements with MOC for the processing of crude oil under which they were entitled to receive refined products in return for the feedstocks they supplied. MTI supplied bunkers to ships calling at Fujairah and from time to time also sold cargoes of bunker fuel, mainly for export to the Far East. In order to enable it to finance its business MTI obtained support in the form of credit lines from a number of banks, principally Banque Trad Crédit Lyonnais.

3. When the Metro Group collapsed Glencore and the other depositors discovered that the quantity of oil held in the floating storage facility was far smaller than ought to have been the case. The Metro Group was hopelessly insolvent and accordingly many of those involved, including the banks who had lent money to MTI to enable it to buy oil, asserted proprietary claims against the oil remaining in its possession and also asserted rights to recover amounts due to MTI in respect of the outstanding price of bunkers supplied to third parties. Other claims were made against those who had received shipments of bunker fuel on the grounds that they had wrongfully interfered with goods that did not belong to them. Glencore also made a claim against its insurers in respect of the loss of oil that had been delivered to MTI for storage but was no longer in its possession and could not be accounted for and a similar claim in respect of refined products that MOC had produced under the processing agreement between them but were likewise no longer in its possession and could not be accounted for.

4. At the outset the Metro Litigation comprised 35 separate actions involving over 50 different parties. In order to enable them to be tried in an orderly and efficient manner, the court directed that they should be managed together and that the issues to which they gave rise should be determined in groups in relation to all the actions. Accordingly, in Phase 1 I decided a number of issues of principle relating to the law applicable to the passing of title to oil held in storage in Fujairah. In summary, I held that under English conflicts of laws rules the transfer of title to movable property is governed by the law of the place where that property is situated, that being in the present case the law of Fujairah. My judgment is now reported at [2001] 1 Lloyd's Rep. 283.

5. The nature of the contractual relationships between each of the various oil company claimants and MTI was to be determined as Phase 2, but in the event the trial was confined to the relationship between Glencore and MTI. In my judgment at Phase 2 I considered the contractual arrangements between MTI and Glencore and the effect under the law of Fujairah of various dealings with the oil held in storage in the context of those contractual arrangements. My judgment is not reported, but can be obtained on-line from the British and Irish Legal Information Institute ("BAILII").

6. Phase 3 was to involve the determination of the proprietary claims of the banks that had provided financial support to MTI to enable it to buy the oil it needed for its business, but in the event those claims were resolved without the need for a trial.

7. It was envisaged at an early stage that the competing claims to the oil remaining in store based on the decisions reached in Phases 1 to 3 and a detailed analysis of the movements of oil into and out of the floating storage facility at Fujairah would form Phase 4 of the litigation, leaving Glencore's claim against its insurers to be determined as Phase 5. However, tracing the movements of oil into and out of the facility has proved to be a complex and difficult task and it has therefore been postponed to a later date. As a result I am now concerned with Phase 5 of the litigation, namely, Glencore's claim against its insurers, Alpina Insurance Co. Ltd ("Alpina") and certain other insurers in the Swiss market, in respect of oil misappropriated by MTI from the floating storage facility at Fujairah and refined products derived from feedstock delivered under its processing agreement with MOC which are said to have been misappropriated by MOC.

   2. Background to the dispute

8. The circumstances in which Glencore delivered oil to MTI at Fujairah and the manner in which the relationship between them developed between 1992 and 1998 are described in my judgment at Phase 2 and do not need to be repeated here. However, it is necessary to say a little more about the nature of Glencore's business operations since they are relevant to the issues that arise in the present case.

9. Glencore is a Swiss company originally known as Marc Rich & Co. A.G. It had a number of subsidiaries including an English company, Marc Rich & Co. Ltd. Following a management buy-out in 1993 the names of the companies were changed to Glencore International A.G. and Glencore UK Ltd. Mr. Andrew Gibson, who until February 2000 was head of the insurance department, described Glencore in his evidence as a diversified natural resources company involved in the production and trading of metals and minerals, energy products and agricultural products. It is well-known as a trading company dealing in, among other things, crude oil and oil products, grain, sugar, rice and other staple commodities, but its activities are not confined to commodity trading and extend, for example, to the operation of industrial plants.

10. For many years Glencore has made use of open covers in connection with its commodity trading activities for the purposes of insuring goods in transit. This form of policy, under which goods in transit falling within the terms of the cover are automatically insured as soon as the policyholder acquires an interest in them, provides the flexibility and continuity of cover which are essential to enable a large trading organisation to carry on business in the modern world. Commerce could not function efficiently if traders handling any significant volume of business had to negotiate separate terms for each individual shipment. Open covers also provide benefits for underwriters because they enable large numbers of similar risks to be underwritten without the need for individual presentations.

11. From the early 1980's Glencore's open cover in respect of oil and oil products was placed in the London market through the brokers Lloyd Thompson (now Jardine Lloyd Thompson). Over the years the policy wording developed in response to changes in the nature of Glencore's business but the underlying philosophy throughout was to procure a broad and flexible contract providing cover against all risks of loss and damage to oil in which Glencore acquired an interest wherever it was situated. As the insurers were aware, Glencore's involvement in the purchase and sale of crude oil and products was solely as a trader; it did not participate either in the production of crude oil or in the final distribution of products.

12. As might be expected in these circumstances, the contract was worded in very broad terms. It took the form of a marine open cover in favour of Glencore and its associated and subsidiary companies and other interested parties for whom they might receive orders to insure. It covered transit by land, water, air and pipeline from any ports or places in the world to any other ports or places in the world. It covered crude oil, kerosene, gasoil, naphtha, liquid petroleum gases and all products incidental to Glencore's business or otherwise as might be declared. It provided cover on four different bases at the insured's option, that which was most commonly used being based on the American Institute Cargo Clauses. To these basic conditions was added a large number of general conditions dealing with various aspects of the business insured. It will be necessary to refer to some of those in detail at a later stage. Subject to the insured's right to choose the basis of cover, the contract was obligatory on both sides as regards transit risks. There was also a separate section providing cover, if required by the insured, against political risks (confiscation etc., usually known as 'CNED' risks) at a rate to be agreed by underwriters. The policy ran for 12 months from 1^st April each year. In the year 1994-95 approximately 79.75 million tons of crude oil and products with a total value of US$4.9 billion were declared to the open cover generating a gross premium of about US$5.5 million.

13. It is necessary at this point to mention the following two conditions dealing with transit and storage risks respectively which formed part of the cover provided by the London market for the 1993-94 year and which, having been retained as part of the cover in subsequent years, have assumed particular importance in this case:

    "1.15 Cover to attach from the time the Assured becomes at risk or assumes interest and continues in transit and/or store (other than as below) or wherever located and until finally delivered to final destination as required . . . . . . .

    1.18 Including, if required, storage and blending prior to shipment or after final discharge in Land Tanks, Refineries, and storage in barge(s), irrespective of whether transit covered hereunder subject to Policy terms and conditions but excluding loss and/or damage caused by faulty blending on amount at risk at particular location involved at Additional premium of 0.015% each 30 days or pro-rata in excess of first 30 days if transit covered hereunder, otherwise, 0.015% each 15 days or part thereof."

14. In December 1993 Glencore entered into arrangements with MTI for the purchase and storage at Fujairah of four or five parcels of Iranian fuel oil as described in paragraphs 27-28 of the Phase 2 judgment. That was part of a joint venture between Glencore and MTI designed to take advantage of contango conditions in the fuel oil market. The vessel Mount Athos was chartered by MTI specifically for the purpose of holding these cargoes, but although the arrangements were made in December 1993, it was not until March 1994 that Glencore approached Lloyd Thompson for confirmation that the oil held on board the Mount Athos was insured under the open cover. In response to that enquiry Lloyd Thompson prepared an indorsement in the following terms which was shown to the leading underwriter, Mr. Hill, on 28^th March 1994:

    "Assured advise that they will be utilising vessel "Mount Athos" for floating storage. Vessel will be anchored off Fujairah.

    Agreed to provide coverage in accordance with Conditions 1.18 but rate 0.015% on annual throughput (estimated approx. USD 95,000,000).

    All other terms and conditions remain unchanged.

    Dated in London: 25^th March, 1994

    INFORMATION (N.L.O.W.)

    Assured anticipate annual throughput approx. 1,500,000 tonnes and average period in store not exceeding 15 days."

    Mr. Hill deleted the proposed rate of 0.015% and scratched the indorsement "Rate to be agreed." One of the other leading underwriters agreed to hold Glencore covered subject to receiving information about the vessel's mooring arrangements.

15. In the event MTI stopped using the Mount Athos to store oil before any rate had been agreed and the vessel left Fujairah on 16^th May 1994. On 7^th June underwriters scratched another indorsement by which they noted that Glencore would be using the Metrotank as floating storage for oil to the value of approximately US$80 million per annum and agreed to provide cover in accordance with condition 1.19 (previously 1.18) at a flat annual premium of US$20,000. The information provided to the underwriters at the time and noted on the indorsement indicated that Glencore expected a total volume of 1 million tonnes of oil to be put into storage at Fujairah with a throughput of approximately 85-90,000 metric tons a month. No further reference to floating storage at Fujairah appears to have been made until the end of February 1995 when Mr. Warren, Mr. Gibson's deputy, advised Lloyd Thompson that the Mount Athos was again being used to store a quantity of fuel oil and asked that she be added as a named vessel to the addendum covering the Metrotank. Underwriters scratched an indorsement dated 1^st March giving effect to that request.

    II. The Alpina policy

    1. The policy terms

16. Alpina is a member of the Zurich Insurance group of companies; the head of its marine insurance division is Mr. Hugo Panzeri. In 1994 Alpina had set out to recruit underwriters who spoke English fluently and had experience of the London market. As a result two new members of the underwriting team were appointed, Mr. Wayne Krause and Mr. Alan Edwards, both of whom had spent time working in London. Mr. Krause joined Alpina in December 1994 and Mr. Edwards in January 1995.

17. Shortly after he moved to Alpina Mr. Edwards received an approach from the broker at Lloyd Thompson responsible for Glencore's account, Mr. Eddie Vanderson, who enquired whether Alpina would be interested in writing the Glencore oil open cover. A similar approach had been made to Alpina in the previous year through the Dutch brokers who handled many of Glencore's other insurances, but it had not been pursued. By early 1995 various factors had led Mr. Vanderson to think that it would be difficult to obtain the most favourable rates in the London market for cargoes carried on older vessels and a chance contact with a colleague at Lloyd Thompson who had done business with Mr. Edwards on a previous occasion led him to renew the enquiry.

18. Following an initial telephone conversation, therefore, Mr. Vanderson wrote to Mr. Edwards on 18^th January 1995 enclosing various documents including the expiring policy wording and loss statistics. Mr. Edwards replied by fax on 26^th January indicating the general terms on which Alpina would be willing to consider writing the cover and drawing attention to the fact that Alpina would wish to delete certain clauses from the existing terms. The terms to which he objected were identified in a fax sent on 30^th January and were the subject of a lengthy discussion between Lloyd Thompson and Alpina at a meeting at Alpina's offices in Zurich on 20^th February. The outcome of that meeting was sufficiently encouraging for Lloyd Thompson to pursue negotiations with Alpina to the exclusion of the London market. A second meeting took place in London on 14^th March 1995 and a week later Mr. Panzeri felt sufficiently confident of obtaining the business to contact other insurers in the Swiss market whom he thought might be interested in taking part of the risk. The negotiations between Lloyd Thompson and Alpina were ultimately successful and the Alpina open cover came into force with effect from 1^st April 1995.

19. The policy was written on the same terms as the expiring Lloyd's and London market cover, including clauses 1.15 and 1.18 to which I have already referred. It made no express provision, however, for oil held in the floating storage facility at Fujairah. On 3^rd April Mr. Warren sent a fax to Mr. Vanderson informing him that the Metrotank and the Mount Athos were again being used to store fuel oil. Glencore calculated that there would be a combined throughput of approximately 1.75 million metric tons a year with an insured value of US$125 per ton. The same day Mr. Vanderson sent a fax to Mr. Edwards at Alpina in the following terms:

    "Last year we obtained Lloyd's underwriters' agreement to provide floating storage coverage in respect of fuel oil stored on board vessels "Metrotank" and Mount Athos".

    Both vessels anchored off Fujairah and estimated annual throughput was 1 million tonnes with approx. annual insured values in region USD 80 million.

    We have now been advised by Glencore that for the next 12 month period they calculate a combined throughput of approximately 1.75 million metric tonnes per annum with insured values at risk of USD 218,750,000.

    Last year Lloyd's underwriters provided coverage subject to flat annual premium of USD 20,000 (copies of agreements attached hereto) in accordance with cover conditions for storage in land tanks etc – there have been no losses reported to date in respect of this exposure.

    This is purely a storage risk exposure and there are no blending risks involved.

    Glencore are not normally responsible for the insurance on fuel oil being delivered to these floating storage vessels but usually provide coverage for onward transit following storage risk.

    Based on increased activity for next 12 month period we would suggest flat annual premium say USD 50,000 and we would appreciate receiving your confirmation to this effect.

    Thanks your assistance."

    Attached to the fax were copies of the two endorsements to the London market open cover dated 2^nd June 1994 and 1^st March 1995.

20. By an indorsement dated 7^th April 1995 (subsequently reissued in identical terms on 18^th October 1995) Alpina agreed to extend cover to oil on board the Metrotank and Mount Athos in accordance with condition 1.18 at an 'in full' premium of US$50,000 and the matter rested there until 14^th September 1995 when Alpina signed a further indorsement noting that Glencore occasionally made use of one or two storage tanks at Fujairah terminal when the storage vessels were nearing capacity or for other operational reasons.

21. Two further endorsements dealing with floating storage at Fujairah followed in January 1996. In the first Alpina noted that, where it was not possible for logistical reasons for Glencore to use land tanks for additional storage, cargoes might remain on board the carrying vessel pending delivery into dedicated floating storage and agreed that cover for that period would be deemed included in the marine rate in accordance with clause 1.18. This particular indorsement clearly resulted from a misunderstanding by Mr. Divine of Lloyd Thompson of instructions he had received from Mr. Warren because it was followed a few days later by a further indorsement dated 11^th January in the following terms:

    "Further to endorsement dated 5^th January 1996 and with reference to endorsement dated 18^th October 1995 Assured advise for clarification purposes that where it is not possible for interest to remain on board the delivering vessel prior to ultimate delivery it may be discharged onto either MT "EVOIKOS", "FAY" or "SHERVAN" for incidental (i.e. usually no more than one week) storage (where Assured may acquire interest) prior to final delivery to the dedicated floating storage vessels;

    which noted and agreed that the latter three vessels be included under the coverage afforded by endorsement dated 18^th October 1995 in their capacity as incidental floating storage vessels."

22. When the policy was renewed for the year beginning 1^st April 1996 the endorsements relating to oil in store at Fujairah were replaced by a new clause, condition 1.29, which had been drafted by Mr. Divine. Although described in correspondence as merely a consolidation of the earlier endorsements, condition 1.29 was in fact a new provision which, while following quite closely the language of the endorsements, in some respects differed from them and extended the scope of the cover. Condition 1.29 read as follows:

    "Underwriters note and agree to provide coverage hereunder in accordance with Conditions 1.18 in respect of cover interest whilst on board floating storage vessels "METROTANK" and "MOUNT ATHOS" whilst anchored off Fujairah (including incidental storage on-shore Fujairah and/or on board final supplying vessels (such as "EVOIKOS", "FAY" and/or "SHERVAN") pending delivery into dedicated storage vessels as required), subject to an annual premium of USD50,000."

23. Mr. Vanderson's fax of 3^rd April 1995 and the endorsements of 7^th April and 18^th October had all referred to an estimate of throughput for the current year. However, when it came to renewing the cover for the year beginning 1^st April 1996 Glencore failed to provide any estimate of throughput for the coming year and did not provide any information about the quantity of oil that had passed through the facility during the expiring year or the year before that.

24. The policy for the 1996-97 year was extended by agreement to 30^th June 1997. Negotiations for the next renewal therefore took place rather later than in previous years, culminating in a meeting at Alpina's offices in Zurich on 30^th May 1997. There is a fierce dispute as to what, if anything, was said in the course of that meeting about Glencore's estimate of throughput at Fujairah for the coming year, but it is common ground that no information was given about the actual throughput in 1996-97 or any previous year. The open cover was renewed on substantially the same terms, including condition 1.29.

    2. The losses

25. Although there remains a dispute as to the precise extent of Glencore's losses, it is clear that during 1997 and the early weeks of 1998 MTI misappropriated large quantities of oil that it was holding in store for Glencore. It is also clear that during the same period some cargoes carried under bills of lading made out to the order of Glencore were diverted by MTI to other destinations after the carrying vessel had reached Fujairah, sometimes without any part of the cargo being discharged. The precise nature and extent of the losses will be determined at the postponed Phase 4 of this litigation. For present purposes I proceed on the assumption that MTI and MOC misappropriated oil belonging to Glencore in some or all of the ways discussed later in this judgment over a period of many months, mainly by drawing without authority on stocks held to the order of Glencore in order to produce blended bunker fuel for supply to third parties or to meet contracts for the sale of refined products. Glencore contends that MTI and MOC together have misappropriated various quantities and grades of oil with a total value in excess of US$250 million and seeks to recover that loss from Alpina under the open cover.

    3. The issues

26. Alpina declined liability for the losses claimed by Glencore on a number of different grounds which can be conveniently grouped under the headings 'fraud', 'misrepresentation and non-disclosure' and 'coverage issues'.

    (a) Fraud

27. Glencore first notified Alpina of a claim under the open cover on 9^th February 1998, two days after the meeting in Athens with Mr. Kilakos described in my judgment at Phase 2 at which the losses were discussed. In April 1998, a little over two months later, Glencore submitted its formal claim under the open cover. The claim proceeded on the footing that the quantities and grades of oil that had been lost were represented by the difference between the quantities that Glencore had discharged into storage and the quantities that it had sold to MTI or withdrawn for its own purposes.

28. From an early stage in their relationship Glencore sold parcels of oil held in storage at Fujairah to MTI by what became known as 'ITT contracts'. Such contracts provided for payment by letter of credit against the issue of a stock transfer certificate evidencing the passing of title to the oil in question. However, in my judgment at Phase 2 I held that the course of dealing between Glencore and MTI was such that oil sold to MTI under an ITT contract was released to MTI on the 'ITT date', that is, the deemed stock transfer date for payment purposes, which was some time ahead of the date on which Glencore actually received security for the payment of the price in the form of a letter of credit. Furthermore, in the course of the litigation, particularly during the Phase 2 trial, it became apparent that a few parcels of bunkers and one parcel of naphtha had left Fujairah after the collapse of the Metro Group during a period in which Glencore was controlling MTI's operations at Fujairah. They included a parcel of fuel oil shipped on the Horizon XII which was the subject of a dispute between Glencore and Fal Oil at Phase 2 trial and which is also the subject of a dispute between Glencore and Alpina in this Phase.

29. Alpina took the view that those at Glencore who had been responsible for formulating the claim must have been aware of the way in which Glencore had done business with MTI and could not honestly have believed that oil released to it pursuant to ITT contracts had been wrongfully taken. It also took the view that they must have known that the parcels of fuel oil and naphtha that were shipped out during the period in which Glencore was controlling the Metro Group's activities in February 1998 had been disposed of with its authority and had not been wrongfully taken. Accordingly, in July 2002 Alpina amended its particulars of defence to allege that Glencore had dishonestly exaggerated its losses or had dishonestly sought to enhance the prospects of obtaining a recovery under the policy.

    (b) Misrepresentation and non-disclosure

30. Alpina has purported to avoid the policies for both the 1996-97 and 1997-98 policy years on the grounds that Glencore failed to disclose, or misrepresented, the throughput at Fujairah, failed to disclose various aspects of its relationship with MTI, failed to disclose that MTI was conducting blending operations at Fujairah and indeed made a positive misrepresentation that no blending of its oil was taking place there.

    (c) Coverage issues

31. Finally, Alpina has raised a number of issues on the construction of the policy which, if correct, would exclude or limit the extent of any liability to Glencore.

    III. Fraud

32. At the close of Glencore's case, during which all those who were directly concerned in the formulation of the claim under the open cover had given evidence, Mr. Kendrick very properly informed the court that Alpina did not wish to pursue the allegations of fraud made against Glencore and abandoned that part of its defence. It is unnecessary, therefore, to say any more about this part of the case other than to make it clear for the benefit of those against whom allegations of dishonesty were made that they have been wholly withdrawn.

    IV. Misrepresentation and non-disclosure

    1. The duty of disclosure

33. Before turning to the specific allegations of misrepresentation and non-disclosure made by Alpina in this case I think it is helpful to consider how the duty of utmost good faith in relation to disclosure operates in the context of an open cover of this kind. The insured is bound to disclose all matters known to him that are material to the risk the insurer is being asked to accept and it follows, therefore, that the scope of his duty of disclosure is to be determined by reference to the subject matter of the proposed insurance. The primary risk that Alpina was being asked to assume was that of loss of or damage to goods in transit, but, as is common in insurances of this kind, the policy also extended to goods in store, both prior to shipment and after discharge at their final destination. Moreover, since in the present case the policy also extended to goods in store which had not been, and were not intended to be, the subject of an insured transit, the range of risks that the insurers were being asked to assume included the risk of loss of and damage to goods in store generally. The proposed contract also provided cover against political risks, but since these had to be individually presented to the insurers for agreement on the rate of premium, different considerations arise in relation to that element of the cover.

34. Two aspects of the risk that was presented to Alpina are particularly significant for present purposes. The first is the extremely broad scope of the proposed cover which extends to transit by land, water, air and pipeline from any place in the world to any place in the world. Similarly, cover against risks of loss of or damage to goods in store is worldwide with no restrictions as to location. It is difficult to imagine a policy on goods being written in wider terms than these. The second is the nature of the insured's business. Glencore is, and was known by Alpina to be, one of the largest independent traders in crude oil and products with interests all over the world. A successful commodity trader must take advantage of market opportunities as and when they arise; he must also be prepared to be innovative in the ways he does business. An insurer who is asked by a commodity trader to provide an open cover on goods must in my view be taken to know that. As Lord Mansfield said in Carter v Boehm (1766) 3 Burr. 1906 at page 1911

    "every underwriter is presumed to be acquainted with the practice of the trade he insures; if he does not know, he ought to inform himself."

    A similar statement of principle is to be found in Noble v Kennoway (1780) 2 Doug. K.B. 510, 513.

35. Although these principles were not controversial in themselves, there was a measure of disagreement between the parties about their application to open covers. Mr. Kendrick submitted that even though the insurer may have agreed to accept a very wide range of risks which on the face of it would allow the insured to conduct his business in a variety of ways, some less hazardous, some more so, the insured was under a duty to disclose any specific plans he might have to conduct his business in one of the more hazardous ways. Mr. Sumption submitted, on the other hand, that an insurer who has agreed to provide cover to a trader in terms as broad as those of the present policy must be taken to be aware of the whole range of different ways in which such a person may ordinarily be expected to conduct his business. The insured need not, therefore, disclose the fact that he plans to trade in a way that will expose the insurer to greater risk, provided that what he intends to do does not fall outside the range of possibilities that the insurer ought to have in mind.

36. A number of authorities were cited as illustrating and supporting these propositions. In Harrower v Hutchinson (1870) L.R. 5 Q.B. 584 the plaintiffs insured a cargo of bone and bone ash for a voyage from a port or ports in the province of Buenos Aires to the United Kingdom. They planned to load part of the cargo in Buenos Aires itself and complete at a port along the coast known as Laguna de los Padres which was unsafe in certain conditions. The plaintiffs did not inform the underwriters that the vessel was to call at Laguna de los Padres and if the underwriters had known of that fact they would have charged a higher premium. The vessel and part cargo were lost on the way back from Laguna de los Padres to Buenos Aires and the insured made a claim under the policy, but the court held that the underwriters were entitled to avoid the policy for non-disclosure.

37. Mr. Kendrick submitted that this case supports the conclusion that even where the insurer accepts a range of risks – in this case the risk of loading at any ports in the province of Buenos Aires – the insured must still disclose the fact, if it be the case, that he intends to act in a manner that will expose the insurer to a risk at the more hazardous end of the range that he is being asked to undertake. However, in my view that is not what it decides. The case was a strong one because the court took the view that the insured had deliberately concealed from the underwriters the fact that the vessel was intending to call at Laguna de los Padres knowing that a higher premium would be charged if they disclosed that fact. But apart from that it is clear from the judgments that neither the underwriters concerned nor underwriters in general were aware at the time that cargoes were being loaded at Laguna de los Padres so that the possibility that the vessel might load there was not a matter that an underwriter familiar with the trade could be presumed to know. The risk of calling at the port was therefore outside the range of possibilities that a prudent underwriter could be expected to have in mind.

38. In Cheshire v Thompson (1918) 29 Com. Cas. 114, another case relied on by Mr. Kendrick, the plaintiffs were warehousemen in Liverpool and Newport who insured with the defendant their profits for the handling of four cargoes of nitrate of soda carried from South America to United Kingdom. They wished to obtain cover against marine perils, war risks and the risk that the cargoes might be diverted to other ports by reason of a change of mind on the part of the owners. In the event cover was placed through brokers on terms that the underwriters would pay a total loss if the vessel did not reach the destination named in the policy "through any cause that may arise", but the risk of diversion was not mentioned as such when the cover was placed. In the event one of the insured cargoes was diverted by orders of the Admiralty. The plaintiffs claimed under the policy and the underwriter declined to pay on the grounds that the risk of diversion had not been accepted. Bailhache J. held that the risk of diversion was a particular and unusual risk not comprehended within the ordinary marine and war risks which was not covered by general wording of the kind used in that case unless it had been made clear to the underwriter what was intended. I agree with Mr. Sumption that the decision turned on the construction of the policy rather than the scope of the duty of disclosure, though it may be said to provide a further illustration of the fact that the underwriter will be presumed to have in mind only such matters as would be within the contemplation of one who is familiar with the trade in question.

39. Property Insurance Co. Ltd v National Protector Insurance Co. Ltd (1913) 108 L.T. 104 concerned a contract of reinsurance of a vessel on the same terms as original. The original policy included a liberty to the insured to navigate in the Great Lakes during the season at an additional premium, but that fact was not disclosed to the reinsurers. Scrutton J. found that a provision of that kind was so unusual that in the ordinary way it ought to be disclosed, although in that case there were other provisions of the contract that dispensed with notification. I think Mr. Sumption was right in saying that the approach adopted by Scrutton J. supports the conclusion that where the insurer agrees to accept a range of risks described in general terms, only matters that fall outside the range of what can reasonably be contemplated by a person familiar with the trade must be drawn to his attention.

40. In Mann MacNeal and Steeves Ltd v Capital and Counties Insurance Co. Ltd [1921] 2 K.B. 300 claims were made on two policies of insurance on a wooden-hulled schooner with auxiliary engines on a voyage from the United States to France and back. On the outward voyage the cargo included a parcel of gasoline in drums. The vessel was destroyed by fire on her return voyage and the underwriters sought to avoid the policies on the grounds that the insured had failed to disclose that it had entered into the contract for the carriage of gasoline. It was quite common at the time, however, for gasoline in drums to form part of a general cargo on such a voyage. Greer J. held that the contract for the carriage of gasoline was material and despite some obvious doubts the Court of Appeal did not disturb his finding. However, all three members of the court were of the view that disclosure had been waived by virtue of the fact that underwriters were presumed to know the kinds of cargo ordinarily carried on such voyages. One of the reasons given for his decision by Greer J. was that although the insurer may be willing to take the risk of a hazardous cargo, he should be told if the risk is in fact a certainty. This argument, which bears a close similarity to the way in which Mr. Kendrick put this part of his case, was rejected by Atkin L.J. in the following terms at page 312:

    "Greer J. says with force that while the insurer may be prepared to risk the chance of a hazardous cargo, he must not be taken to be prepared to incur the certainty of a hazardous cargo. I feel the weight of this, but I think the answer is that included in the risk he takes is the risk that there is an already concluded engagement for hazardous cargo, just as there is the countervailing possibility that he runs no risk of a hazardous cargo at all, by reason of an absolutely safe cargo having been agreed."

41. In the light of these authorities I think Mr. Sumption is right in saying that when an insurer is asked to write an open cover in favour of a commodity trader he must be taken to be aware of the whole range of circumstances that may arise in the course of carrying on a business of that kind. In the context of worldwide trading the range of circumstances likely to be encountered is inevitably very wide. That does not mean that the insured is under no duty of disclosure, of course, but it does mean that the range of circumstances that the prudent underwriter can be presumed to have in mind is very broad and that the insured's duty of disclosure, which extends only to matters which are unusual in the sense that they fall outside the contemplation of the reasonable underwriter familiar with the business of oil trading, is correspondingly limited. It also means that the insured is not bound to disclose matters which tend to increase the risk unless they are unusual in the sense just described.

42. These are important considerations to bear in mind when considering condition 1.29 and the storage of oil at Fujairah because there is a natural temptation in the light of what subsequently occurred to approach the question of disclosure as if oil in store there had been the subject of a separate policy. Mr. Sumption accepted that none of the vessels making up the floating storage facility could properly be described as "barges" and that condition 1.18 was therefore not apt to cover oil held there. It follows, therefore, that oil held in the facility was covered during the 1995-96 year only under the endorsements and during subsequent years only under condition 1.29 and to that extent it can be said that that condition provided separate cover in respect of a specific risk. However, it would be wrong to view those provisions in isolation from the remainder of the cover. As their wording indicated, condition 1.29 and the endorsements that preceded it merely extended the cover provided under condition 1.18 to oil held in floating storage at Fujairah and must be understood in that light. As such it represented a small part of the risks covered by the open cover as a whole.

    2. Throughput

43. The primary ground on which Alpina purported to avoid the open cover was that Glencore failed to provide proper information relating to throughput at Fujairah, both as to the volume expected to pass through the facility during the coming year ('estimated throughput') and the volume of oil that had passed through it in previous years ('historic throughput'). Mr. Kendrick submitted that both were material matters which ought to have been disclosed at the time of the original presentation in April 1995 and at the time of renewal in 1996 and 1997.

44. Figures for historic throughput were not provided to underwriters at any stage and the failure to provide them forms one of the grounds on which Alpina has sought to avoid the policies for the 1996-97 and 1997-98 years. An estimate of throughput for the coming year was given to Alpina when the cover on goods in floating storage at Fujairah was first placed in April 1995, but it was common ground that nothing was said about estimated throughput when the policy was renewed in 1996. Mr. Kendrick submitted that in the light of the circumstances surrounding the renewal in March 1996 Glencore had impliedly represented that the estimate given almost a year earlier remained good for the coming year. In fact, however, no one appears to have given any thought to the question, but the volume of oil being shipped into Fujairah by that time suggests that if a fair estimate of throughput had been given the figure would have been greatly in excess of that given the year before. Whether anything was said about estimated throughput at the time of renewal in 1997 was hotly disputed and is a matter to which I shall return in due course.

45. The issues that arise in relation to the disclosure of estimated throughput and historic throughput are rather different and the two can therefore best be considered separately.

    (a) Historic throughput

46. It is convenient to begin by considering the relevance of historic throughput. The starting point is the basis on which cover on oil held in floating storage at Fujairah was rated. The underwriters who gave expert evidence, Mr. Gooding and Mr. Blaber, agreed that the magnitude of the risk run by underwriters insuring goods in store depends on the value of those goods and the period of their exposure to insured perils. One way of assessing that risk where goods of the same kind are constantly being moved in and out of the same storage facility is by reference to throughput, a method which Mr. Blaber was inclined to support. Underwriters charge a premium by reference to average throughput over a defined period, usually a month, as broadly representing the exposure to loss in relation to a fluctuating quantity of goods. However, the use of throughput as a basis for assessing risk has its disadvantages, as Mr. Gooding forcefully observed, because it may not make sufficient allowance for the time goods remain in store exposed to the insured perils. Thus a low throughput may reflect a large but relatively static stock in which the quantity exposed to the insured perils at any one time is quite large, whereas a high throughput may result from a small but rapidly moving stock in which the quantity exposed to insured perils at any time is relatively small. An alternative method, and indeed the one favoured by Mr. Gooding, is to assess the risk by reference to the average quantity exposed to risk at any one time, but it is unnecessary for present purposes to debate the competing merits of different competing methods of achieving the same end because in the present case both the Lloyd's and London market underwriters who wrote the cover for the 1994-95 year and Alpina who wrote the cover for the following three years were content to base their rating simply on throughput.

47. The most significant feature of the cover on oil in store at Fujairah in the present case is that insurers were content to charge an 'in full' premium based on an estimate of annual throughput. That meant that the premium was fixed once and for all by reference to the insured's estimate of throughput and that the insurers were precluded from charging a premium proportionate to the actual risk if the throughput turned out to be significantly higher than the estimate they had been given. Equally, of course, the insured ran the risk of paying an inflated premium if throughput fell significantly below his estimate. It is not difficult, therefore, to understand why Mr. Gooding described a contract of this kind as a lottery or why he was at pains to point out that its obvious disadvantages for the underwriter can easily be avoided by stipulating that the premium is to be adjusted at the end of the year by reference to actual throughput. In his view an underwriter would only grant cover on terms like these for commercial reasons as part of a larger piece of business and by doing so would in effect be making it clear that he was not interested in actual throughput at all.

48. The consequences of writing an 'in full' premium based on an estimate of throughput were not lost on Alpina's expert, Mr. Blaber, either, but he approached the problem from a different perspective. He expressed the view that if an estimate of throughput is used as the basis for fixing the premium, the underwriter is entitled to assume not only that it has been made in good faith, but that it is likely to be accurate and not (in his words) 'prone to unreliability'. He said that if the insured could not provide an estimate that was likely to be highly accurate, he should say as much and decline to provide one at all.

49. In my view Mr. Blaber's approach fails to take sufficient account of the nature of an oil trader's business. The underwriter of goods in store is concerned both with the value of the goods he is insuring as well as the period for which they are exposed to risk. The market value of most, if not all, traded commodities fluctuates over the course of time, and may sometimes change quite rapidly in the space of a few days. That is as true of the oil market as any other. Accordingly, even if the insured were able to estimate with a high degree of accuracy the quantity of goods likely to pass through a given storage location during the policy year, any estimate of throughput in terms of value would be liable to falsification by movements in the market. In the case of a commodity trader, however, any estimate of throughput in purely quantitative terms is inevitably prone to unreliability because the nature of trading is essentially opportunistic. A trader will make use of storage facilities, usually for fairly short periods of time, to store parcels of goods in order to take advantage of opportunities as they arise. Although the nature of the business may be broadly predictable at a certain level, it is highly unpredictable at the level of individual contracts which are made in response to specific opportunities. An underwriter familiar with this kind of trade ought to be aware that the unpredictable nature of the business would render any estimate of throughput at any given storage location prone to unreliability, as Alpina's underwriters in fact were. Mr. Blaber did not seem to me to have a strong understanding of the way in which commodity traders in general, and oil traders in particular, do business and indeed did not appear to accept the general proposition that an underwriter is expected to be familiar with the characteristics of the trades he insures. In the event I found much of what he said both about the insured's duty in relation to estimates of the kind given in this case and the prudent underwriter's assessment of them unpersuasive.

50. Mr. Kendrick submitted that historic throughput was material and should therefore have been disclosed on renewal because any significant divergence from the estimate given a year earlier might lead the underwriter to question whether he should continue to write the risk for an 'in full' premium based on estimated throughput. That was certainly Mr. Blaber's view, but it was not shared by Mr. Gooding who considered that by expressing a willingness to consider writing the risk on terms such as those the underwriter was indicating that he was not interested in actual throughput at all. In his view the prudent underwriter would be well aware when dealing with a trader of this kind that however much care had been taken in producing an estimate of throughput, it could well turn out to be seriously inaccurate. If historic throughput showed that to be the case, it would only bear out something that he had already taken into account. If the estimate proved largely accurate, that could only be regarded as a happy coincidence and would not provide grounds for assuming that the estimate for the coming year would prove to be equally reliable.

51. On this question too I prefer the evidence of Mr. Gooding because in my view it reflects the reality of underwriting risks associated with this kind of business. Given the inherent variability of the market and of the nature of an oil trader's business, it is not possible to draw any conclusion from the fact that the estimate given at the beginning of the year does or does not prove to have been accurate by the time the year reaches its end. An underwriter who is prepared to write a risk of this kind for an 'in full' premium based on the insured's estimate of throughput is taking a gamble which cannot be justified in ordinary underwriting terms but may be justifiable on purely commercial grounds as part of a much larger contract. In those circumstances I am unable to accept that historic throughput is material to an underwriter considering accepting a risk of this kind on these terms. My conclusion is fortified by that fact that Alpina did not at any stage request figures for actual throughput even though its underwriters must have realised that the information was readily available.

52. This makes it unnecessary to decide whether, as Mr. Sumption submitted, Alpina must be presumed in the present case to have had sufficient knowledge of historic throughput from the information contained in the declarations bordereaux. Nonetheless, I think it appropriate to deal with the issue briefly.

53. At monthly intervals during the life of the cover Glencore sent Alpina through Lloyd Thompson declarations bordereaux identifying the cargoes declared to the open cover for transit risks and cargoes insured in store. Cargoes were mainly declared three months in arrears, although some were declared even later than that. Until July 1996 the bordereaux identified in relation to each cargo the quantity, grade of oil, loading and discharging ports, its value and the terms on which it was being handled. Many cargoes were shown in the bordereaux as being carried to Fujairah with the notation 'FOB-STOR' in a column headed "Terms". After July 1996 Lloyd Thompson prepared bordereaux in a new form containing additional information about the carrying vessels but excluding information about the voyages and ports of discharge. There was some evidence that Lloyd Thompson continued to provide bordereaux in the old form as well as the new, but Mr. Krause and Mr. Flett of Alpina both said that they had not received them and none were disclosed. I think it unlikely that more than one set of bordereaux were routinely supplied after July 1996 and in the light of the evidence as a whole I am satisfied that they were invariably in the new form.

54. Mr. Sumption submitted that by the time of the renewal of cover in 1996 Alpina had received monthly declarations bordereaux for the first seven months of the expiring year which made it clear that the throughput for that year was bound to be much larger than the estimate previously given. That was the best information of throughput over the expiring year that Glencore itself was then in a position to produce and therefore there was no failure on its part to disclose historic throughput. Alternatively, having been told that throughput was much greater than had been estimated and having failed to ask for more information, Alpina had waived any right to further disclosure.

55. It was common ground that the insured need not disclose any circumstance which is known or presumed to be known to the insurer and that the insurer is presumed to know matters which he ought to know in the ordinary course of his business. These may be matters of public knowledge or they may be matters which are, or ought to be, reasonably clear to the insurer from information already in his possession: see MacGillivray on Insurance Law, 10^th ed. §17-73. In London General Insurance Co. v General Marine Underwriters Association [1921] K.B. 104 the plaintiffs, who had insured a cargo on the vessel Vigo, effected a policy of reinsurance on the cargo with the defendants "lost or not lost". On the previous evening it was known at Lloyd's that part of the cargo on the Vigo had been destroyed and during the morning a casualty slip containing that information was sent round to various underwriters, including the plaintiffs. However, through pressure of work they failed to read it and the reinsurance was placed later that day by brokers acting on their behalf while both parties were still ignorant of the loss. When they discovered that the loss had already occurred by the time they accepted the reinsurance the defendants sought to avoid the policy. The court agreed that the policy was voidable for non-disclosure. The plaintiffs had information in their hands in the form of the casualty slip which related to a cargo in which they were already interested and it was no excuse for them to say that they had not had time to look at it. The casualty was therefore a matter of which they were presumed to have knowledge. The defendants, on the other hand, were not interested in the cargo until the reinsurance was placed. Even if they had received the casualty slip they could not have been expected to take notice of a loss in which they had no interest or to have borne it in mind when they were asked to accept the reinsurance later in the day. The decision in this case is of relevance for two reasons: first, because it demonstrates that actual knowledge on the part of the insurer is not required in circumstances where a reasonable insurer would have taken the steps necessary to obtain the information from materials in his possession; secondly, because it demonstrates that the steps that a reasonable insurer will be expected to take in obtaining information from materials already available to him will depend on the circumstances of the case.

56. Mr. Kendrick submitted that even the bordereaux produced up to July 1996 did not contain enough information to enable Alpina to understand which cargoes were being put into store at Fujairah, but I think that overstates the position. It is true that when Mr. Gooding was asked in cross-examination to explain the meaning of some of the entries in the bordereaux he was a little hesitant, but in general he was able to understand them and I have little doubt that in cases where the discharge port was stated to be Fujairah and the terms were noted as 'FOB-STOR' a reasonable underwriter would understand that to mean that the cargo had been bought f.o.b. and was to be discharged into store at Fujairah. However, it is correct that the same information could not be gleaned from the bordereaux sent after July 1996.

57. There was no real dispute that the declarations bordereaux were sent to Alpina for underwriting purposes and were reviewed on a regular basis, albeit briefly, by members of the underwriting department. I have no doubt, therefore, that Alpina must be presumed to have been aware of the individual items of information contained in them. However, it does not necessarily follow that it is to be presumed to have had the kind of information that would be gained only from an extended analysis of the bordereaux over the whole of the previous seven months. Although Mr. Gooding said that he would expect a prudent underwriter who had agreed to insure oil in storage for an 'in full' premium to monitor throughput carefully during the year, he agreed that it would be difficult for the underwriter to conduct a detailed analysis of throughput at Fujairah from these particular bordereaux and Mr. Blaber was of the view that no underwriter would even think of doing so. In my view it would have been possible in March 1996 to extract from the bordereaux a reasonably reliable indication of the level of throughput over the preceding months, but only by analysing them all in some detail. Storage risks at Fujairah were just one aspect of the contract and at that time were not regarded as of particular importance. I am unable to accept that an underwriter of an open cover like this could reasonably be expected in the ordinary course of business to carry out an investigation of that kind into one particular area of the cover or that the insured is entitled to assume that he has done so.

58. Mr. Sumption submitted, however, that in the present case Mr. Krause had in fact reviewed the bordereaux at the request of Mr. Panzeri prior to renewal in order to evaluate the pattern of business over the expiring year and that Alpina must therefore be taken to have been aware of the general level of throughput. I am unable to accept that submission for broadly the same reasons. I accept that Mr. Panzeri did delegate the task of reviewing the bordereaux to Mr. Krause and that the purpose of doing so was to obtain information that might be relevant to decisions that had to be taken on renewal, but I do not think he was expected to or did carry out the kind of analysis that would be necessary to produce any useful information on throughput at Fujairah. As Mr. Panzeri pointed out, the policy covered worldwide trading and there was no reason to identify storage at Fujairah as an area that called for detailed consideration. What was of interest to Alpina at the time of renewal was information of a broader nature altogether, such as whether declarations were generally being made in a satisfactory manner and how the value of shipments declared for transit related to the policy limit. It is true to say that, once one's attention is directed to it, it is possible to see from the bordereaux that there are quite substantial movements of oil into Fujairah, but I do not think that the underwriter can for that reason be expected to have examined the bordereaux with a view to working out how much cargo was being directed to that particular destination.

59. I am unable to accept, therefore, that in this case Alpina is to be presumed to have been aware from the declarations bordereaux that the level of throughput in the 1995-96 policy year was significantly higher than the estimate it had been given or that if Glencore would otherwise have been under an obligation to disclose historic throughput Alpina had waived such disclosure.

60. However, even if I am wrong in finding that historic throughput was not material, there remains the question whether Glencore's failure to disclose it induced the insurers to renew the cover on the terms finally agreed, either in 1996 or 1997.

61. Both Mr. Krause and Mr. Panzeri suggested at points in their evidence that, if they had been aware of the extent of the discrepancy between the estimated and actual throughput for 1995, they would have considered charging an adjustable premium calculated by reference to actual throughput, but the likelihood that they would have insisted on a change of that kind has to be considered in the context of what they already knew and of the negotiations as a whole. As to their existing knowledge, both Mr. Panzeri and Mr. Krause appreciated that estimates of throughput were likely to be unreliable but that had not led them to question the wisdom of writing the Fujairah cover on an 'in full' premium based on estimated throughput. The negotiations had been protracted and in some respects frustrating for Alpina mainly as a result of Glencore's insistence on restructuring the rating bands which had led to significantly greater reductions in premium than Alpina had hoped to concede. Despite that, however, the cover was expected to generate substantial profits and the business was sufficiently attractive for Alpina to wish to take a larger proportion itself than it had in the previous year. The actual throughput for the 1995 policy year was something over 6.5 million tons and there was no reason to think that it would increase very substantially in 1996-97. On that basis the change to an adjustable premium might have resulted in an increase in income of up to US$125,000, assuming that the premium based on estimated throughput would otherwise have remained unchanged.

62. According to Mr. Krause, by the end of the negotiations Alpina had reached its sticking point. That may be so in the context of the negotiations as they had developed, but the question that has to be considered in this context is whether Mr. Panzeri would have refused to renew the policy if Glencore had insisted on retaining the 'in full' premium for Fujairah based on an estimate of throughput. Every underwriter has his limit, of course, but I do not think that Mr. Panzeri would have declined to renew such a profitable contract for the sake of US$100,000 or thereabouts of premium income. He was aware that an 'in full' premium calculated on estimated throughput might well not properly reflect the true risk, but this was a relatively small element in the overall scheme of things. His evidence in cross-examination suggested that he did not consider historic throughput to be of any relevance and I am not persuaded that if the actual throughput for 1995-96 had been disclosed Alpina would have declined to renew on the same terms. In accordance with the practice of the Swiss insurance market the co-insurers were content to follow Alpina in relation to the underwriting of this part of the cover without expecting to be separately consulted.

63. The position at the time of renewal in 1997 is essentially the same. Again, the question is whether Alpina would have declined to renew the cover on the same terms if there had been disclosure of throughput in earlier years and for the same reasons I am not persuaded that it would.

    (b) Estimated throughput

64. For reasons that no longer matter, when Glencore first sought cover on oil in store at Fujairah it suggested that the premium be assessed by reference to throughput rather than the value of individual parcels delivered into store. Lloyd Thompson's original proposal to underwriters was for cover to be provided on the same terms as condition 1.18 at 0.015% on annual throughput, but in the event the underwriters agreed to provide cover for a flat rate premium of US$20,000 on the basis of an annual throughput value of approximately US$80 million, equivalent to a rate of 0.025%. It is not entirely clear what persuaded underwriters to charge an 'in full' premium based on estimated throughput rather than a premium adjustable by reference to actual throughput; it may well have represented a commercial decision of the kind envisaged by Mr. Gooding. However, once that basis of rating had been agreed, it became entrenched, largely as a result of the transfer of the business from the London market to the Swiss market in 1995. Although the original Lloyd's indorsement made it clear that the premium was related to estimated throughput, it was unnecessary for later documents to identify the underlying rate because the premium was stated 'in full'. Equally, when Alpina agreed to continue the cover in April 1995 the indorsement was framed in similar terms.

    (i) The 1996 renewal

65. It was common ground that nothing was said about estimated throughput at Fujairah for the coming year during the negotiations for the renewal of the policy in March 1996. In those circumstances Mr. Kendrick submitted that Glencore had not only failed to disclose its current estimate of throughput but had impliedly stated that the estimate given the previous April had not changed, thereby misrepresenting the true position. Mr. Sumption accepted that any estimate of throughput was material and ought to have been disclosed, but he took issue with the suggestion that there had been any misrepresentation in this instance and challenged the conclusion that the absence of an estimate had affected Alpina's decision to renew on the terms agreed. In order to consider these arguments it is necessary to describe briefly the course of the renewal negotiations insofar as they related to storage risks at Fujairah.

66. Discussions on renewal occupied the parties on and off from the beginning of January until the latter part of March. The main topic of debate was the reduction in rates, especially on older vessels, which Lloyd Thompson knew Glencore was determined to achieve. Alpina did not strongly resist the suggestion that rates should be reduced generally, but was very reluctant to concede a reduction in rates on older vessels because it had been a fundamental part of its underwriting policy on this cover to impose rates which favoured the use of younger vessels. The insurance of goods in storage, whether at Fujairah or elsewhere, does not appear to have given rise to any specific discussion. Early in March 1996 Mr. Divine produced a revised form of wording in which, among other things, he sought to consolidate the various endorsements relating to cover on goods in floating storage at Fujairah into a new provision, condition 1.29. There was a good deal of argument before me about the way in which Mr. Divine went about that task, but for present purposes it is sufficient to note that although the new clause provided for an 'in full' premium of US$50,000, it made no reference to throughput as such or to the basis on which the premium was calculated. On 5^th March 1996 Mr. Divine sent Mr. Krause a copy of the revised wording on the basis of which Glencore proposed that the cover should be renewed. It is clear that Mr. Krause read through the wording with some care, but only three provisions excited any comment and condition 1.29 was not one of them. As a result Alpina accepted condition 1.29 in the form proposed by Mr. Divine.

    (A) Non-disclosure

67. Mr. Sumption conceded that if Alpina was to be invited to renew cover on oil in floating storage at Fujairah at an 'in full' premium, Glencore's estimate of throughput for the coming year was material and should have been disclosed. That was Mr. Gooding's view and in the light of the way in which the proposal for renewal was made, I think that must be correct. This was not a case in which negotiations had proceeded on the basis that insurers would expect to charge a particular rate on the insured's estimate of throughput; cover at Fujairah simply did not feature in the parties' discussions at all. By their fax of 5^th March Lloyd Thompson on behalf of Glencore offered to renew on the terms contained in Mr. Divine's revised wording, including condition 1.29. Glencore's current estimate of throughput at Fujairah for the coming year was plainly a matter that would influence the judgment of a prudent underwriter in deciding whether to accept the proposed premium or demand a higher one.

    (B) Misrepresentation

68. The other way in which Mr. Kendrick put this part of the case was that by failing to give a revised estimate of throughput for the coming year and inviting Alpina to renew at an 'in full' premium of US$50,000 Glencore impliedly represented that its estimate of throughput was the same as it had been in the previous year when the premium had also been US$50,000. This view was supported by the evidence of Mr. Blaber. He expressed the opinion that if an underwriter who had written a risk of this kind on the basis of an estimate of throughput were not given a fresh estimate for the ensuing year he would be entitled to assume that the previous estimate still held good, especially in a case where the premium proposed by the insured was consistent with such an estimate. Mr. Gooding, however, disagreed. In his view an oil trader's estimate of throughput at a given storage location in any year was not the kind of information that could be expected to hold good in relation to subsequent years. The views of the experts are, of course, helpful insofar as they enable the court to understand more clearly how underwriters approach these matters, but in this case the question is not one that turns on the practice of the insurance market as much as on the particular communications between the parties in this case.

69. If nothing had passed between the parties that had any bearing on the issue of throughput, I think it would have been difficult for Alpina to argue that Glencore had impliedly represented that the estimate of throughput in the 1995-96 year given almost a year earlier continued to apply in relation to the coming year. In my view the nature of a trader's business is so inherently variable that one could not reasonably draw that inference. I am unable, therefore, to accept Mr. Blaber's suggestion that in a matter of this kind a statement made twelve months earlier is to be taken to have been impliedly repeated if nothing more is said. No doubt that may be true of some kinds of statements, but not one which relates to a matter as prone to variation as this. However, the fact that Lloyd Thompson put forward a proposal which included an 'in full' premium of US$50,000 takes the matter a stage further. When the cover on oil at Fujairah was placed with Alpina it was clear that the premium was related to the throughput (despite the fact that the rate was not spelled out), if only because Mr. Vanderson had suggested an increase from $20,000 to US$50,000 to cater for the increase in estimated throughput from US$80 million to US$218.75 million. Although Glencore was seeking a significant reduction in transit rates, it had not proposed a reduction in the rate for storage under condition 1.18 which remained at 0.015% and there is nothing to suggest that anyone thought that there should be a change in the rate at which oil in floating storage at Fujairah should be covered or in the basis of calculation of the premium. Accordingly, by proposing that cover be renewed at the same premium as under the expiring policy I think that Glencore was impliedly representing that its estimate of throughput was substantially the same as that which had been given the previous year. This representation was false because no one at Glencore had turned his mind to the question of throughput for the coming year. If anyone had been asked to give such an estimate it is likely that it would have exceeded that given in April 1995 by a substantial margin.

    (C) Inducement

70. Whether the misrepresentation or non-disclosure played any part in inducing Alpina to write the open cover on the terms finally agreed is a more difficult question, however. The expiring year had been successful from an underwriting point of view, as was demonstrated by the statistics which Lloyd Thompson sent Alpina at the beginning of the renewal discussions, and there is every indication that Mr. Panzeri was keen to participate in other areas of Glencore's insurance programme. Moreover, it is apparent from the documents charting the course of the negotiations that both parties were concentrating on the big picture. In particular, Glencore was not only seeking a significant reduction across the board in transit rates to reflect a weaker market, but was determined to reduce the rates charged on older vessels. This was resisted by Alpina and became the primary focus of the debate. As I have already pointed out, no attention seems to have been directed to the rate for land storage under condition 1.18.

71. The wording proposed by Mr. Divine was examined by Mr. Krause. His attention had been drawn to the significant changes and he appreciated that, since the document formed the basis on which Alpina was being asked to renew, it was necessary to examine it with a degree of care. Most of the changes, including the new wording in relation to Fujairah, were uncontroversial and elicited no comment beyond agreement. In preparation for the renewal negotiations Mr. Krause had reminded himself of the policy terms and the endorsements, but he was not confident that he had looked at Mr. Vanderson's fax of 3^rd April 1995 in which the request for cover at Fujairah had first been made and I think it unlikely that he had. However, if he had looked at the original indorsement relating to Fujairah he would have seen that at that stage the estimated throughput by value was US$218.75 million.

72. Mr. Krause was generally a fair witness, although I think he was sometimes inclined to exaggerate the strength of his recollection. He accepted that since Alpina was underwriting the Fujairah risk for an 'in full' premium it was important to obtain an up to date estimate of throughput for the coming year and he was well aware of the unpredictable nature of commodity trading in general and oil trading in particular. He could not explain, however, why he had not taken any steps to obtain an estimate of throughput from Glencore. Although at one point he said that when he read condition 1.29 in Mr. Divine's draft he had assumed that the estimate was the same as it had been for the previous year, he was candid enough to say that he did not specifically recall thinking about the point. That is not a matter for criticism, of course, since these events occurred over seven years ago at a time when this aspect of the cover was not controversial. The cover in relation to Fujairah was, after all, a small part of a much larger picture. It does, however, make it difficult for me to be satisfied that Mr. Krause actually directed his mind to the question of throughput at all.

73. Mr. Panzeri had allowed Mr. Krause to take charge of the renewal negotiations to enable him to gain experience of handling an account of this kind. Indeed, he hoped that in due course renewals could be handled with very little participation on his part. However, Mr. Krause had no underwriting authority so all decisions ultimately had to be taken by Mr. Panzeri himself. Having put him in charge of the negotiations, however, Mr. Panzeri relied almost entirely on Mr. Krause to provide him with the information he needed for that purpose. In these circumstances it is not surprising that Mr. Panzeri was unable to recall what documents relating to the cover he had read prior to, or in the course of, the negotiations.

74. In his witness statement Mr. Panzeri did not say that in the absence of any indication to the contrary he had understood the estimate of throughput at Fujairah to remain the same, but in his oral evidence he, like Mr. Blaber, said that as an underwriter he was entitled to assume that information given in the course of an earlier presentation remained reliable unless the broker told him otherwise. I find it difficult to reconcile that, however, with his clear understanding of the nature of an oil trader's business and the use of storage to take advantage of market opportunities. Mr. Panzeri is clearly an intelligent and experienced underwriter who is well capable of recognising that some kinds of information are inherently unlikely to remain valid for very long. Projections which depend on market movements and the exploitation of commercial opportunities are a prime example of such information. Mr. Sumption suggested to both Mr. Panzeri and Mr. Krause that Alpina was concerned only with the overall loss record and premium and therefore did not mind whether the premium charged for floating storage at Fujairah adequately reflected the risk represented by that particular element of the cover. They both denied that, of course, and I do not think that either of them had formed a positive view that the cover on oil stored at Fujairah could in effect be disregarded. However, there is no evidence to suggest that it was specifically discussed between Mr. Krause and Mr. Panzeri at any stage during the renewal process and I find it difficult to accept that Mr. Panzeri actually turned his mind to the question or to the significance of the absence of an up to date estimate of throughput for the coming year. If he did, it is all the more difficult to explain why Alpina failed to ask for confirmation of the estimate for the coming year. In the light of all the evidence I think that the most likely explanation for the omission is that this was regarded as a relatively insignificant element of the cover as a whole and was simply overlooked both by Lloyd Thompson and by Alpina. The evidence does not suggest that either Mr. Panzeri or Mr. Krause actually understood Glencore to be saying that its estimate of throughput at Fujairah in the coming year was the same as the estimate it had given in 1995, much less that the decision to renew on the terms proposed by Mr. Divine was affected by any understanding to that effect. In those circumstances there is no basis for concluding that the misrepresentation induced Alpina to enter into the contract.

75. I turn then to the other way in which Alpina puts this part of its case, namely, non-disclosure. An underwriter seeking to avoid on this ground must show that if proper disclosure had been made he would not have written the policy in the same terms, but in order to consider that question it is necessary to know what circumstances would have been disclosed to the insurer if the insured had complied with his duty. One obvious difficulty in the present case is that no estimate of throughput for the coming year appears to have been made, and even if it is permissible (which I doubt) to approach the question on the basis that an estimate could have been provided if necessary, there is little basis for knowing what it would have been. Estimating throughput at any storage facility over the course of a year is essentially a matter of judgment which involves assessing, as far as that is possible, the way in which storage requirements at that location are likely to develop during that period. If someone at Glencore (probably Mr. Heuzé) had been asked to carry out that exercise he would no doubt have assumed that the existing business with MTI under what was known as 'JV1' would continue, but whether its volume would reduce or increase would depend to a substantial extent on MTI's success as a bunker supplier. That would not have been the only matter to consider, however. For example, it would also have been necessary to consider whether Glencore was likely to sell cargoes or parcels of fuel oil on its own behalf using oil shipped to Fujairah and blended for its account by MTI outside the JV1 arrangements and there may well have been other aspects of the trade that could have affected the use of MTI's floating storage capacity. The actual throughput at Fujairah in the policy year 1996-97 (a 15 month period) was between 6 and 7 million tons. I do not think that actual throughput is necessarily a reliable indication of what Glencore's estimate would have been in March 1996, but, if one takes it as a rough yardstick, it suggests that Alpina would have been expecting to charge a premium in the region of US$150-175,000.

76. At this point one comes back to the question whether Mr. Panzeri would have insisted on an increase in premium of US$100,000 or so in respect of this element of the cover and for the reasons I have already given I do not think he would. The fact that Alpina was willing to write the Fujairah risk for an 'in full' premium based on estimated throughput in the first place suggests that it was willing to treat it as a commercial concession and I am not persuaded that a sum of that amount would have been regarded as sufficiently important in the ultimate analysis.

77. For these reasons I am unable to accept that Glencore's failure to provide an estimate of throughput for the coming year entitles Alpina to avoid the cover for the 1996-97 policy year.

    (ii) The 1997 renewal

78. The renewal of the policy for the 1997-98 year requires more detailed consideration. The central issue between the parties is what was said at a meeting between Lloyd Thompson and Alpina in Zurich on 30^th May 1997 during which the negotiations for renewal were substantially completed. Alpina's case was that there was once again no mention of throughput and that Mr. Panzeri and Mr. Krause therefore proceeded on the basis that the estimate given in April 1995 continued to hold good for the coming year. Glencore accepted that nothing was said in the course of the renewal negotiations about historic throughput, but maintained that in the course of the meeting Alpina was told that Glencore's estimate of throughput for the 1997-98 policy year was 7 million tons with a value of US$840 million.

79. Mr. Gibson and Mr. Warren began their preparations for renewal in May 1997. The market at that time was still soft; indeed, it was appreciably softer than it had been a year earlier and Mr. Gibson was therefore expecting to obtain a further significant reduction in premium. However, he also wanted to take the opportunity of reviewing the policy generally and therefore suggested that the members of the insurance department should each examine the current wording with a view to suggesting improvements for discussion at an internal meeting a few days later.

80. In preparation for those discussions Mr. Warren approached Mr. Hawkins for an estimate of throughput for the coming year and was given a figure of 7 million tons with a value of about US$840 million. When Mr. Warren reported those figures to Mr. Gibson he was very surprised. He asked Mr. Warren to check whether there had not been a mistake and to find out what proportion of the cargoes discharged into store at Fujairah were declared for transit risks under the policy. Faced with such a large increase in throughput he thought that Glencore could make use of the 30 days' free storage cover that was provided as an adjunct to transit cover under condition 1.18 as a bargaining counter to keep the premium for storage down. Mr. Hawkins confirmed that the figures he had given to Mr. Warren were correct and that about 90% of the oil in store was declared for transit risks under the open cover. However, according to Mr. Warren, Mr. Hawkins said nothing to him about blending or about how long oil generally remained in storage at Fujairah. Following these exchanges Mr. Warren produced a note headed "Renewal Proposals" in which he set out a number of ways in which he thought the policy could be improved. It included the following item:

    "Fujairah throughput 7mt - $840m – Large ppn [proportion] is cargo voyages declared. As much as 90%."

81. There was some debate about whether the note was written before, at, or after a discussion that took place at about that time between Mr. Warren, Mr. Gibson and Miss Went. I am inclined to think that Mr. Warren wrote it in advance of their meeting when he was going through the policy to identify the points he thought worth raising, but nothing really turns on the precise timing of it. I also think it likely that although Mr. Warren aired his points at that meeting, no decisions were taken on many of them at that stage.

82. Mr. Gibson and Mr. Warren met Mr. Vanderson, Mr. Divine and Mr. Roberts of Lloyd Thompson at Glencore's offices on 22^nd May to talk about the renewal of the policy. The primary topic for discussion was a modification of the existing rating bands for transit cover in order to obtain a substantial overall reduction in premium. At some point, probably after the main business had been concluded, Mr. Warren was invited to raise the points in his note. It seems reasonably clear that they were considered one by one since the file note later produced by Mr. Divine refers to a 'Wording amendments "wish list"' and follows the lines of Mr. Warren's note quite closely. There is no doubt that floating storage cover at Fujairah was discussed, but there is a dispute about precisely what was said about it. The relevant part of Mr. Divine's file note (which is the only surviving note of the meeting) reads as follows:

    "5) Regarding floating storage MW advises that of the current throughput values of approx. USD 840 M approx. 90-95% is blended and off-loaded within 30 days. Approx. 30,000 MT through per day (7 M MT !). Therefore consider USD50,000 premium unreasonable.",

    but Mr. Gibson and Mr. Warren did not accept that either of them had made any comments about blending or residence time.

83. One can see at a glance that although the throughput figures are correctly noted, the rest of the information recorded by Mr. Divine does not correspond with what Mr. Warren said he was told by Mr. Hawkins. In particular, there is no reference to the fact that most of the cargoes going into Fujairah were declared for transit risks, and instead of a reference to 30 days' free storage cover there is a statement that most of the oil is blended and off-loaded within 30 days.

84. Although those who were present at the meeting all said they could remember the substance of what was, and was not, said, I am not persuaded that any of them had a very clear recollection of the course of the discussions. The most important question from Glencore's point of view was a reduction in transit rates and how that might best be achieved and it was only after that part of the discussions had been concluded that Mr. Warren was invited to raise the points contained in his "wish list". Mr. Divine could not remember much about the discussion on rates, but I have little doubt that it was quite extensive and this of itself suggests that his recollection of the meeting is somewhat hazy. His file note contains no reference to this part of the discussion, despite the fact that he had apparently been taking a note throughout the meeting. I think the most likely explanation is that he made separate notes of the discussion on rating and the discussion of Mr. Warren's "wish list". He used his note of the discussion on rating as the basis for a fax which Lloyd Thompson sent to Alpina on 28^th May and then destroyed it as being of no further value. He then used his note of the discussion of the "wish list" as the basis for the file note before destroying that as well.

85. Mr. Gibson said that when he sent Mr. Warren back to Mr. Hawkins to find out what proportion of cargoes in store were declared for transit risks he did not ask him to find out how long oil typically remained in storage at Fujairah and he could not recall anyone saying at the meeting that the residence time was less than 30 days. Mr. Warren initially denied having said anything about residence time at the meeting and said that he did not know that blending was going on at Fujairah. However, at a later stage in his evidence he said that he had asked Mr. Hawkins at some point how long oil remained in store and although he could not remember what answer he had received, he did think that the turnover or 'residence' time was less than 30 days. It is interesting to note that although a residence time of under 30 days does not accord with the understanding then held by Glencore of the time at which property in oil passed to MTI, it does generally accord with the period for which stocks were shown in Mr. Hawkins spreadsheets as held to Glencore's order before being released to MTI following the issuing of ITT contracts.

86. Mr. Vanderson said that he remembered being told at the meeting on 22^nd May that 90-95% of the oil stored at Fujairah was insured for transit and that he was aware that in such cases it would attract 30 days' free cover in land-based storage. He did not recall any discussion of blending and off-loading, but he thought that Mr. Divine's file note was likely to be more reliable than his memory on this point. In my view Mr. Vanderson did not have a very good recollection of the meeting at all. That is not said by way of criticism given that these events took place six years ago and this matter was not regarded at the time as particularly important, but it does mean that he was unable to add very much to what was said by others.

87. Mr. Divine's recollection of this initial meeting was also generally hazy. He said he remembered being given a very large figure for throughput at Fujairah and recalled a discussion involving two points: declaration for transit and residence time. He was confident his note was accurate and firmly denied any suggestion that he had deliberately embellished it by adding the reference to residence time.

88. I am satisfied on the evidence as a whole that Mr. Divine prepared his file note from notes he had taken during the course of the meeting. However, I think it likely that those notes were very brief and that when he came to draft the file note five days later he drew on his recollection as well. Although the file note records comments made or positions taken by different individuals, I do not think it can be relied on as a verbatim note but should be treated as a summary of what that person said. Mr. Warren brought to the meeting two pieces of information about oil stored at Fujairah: the estimated throughput and the proportion of cargoes declared for transit cover. Information about declarations was obviously relevant to Glencore's ability to make use of the 30 days' free storage cover under condition 1.18 in its negotiations with Alpina, so it would be surprising if that had not been discussed, if only briefly. It is not a great leap from that point, however, to ask oneself whether a significant proportion of cargoes ceased to be on risk within that period. If they did, that would add considerable strength to the argument that the premium for floating storage should be very low. I should be surprised if that point had not occurred to Mr. Vanderson and Mr. Divine and if it did I should expect them to have raised it.

89. I think it highly unlikely that Mr. Divine would have referred to residence time in his file note if nothing at all had been said about it at the meeting. Moreover, although declarations for transit and residence time are separate and distinct points, in the context of any discussions about the appropriate level of premium for oil in storage they were quite closely related. I think that Mr. Warren's information about the proportion of cargoes declared for transit risks sparked a brief discussion in which one of the brokers raised the subject of residence time. Mr. Warren understood from his earlier discussions with Mr. Hawkins that a large proportion of cargoes were turned around in less than 30 days and said something to that effect. When he came to write up his file note a few days later Mr. Divine conflated the two ideas which he then expressed in his own words, although he attributed them to Mr. Warren.

90. Towards the end of May a meeting was arranged between Lloyd Thompson and Alpina for 30^th May in Zurich to discuss the renewal of the cover. On 28^th May the brokers sent a fax to Alpina enclosing statistics which showed that the contract could be expected to have produced a gross premium of about UD$9.5 million for the 15-month period to the end of June 1997 and reminded Alpina that the market had become extremely competitive. They then set out their specific proposals for amendments to the rating bands for transit risks together with a request for a 25% reduction in all other rates. Following a request from Alpina for additional information, Lloyd Thompson sent a further message on 29^th May enclosing a list of five wording and coverage issues and one matter of administration which they wished to raise at their meeting. These included an increase in the policy limit to US$80 million, and a review of the premium charged under condition 1.29.

91. The meeting on 30^th May was attended by Mr. Vanderson and Mr. Divine for Lloyd Thompson and by Mr. Panzeri, Mr. Krause and Mr. Stachel for Alpina. It was held in Alpina's boardroom around one end of an oval table. All those present were sitting within a few feet of each other and should, therefore, have been able to hear everything that was said. The discussions were conducted in English which is Mr. Krause's first language. It is not the first language of either Mr. Panzeri or Mr. Stachel, but Mr. Panzeri has a very good command of the language and Mr. Stachel spoke it competently and certainly well enough to understand what was going on. The main topic for discussion was inevitably the proposed restructuring of the rating for transit cover combined with an overall reduction in premium of 25%. Although they had other matters to discuss, both sides considered them to be of much less importance. It was Alpina's practice to avoid, if possible, giving any final decision in response to a presentation of this kind. Mr. Panzeri preferred to consider a broker's proposals, discuss them as appropriate within Alpina and take his time to consider them before communicating his decision in correspondence. This was not a meeting, therefore, at which Alpina was likely to make a final decision on rates, but it was one at which the brokers could be expected to put forward all relevant information in their possession.

92. The only real dispute between the parties about what was said at the meeting relates to the cover on oil in floating storage at Fujairah. In the light of subsequent developments that part of the discussion has naturally assumed great significance, but it is important to place it in its proper context. As the statistics showed, the contract had proved very profitable over the previous two years. The market had weakened considerably since 1996 so a demand for a significant reduction in overall premium could not really be resisted. Nonetheless, even the "as if" figures based on Lloyd Thompson's proposals indicated that the business would generate a total net premium income of over US$4.5 million in the coming year which, judging by past experience, could be expected to result in a substantial profit. This was not business that Alpina would have been happy to lose. Storage risks generally were not a matter of concern and did not give rise to any separate discussion. Floating storage at Fujairah was raised, however, because Lloyd Thompson had warned Alpina that they wished to discuss it, but there is no reason to think that prior to the meeting Alpina regarded that as an important part of the cover. From Lloyd Thompson's perspective, however, it was an issue that required careful handling because the figure they had been given for the estimated throughput was so much higher than anything previously notified to Alpina that it was likely to lead to a demand for a significant increase in premium.

93. There was a sharp conflict of evidence about what was said at the meeting in relation to this element of the cover. Both Mr. Vanderson and Mr. Divine said that Mr. Krause was told in terms that the estimated throughput at Fujairah for the coming year in terms of monetary value was US$840 million but that 90-95% of cargoes were declared for transit risks and were turned round within 30 days. Mr. Krause said that no specific estimate of throughput was given; on the contrary, he was told that activity at Fujairah was reducing and that the exposure there was now very limited.

94. Both Mr. Divine and Mr. Krause made notes during the meeting, but in neither case have they survived. Some time after the meeting Mr. Divine produced a typed file note and then destroyed his manuscript. The paragraph dealing with floating storage reads as follows:

    "8) Floating storage. In line with other policy reduction requests Assured looking for significant saving on this in-full annual premium. In addition to the overall package to consider advised that approx. 90-95% of the annual throughput values (approx. USD 840 M) are blended and off-loaded within 30 days, therefore in line with other policy concessions (i.e. 30 days included in transit) USD 50,000 seems unreasonable to the Assured. WK confirms that they will look to a greater than 25% saving in view of comparatively limited exposure in excess of 30 days."

95. On 3^rd June Mr. Krause drafted two fax messages, one to Lloyd Thompson and one for circulation to Alpina's co-insurers, and then destroyed his manuscript. The material part of his fax to Lloyd Thompson read as follows:

    "5. In terms of the limited storage exposure now incurred on the floating storage vessels, we confirm that we are prepared to reduce the in full premium under Clause 1.29 to USD 30,000."

96. The material part of his fax to the co-insurers read as follows:

    "The policy wording to be amended as follows:

    . . . . . . . . . . . . . . . . . . .

    4. In terms of a now very limited storage exposure incurred on the floating storage vessels, a reduction in the in full premium under Clause 1.29 to USD 30,000."

97. Mr. Divine's file note extends to almost two pages of quite closely typed text and is clearly much fuller than his original manuscript which is likely to have been quite brief. In drafting it Mr. Divine drew heavily on other sources - either his recollection of the meeting or his imagination. The second possibility has to be considered because Alpina challenged the authenticity of the section of the note dealing with floating storage at Fujairah. Mr. Kendrick suggested that Mr. Divine had typed the note some considerable time after the meeting when he could no longer recall whether he had specifically mentioned the figure for estimated throughput. He therefore referred to his own file note of the meeting with Glencore on 22^nd May together with Mr. Krause's fax of 3^rd June in order to enable him to reconstruct the discussion in a way that would support the reduction of the premium to which Alpina had agreed.

98. In deciding what was said about storage at Fujairah at the meeting of 30^th May I think it is helpful to begin by considering what the brokers and the insurers had in their minds as they approached it. Perhaps the first point to emphasise is that neither side regarded floating storage at Fujairah as a matter of great importance in the overall scheme of things. Lloyd Thompson had put it on the agenda for discussion, but the primary focus of their attention remained the rating for transit cover which they had addressed in some detail in their fax of 28^th May. Mr. Krause, who was going to take a leading role in the negotiations, did no more at that stage than note that the brokers wished to raise condition 1.29. If he had thought about it, he would probably have realised that Glencore ought to be providing an estimate of throughput for the coming year, but I do not think the matter was of sufficient importance at that stage for him to direct his attention to it.

99. Mr. Vanderson and Mr. Divine knew they had to raise the question of floating storage, however, because Glencore had given them an estimate of throughput which they were bound to pass on to Alpina. They were well aware that the estimate was likely to provoke a demand for an increase in premium, but neither they nor Glencore were willing simply to acquiesce in an increase if they could put forward a persuasive argument to the contrary. They had discussed the position with Mr. Gibson and Mr. Warren and understood it to be as reflected in Mr. Divine's file note of 27^th May. Since under condition 1.18 Glencore was entitled to 30 days' storage cover on cargoes insured for transit risks at no additional premium, they could argue that cargoes delivered into floating storage at Fujairah should be treated in the same way. Moreover, if most cargoes were turned round in less than 30 days, the extent of the insurer's exposure over and above that already covered by the transit premium was much smaller than the throughput figure alone would suggest. On that basis it would suit Glencore to emphasise exposure by reference to residence time rather than throughput.

100. In the event Mr. Vanderson and Mr. Panzeri both left it to their senior assistants, Mr. Divine and Mr. Krause respectively, to deal with the details of the proposals, confining themselves mainly (in the case of Mr. Vanderson) or entirely (in the case of Mr. Panzeri) to a general supervisory role. Mr. Kendrick criticised the way in which Mr. Divine had earlier dealt with various aspects of the cover and suggested that he was at best prone to making blunders and at worst capable of calculating behaviour bordering on dishonesty. He submitted that Mr. Divine had decided in advance of the meeting simply to ask for a reduction in the Fujairah premium to US$30,000 on the basis that Alpina's exposure there was now limited and to avoid, if possible, any mention of the figure of US$840 million. If Mr. Krause asked for an explanation he would tell him that 90-95% of cargoes went off risk within 30 days which was within the period of free storage cover included in the marine premium. Only if Mr. Krause probed further would he disclose that throughput by value for the coming year was estimated to be US$840 million.

101. Mr. Kendrick used the words "blunder" and "guile" when making submissions about the way in which Mr. Divine had previously dealt with various aspects of the cover, but they were really only colourful ways of referring to negligence and dishonesty. There undoubtedly had been errors and oversights in the administration of the cover in the course of the previous two years, but I do not think that they provide a sufficient basis for a finding that Mr. Divine was habitually negligent, much less that he was constitutionally inclined to sharp practice or dishonesty in professional matters. Moreover, one cannot ignore the fact that Mr. Vanderson was also closely involved in the renewal process. He attended the meeting with Glencore on 22^nd May and discussed the presentation of the Fujairah risk with Mr. Gibson and Mr. Warren. He was also going to be present at the meeting with Alpina and could be expected to realise at once if there were any serious shortcomings in the way Mr. Divine dealt with the matter. If Mr. Kendrick's analysis of Mr. Divine's strategy is to be accepted, therefore, Mr. Vanderson must necessarily have been a party to that strategy, yet there is even less reason to accuse him of dishonesty. The suggestion that both brokers wished to say as little as possible about throughput because they were embarrassed about the failure to provide an estimate at the time of the previous renewal does not strike me as at all plausible. In fact, the outcome of the discussions with Glencore a few days earlier suggests that both brokers went to the meeting intending to disclose the estimated throughput but at the same time intending to diminish its significance by emphasising the high proportion of cargoes in store that were previously declared for transit and the fact that most of the oil had a short residence time. Nothing they had been told, however, could justify a suggestion that the level of activity at Fujairah had reduced.

102. Evidence about the meeting itself was given by Mr. Vanderson, Mr. Divine, Mr. Panzeri, Mr. Krause and Mr. Stachel, all of whom, except Mr. Panzeri, were present throughout. It is convenient to deal with Mr. Panzeri's account first. I found him an impressive witness. He answered questions carefully and candidly and was conscious of the need to prevent his evidence from being coloured by hindsight as far as he could. However, his recollection of the meeting and the events surrounding it was not very clear and I am satisfied that he was not in the room when the conversation about floating storage took place, having left to warn the restaurant that they would be late for lunch. His contribution was therefore mainly confined to reporting what Mr. Krause had later told him, namely, that Mr. Divine had asked for a reduction in the premium of more than 25% because of the reduced activity at Fujairah. He said that if he had been told that throughput was estimated at US$840 million but that 90-95% of the oil was turned round in 30 days he would have refused to give any reduction and would have wanted to know what had caused the increase in throughput. As it was, he had agreed to reduce the premium to US$30,000 for commercial reasons in the understanding that throughput was reducing.

103. Mr. Divine said that he had included condition 1.29 in the list of wording issues he wanted to discuss because he needed to raise new factors that had not been considered in previous discussions. These were, of course, the proportion of cargoes declared for transit cover and the residence time. However, it made no sense to raise either of these matters except in conjunction with an estimate of throughput. He said that he knew he had to give Alpina the revised estimate of throughput and was confident that he had stated the figure of US$840 million. He was less sure whether he had also mentioned the figure of 7 million tons. He said that he remembered this part of the meeting clearly because Mr. Vanderson had interrupted him in full flow, somewhat to his annoyance, to emphasise that nearly all the oil going into storage at Fujairah had already been declared for transit cover.

104. Mr. Vanderson's role at the meeting was essentially supervisory. He made some introductory remarks designed to emphasise that the contract had been very profitable, that the market was currently weak and that Glencore was expecting a 25% reduction in premium as well as certain other improvements. He then left it to Mr. Divine to deal with the individual elements of the cover, most of which related to the way in which transit risks were rated. He said that he gained the impression that none of their demands surprised Alpina or were likely to be strongly resisted, and indeed that was later confirmed by Mr. Krause. He said that he was certain that Mr. Divine did not say that storage exposure at Fujairah was very limited. They had no instructions to that effect; indeed Glencore had told them that throughput was going up. He also said that he was sure that Mr. Divine had made the point about residence time and that he had intervened to emphasise that they were looking for a reduction in premium despite the increase in throughput. He could not recall whether he had said that a greater proportion of shipments going into store were being declared for transit than had previously been the case.

105. Mr. Krause confirmed that he had paid little attention to the floating storage cover prior to the meeting. He did not know why condition 1.29 had been put on the agenda and he was concentrating on the more important aspect of the transit cover rating. He knew the market was weak and he expected Lloyd Thompson to be seeking a reduction in the order of 25% all round. He confirmed that the discussion about floating storage had come well towards the end of the meeting. He said he regarded condition 1.29 as part of the 'nuts and bolts' of the policy, by which I think he meant one of the uncontroversial and relatively insignificant parts. He said that the conversation about Fujairah had taken place largely between himself and Mr. Divine and that the others present did not seem to be paying too much attention. However, he thought they could hear what was being said since they were all sitting quite close together. He said that Mr. Divine had told him that activity at Fujairah was reducing and that the exposure was now very limited and on that basis suggested that the premium be reduced to US$30,000. Mr. Divine did not say how low the level of activity had become and he did not ask, but he assumed that Glencore would be storing less oil than before and that there would therefore not be a significant amount on board the storage vessels. In essence he said he thought that throughput was reducing and that Alpina's exposure would be lower as a result. In re-examination he said that he had a very clear recollection of the meeting and was absolutely certain that no figure for throughput had been given in relation to Fujairah.

106. Finally I heard from Mr. Stachel. He was the manager of the claims department and was attending a renewal meeting on this policy for the first time. Although he attended the meeting in order to raise certain matters relating to claims handling, I am satisfied that he followed the discussion of other topics in a general way. He was not an underwriter, however, and his command of English, although good, was not good enough to enable him to follow a conversation about technical underwriting matters with ease. He said that he became aware at one point that a discussion was taking place about premium in relation to floating storage, but that he did not take much interest in it, which is entirely understandable. However, he said that if a figure of US$840 million had been mentioned he would have remembered it.

107. Mr. Kendrick submitted that in general Alpina's witnesses had a better recollection of the meeting than did the brokers and in some respects I think that is true. For example, I prefer Mr. Panzeri's and Mr. Krause's description of the seating arrangements to that given by Mr. Divine, although nothing of any importance turns on it. However, when it comes to the details of their discussions it is important to bear in mind both that the meeting itself took place six years ago and that the issue of cover on oil in floating storage at Fujairah was a relatively minor part of the overall picture. This had been very profitable business for the past two years and, although Alpina was resigned to a further reduction in rates, it was anxious to retain what was still a profitable and prestigious contract. On this occasion, as in the previous year, the main focus of the parties' interest was the overall level of premium based on the rating structure and level of premiums charged for transit cover. It would hardly be surprising, therefore, to find that none of those present had a very clear recollection of the details of their discussion about storage at Fujairah and even when a witness does profess to have a clear recollection of events, his memory may not be reliable. Moreover, although I have no doubt that all those who gave evidence were doing their best to help the court, it is very difficult for a witness in a case such as this to prevent his account from being coloured by reconstruction and hindsight when asked to consider how he would have reacted to statements or events that are in dispute. In these circumstances the contemporaneous documents and the inherent probabilities assume particular importance as was noted by Lord Goff when delivering the opinion of the Privy Council in Grace Shipping Inc. v C. F. Sharp & Co. (Malaya) Pte. Ltd [1987] 1 Lloyd's Rep 207 at 215-216.

108. This brings me back to Mr. Divine's file note. It covers the whole of the meeting, extending to ten separately numbered items in addition to paragraphs summarising what had been said by way of introduction and conclusion. With the exception of the paragraph dealing with floating storage (number 8) it was accepted by Alpina's witnesses as a full and fair summary of what had been said. That is of some importance in itself, because unless it was subject to a later alteration (of which there is no evidence) it suggests that it was drafted at a time when Mr. Divine had a good recollection of the discussions and is likely to be substantially accurate unless it was deliberately intended to be misleading. The final paragraph of the note contains a reminder to chase for a response "next week" and a note that Mr. Krause had promised one by "next Wednesday". Since the note could not have been typed before Mr. Divine returned to the office on Monday 2^nd June, the phrase "next week" is inappropriate unless he had already drafted it in manuscript over the week-end or was composing it from that perspective. Of the two I think it more likely that he had already produced a draft over the week-end, hence the reference to expecting a response by the following Wednesday. It follows that when he drafted the note he had his manuscript notes of the meeting with Alpina and his file note of the meeting with Glencore on 22^nd May available to him for reference, but not Mr. Krause's fax of 3^rd June.

109. If, as I think likely, Mr. Divine drafted the file note over the week-end following the meeting, it is unlikely that he did not recall with reasonable clarity the course of the discussions and since it is accepted that the remainder of his note is accurate, there is no obvious reason why his record of the discussion about floating storage should have gone badly wrong simply through a failure of recollection. Mr. Kendrick submitted that Mr. Divine was prone to making blunders and I accept that he may have partially misunderstood what had been said to him at the meeting with Glencore on 22^nd May. However, I find it difficult to accept that paragraph (8) of the file note can be explained in that way given the fact that on this occasion Mr. Divine was making a note of what he had said rather than recording his understanding of what someone else had said to him.

110. The only other possibility is that he deliberately concocted a false note of this part of the discussion to cover up the fact that he had failed to disclose important information to Alpina and had presented the risk on a false basis. However, there are equally serious objections to that conclusion. It has to be borne in mind that Mr. Vanderson was present during the meeting and by all accounts was sitting next to Mr. Divine. Accordingly, unless he was paying no attention at all, he would almost certainly have caught the gist of what Mr. Divine said in relation to throughput at Fujairah and would have noticed if he had dealt with it in a fundamentally misleading way. In those circumstances he would surely have intervened, unless, of course, he and Mr. Divine had decided before the meeting to handle the issue in that way.

111. Sensing, perhaps, the difficulties inherent in that way of looking at the matter, Mr. Kendrick offered as an alternative explanation the suggestion that Mr. Vanderson may have left this part of the presentation to Mr. Divine and, having done so, did not take any notice of how he dealt with it. As a result he simply failed to realise that Mr. Divine had omitted to give any figure for throughput in the coming year. However, I do not find that a convincing explanation. Although Mr. Vanderson was certainly content to allow Mr. Divine to take the lead, I can see no grounds for thinking that he abdicated responsibility entirely, particularly in relation to an issue that he was aware called for careful handling. His own evidence is that he took a close interest in what was being said and that is what I should have expected. That leaves only the possibility that Mr. Vanderson and Mr. Divine had decided before the meeting not to mention a figure for total throughput but simply to concentrate on the relatively small proportion that was not declared for transit cover and remained in storage for more than 30 days. However, I find that an even less plausible explanation. Both brokers knew that throughput at Fujairah had increased significantly and they both knew not only that their duty of good faith required them to disclose the estimate of throughput, but that they had been instructed by Glencore to do so and to put forward specific reasons why it should not lead to an increase in premium. The inclusion of condition 1.29 as an item for discussion in the supplementary agenda sent to Alpina on 29^th May indicates that they intended to follow that course. It would be extraordinary if they had subsequently abandoned that strategy in favour of making a misleading presentation on a subsidiary issue of this kind and no credible motive for their doing so has been put forward. But unless they were both willing to mislead Alpina, they could not have failed to give a figure for throughput, nor could they have told Mr. Krause that storage activity at Fujairah was reducing and was now very limited, as he asserted.

112. All this suggests that Mr. Divine's file note is probably broadly accurate, but even so, it is not entirely self-explanatory, both because the reference to the annual throughput value of "approx. USD 840 M" has been placed in brackets as if to suggest that the author was making a private comment rather than recording something that had been stated openly and because it contains no reference to any participation by Mr. Vanderson in this part of the discussion. At first sight the use of brackets might well seem to suggest a comment inserted for the benefit of Lloyd Thompson rather than a record of something spoken at the meeting, but in the context of the note as a whole the implication is, if anything, to the contrary. It is clear from parts of the note that are not controversial that in many instances the phrases in brackets record something that was either expressly stated or was the acknowledged basis of the discussion. For example, the phrase

     "(mutually agreed by all to be extremely aggressive)"

     which appears in the second introductory paragraph clearly reflects comments made around the table. Likewise, the phrase

     "(at 0.075% level)"

     in item (1) either reflects something that was expressly referred to or was known to be the basis of discussions because it had been stated in Lloyd Thompson's fax of 28^th May. In fact, although some of the phrases in brackets are more difficult to categorise, until one comes to the last paragraph headed "Conclusion", it is not clear that any of them are in fact private comments by Mr. Divine rather than reflections of comments made openly in the course of the discussions. Only in the final paragraph does one find what is clearly a private comment, namely,

     "(however, we have been surprised before)",

     but in fact the whole of that paragraph is a comment on the outcome of the meeting, not a record of what was said during it. Accordingly, although I do not think that a great deal of weight can be placed on the way in which the document is drafted, the fact that the references in item (8) to a throughput value of US$840 million and to 30 days' cover being included in transit premium are both in brackets provides no support for the argument that these points were not made openly. On the contrary, the fact that they appear at all is more consistent with the conclusion that they were.

113. The absence of any reference in the note to the intervention said to have been made by Mr. Vanderson when Mr. Divine was dealing with the issue of floating storage can, of course, be contrasted with the reference to his participation in the discussion on other occasions, especially in item (4) dealing with limits. The incident itself was described by Mr. Vanderson and Mr. Divine in supplementary statements served shortly before the trial began, neither of them having mentioned it in his original statement. Mr. Kendrick suggested that it was nothing more than an embellishment designed to give greater credibility to their evidence, but I do not find the omission from the note of a reference to an intervention of that kind a very powerful factor. The file note does not purport to record everything that was said during the meeting; that was not its purpose. It was intended to record the gist of discussions and Alpina's reaction to the points that were made. If Mr. Vanderson had made a separate comment of importance, one would expect Mr. Divine to have noted it, as indeed he did when recording the discussion covered by item (4), but if he merely intervened to reinforce a point that had already been made by Mr. Divine himself, there was much less reason for him to do so. I should not expect an incident of the kind described by the brokers to stick in the mind of Mr. Krause or Mr. Stachel who may have seen nothing unusual in Mr. Vanderson's intervening. I accept the brokers' evidence on this particular issue, though I do not think it carries the matter very far.

114. The only other contemporaneous documents that might be expected to shed any light on what went on at the meeting are Alpina's two faxes of 3^rd June. One of the striking features of these faxes, however, is not so much the absence of any reference to the figure of US$840 million as the statement that there was now "limited" (or, in the case of the fax to the co-insurers, "very limited") storage exposure at Fujairah. That is obviously what Mr. Krause understood the position to be at the time and to that extent it can be said that they support his account of the meeting. Another possibility, however, is that Mr. Krause's evidence on this point was based on, or was at any rate heavily influenced by, the wording of these faxes. Mr. Sumption submitted that once the dispute had arisen Mr. Krause would have turned to the documents to remind himself of what had gone on at the meeting. Having no clear recollection of what had been said, he naturally believed that he had been told exactly what they say, i.e. that there was now only a limited storage exposure at Fujairah, and he interpreted that to mean that Mr. Divine had told him that throughput had reduced. Mr. Krause said that he had not looked at the documents before giving his initial account to Alpina's solicitors. I find that surprising, but I do not think it is ultimately of great importance because in either case all that one gets from the documents is an understanding of how he viewed the position at the conclusion of the meeting. How he arrived at his view is another matter altogether.

115. I am quite satisfied that Mr. Vanderson and Mr. Divine knew they had to tell Alpina that the estimated value of throughput at Fujairah in the coming year was about US$840 million and that they intended to do so. I am also quite satisfied that they intended to emphasise as strongly as possible that, because most of the oil was declared for transit under the policy and was turned round in less than 30 days, the premium payable for transit cover was adequate to cover the storage risk as well. Although no figures were mentioned, Mr. Vanderson believed that since the policy was first written by Alpina there had been a significant increase in the proportion of cargoes declared for transit and he intended to make that point as well. The brokers said that they did make these points, but according to Mr. Krause none of them was mentioned. According to him Mr. Divine simply said that storage activity at Fujairah was reducing and was now very limited. The contrast between the different accounts is too stark to admit of confusion, but to accept Mr. Krause's account involves accepting that Mr. Divine and Mr. Vanderson both decided for some unexplained reason not to disclose the information about throughput or to put forward the arguments that they had discussed with Glencore, but instead to deal with the matter on a completely false basis.

116. A number of other points raised by Mr. Kendrick deserve consideration before finally reaching any conclusion on this question. The first concerns the size of the estimate. Mr. Gibson had been surprised when he was told that the throughput was expected to be 7 million tons and Mr. Divine was equally surprised when he was given that figure during the meeting with Glencore on 22^nd May. Mr. Kendrick submitted that Mr. Krause would surely have expressed surprised on being told that the value of throughput in the coming year was expected to be US$840 million. He would have made a note of it and would have reported it both to Mr. Panzeri and the co-insurers. Yet there is no evidence that he did express surprise, or that he made any note or asked any questions and he certainly did not refer to it in his message to the co-insurers.

117. This is a fair point because the figure for throughput was certainly very large as an absolute number, but one's reaction to it is likely to depend to a large extent on having some idea of what the throughput ought to be or had previously been and Mr. Krause had no point of reference by which to assess it. It had been completely overlooked the previous year and he had not reviewed the file in the course of preparing for the meeting in a way that left him with any idea of what figure for throughput to expect. If the point was made that most cargoes were declared for transit and subsequently came off risk within 30 days of discharge, it would be quite possible for him to lose sight of the figure he had been given for throughput while retaining in his mind the point that there was comparatively limited exposure in excess of the 30 days already paid for in the transit premium. That would explain why he told Mr. Panzeri that there was reduced activity at Fujairah and why he described the exposure as limited in the messages he sent out on 3^rd June. In any event, it was not Alpina's practice to inform its co-insurers of the detailed underwriting reasoning of what was a relatively peripheral aspect of the cover. I do not think, therefore, that much additional support can be derived from that quarter.

118. The next point arises out of the request for an increase in the limit of cover from US$60 to US$80 million any one vessel or any one loss any one location. This had been prompted by the shipment during the year of one particular cargo the value of which had been a little over US$70 million. Glencore was concerned that in a rising market there might be a few more shipments of that value, but they were expected to be few and far between and for that reason the increase was requested as "sleep-easy" cover. One can see it described as such in Mr. Divine's file note and in Mr. Krause's message to the co-insurers seeking their approval to the proposed changes. Mr. Kendrick submitted that describing the limit in that way was quite inconsistent with disclosing a level of throughput that would result in the insurers being constantly at risk up to the limit of cover in respect of oil in storage at Fujairah and that the use of the expression "sleep-easy" would have struck Mr. Krause and Mr. Stachel as incongruous. I could see the force of the argument if the limit had been discussed in the context of storage risks, particularly if it had been raised in the context of storage at Fujairah, but it was not. One can see from the file note that the increase in the limit was discussed in the context of transit cover and the value of individual shipments. It does not appear to have been discussed at all in the context of storage risks, whether at Fujairah or any other location, so the point loses much of its force.

119. Next there is the duration of the discussion. Mr. Panzeri left the meeting just before the discussion about floating storage took place, having gone to speak to the restaurant, and by the time he returned it was over. Mr. Kendrick submitted that it could not have taken Mr. Panzeri very long to speak to the restaurant and that the discussion must therefore have been very brief, too brief, indeed, for Mr. Divine to have explained why a reduction in premium could be justified despite a large rise in throughput. This again is a fair point, but it depends on knowing how long Mr. Panzeri was out of the room and on that the evidence is unclear. It would not take very long for the brokers to make the point that the transit premium included a period of storage cover, or to make the point that very few cargoes remained on risk in store for more than 30 days. The time needed to deal with the point would depend largely on whether it was immediately understood or needed further elucidation. No one suggests that there was any prolonged discussion. Mr. Krause was there to listen to what was said, not to negotiate, although according to Mr. Divine's note his response to the request for a reduction in premium was positive.

120. In the event Alpina agreed to reduce the premium for floating storage cover at Fujairah from US$50,000 to US$30,000. None of the witnesses had any very clear recollection of who first put forward that figure which itself tends to emphasise that recollections of the meeting were generally poor. However, I think it likely in the nature of things that if the brokers were seeking a reduction of more than 25% in the current rate they would have had a proposal to make and would not simply have left it to Alpina to come up with a figure. Some support for the conclusion that the figure was first put forward by Mr. Divine in the course of the meeting is to be found in Mr. Krause's fax of 3^rd June in which he confirmed a reduction to US$30,000. There is no mention of that figure in Mr. Divine's file note, but I do not think that significantly undermines its reliability since it may have been put forward as an indication of what the brokers were looking for rather than a firm proposal.

121. Looking at the evidence as a whole I have come to the conclusion that Mr. Divine's file note accurately identifies both the substance of the discussion on the Fujairah cover and the general way in which it was handled. It is understandable that he would wish to make the point early on that most of the cargoes were declared for transit and turned round within 30 days in order to minimise the adverse impact of the throughput figure itself. I am unable to accept that he told Mr. Krause that storage activity was reducing; he knew it was not and I can see no reason why he or Mr. Vanderson should have suggested otherwise. I think it likely that either Mr. Vanderson or Mr. Divine said that exposure was much lower than had previously been assumed, but only as part and parcel of the argument that the risk was already covered by the premium charged for transit cover.

122. The duty of disclosure requires the insured to place all material information fairly before the underwriter, but the underwriter must also play his part by listening carefully to what is said to him and cannot hold the insured responsible if by failing to do so he does not grasp the full implications of what he has been told. Although Mr. Krause and Mr. Panzeri said that it was a matter of importance to them to charge a premium for cover at Fujairah at a rate appropriate to the risk, I am unable to accept that they viewed it in that way at the time. If they had, it is unlikely that they would ever have agreed to write the cover for an 'in full' premium based on an estimate of throughput. I am satisfied that the figure of US$840 million was expressly stated, but the precise manner in which that occurred is more difficult to determine. I think it unlikely that Mr. Divine simply announced the figure baldly before launching into his argument on the transit premium; it is much more likely that he dealt with it in the context of an overall presentation as is suggested by his note. However, that is far from saying that Mr. Divine and Mr. Vanderson failed to put the figure fairly before Mr. Krause. On the contrary, I think they did, but that Mr. Krause was more concerned with the degree of exposure over and above that which could fairly be subsumed in the transit risk. It did not call for more detailed analysis, being a relatively unimportant part of the policy – 'nuts and bolts' as he put it – for which the premium was trivial in the context of the policy as a whole. It follows that in my view this ground of avoidance was not made out.

     3. Glencore's relationship with MTI

123. Alpina contended that the nature of Glencore's relationship with MTI was such that it ought to have been disclosed both when the open cover was first written in 1995 and when it was renewed in 1996 and 1997. However, since Alpina is not seeking to avoid the cover for 1995, it is possible to concentrate on the position as it existed in 1996 and 1997.

124. In my judgment in Phase 2 I described the development of the relationship between Glencore and MTI from its origins in some occasional joint venture sales in 1992 through simple storage arrangements to the long-term storage and sales arrangements known as JV1 and JV2. The arrangements known as JV1 were intended to enable Glencore and MTI to exploit the market for marine bunker fuel in the Gulf and further afield through a joint venture which made use of Glencore's position as an established oil trader and MTI's expertise as a producer of blended fuel oil. The precise nature of these arrangements is set out in some detail in the Phase 2 judgment and need not be repeated here. In essence they involved the purchase by Glencore of oil which was delivered into floating storage at Fujairah where it remained until it was sold to MTI by in-tank transfer prior to being blended to produce bunker fuel. However, although most of the oil brought into Fujairah was sold to MTI in that way, from time to time Glencore drew oil from its stocks at Fujairah in order to satisfy sales to third parties and in some cases parcels were blended by MTI to Glencore's order for that purpose. The profits and losses derived from the sale of bunkers and fuel oil cargoes were shared between the parties.

125. Under the formal contractual arrangements which established JV1 oil brought into Fujairah by Glencore remained its property until a stock transfer certificate had been issued in favour of MTI in respect of a particular parcel pursuant to an ITT contract. However, the parties conducted their business in a way which did not reflect the formal requirements of the contract so that by June 1996 at the latest, and probably well before that, MTI had become entitled to make use of oil sold to it as soon as the ITT contract had been made, even though no stock transfer certificate had been issued. The manner in which that came about is also described in the Phase 2 judgment. Glencore did not realise that the way in which it conducted this aspect of its business had led to that result and continued to believe that by retaining title to the goods it had protected its position. That was not the case, however, because MTI generally used the oil it had agreed to buy quite soon after the ITT contract had been made and as soon as the oil was incorporated into a blended product Glencore's title disappeared.

126. These arrangements were modified in June 1996 with the introduction of JV2. That was not a joint venture in the true sense because instead of sharing the profits derived from the sale of bunkers, MTI bought oil from Glencore at cost plus a small additional sum and retained the profits from the sale of bunkers for itself. Apart from that, however, the nature of the relationship remained essentially unchanged: Glencore bought cargoes of oil which it stored with MTI at Fujairah pending sale in individual parcels by ITT contract. Throughout the period with which I am now concerned oil in storage at Fujairah was held in commingled bulk segregated only by grade.

127. Many of the vessels by which oil was carried to Fujairah pursuant to these arrangements were operating under time charter to MTI. By April 1995 the relationship between the parties had already become so well established that they agreed to dispense with discharge inspectors at Fujairah (except in certain specified cases) on the basis that, if the ship discharged in full, MTI would accept liability for the receipt of the full quantity stated in the bill of lading. From Glencore's point of view this was an attractive course, both because it saved the cost of the inspectors' attendance and because it meant that it did not have to worry about minor losses on the voyage.

128. Alpina submitted that there were a number of features of the relationship between Glencore and MTI which, taken individually or together, enhanced the risk of loss and were so unusual that they ought to have been disclosed. These were (a) the volume of oil traded between the parties under these arrangements; (b) the fact that the oil was to be sold to MTI which carried on the business of blending and selling fuel oil; (c) the lack of arrangements for the inspection of cargoes on arrival and acknowledgment of oil held in stock; (d) the absence of arrangements for the verification of quantities held in store; (e) the lack of independent inspection; (f) the fact that MTI rather than Glencore chose the vessels that made up the floating storage facility. I shall consider these individually, but before doing so it is necessary to say something more about the insured's duty of disclosure in this context.

129. When a trader seeks insurance in the form of an open cover the underwriter is aware that most, if not all, of the risks that will be declared under it are unknown to the insured at the time the cover is placed and so cannot be the subject of any specific disclosure. Moreover, in the absence of any requirement in the policy to do so, the insured is not obliged to notify the insurer of matters affecting any particular risk which falls within the terms of the cover. It follows that the underwriter is content to assess the risk undertaken by giving cover in that form by reference to matters that affect the range of risks he is accepting, not by reference to matters that are relevant only to individual risks falling within the terms of the cover. As I have already pointed out, therefore, the insured's duty of disclosure does not normally extend to matters relating to individual risks, even if he knows of them (such as, for example, a cargo that is about to be carried on a particular voyage falling within the terms of the cover) because they are not material to the underwriter's assessment of the risk represented by the policy as a whole. It is not surprising, therefore, that Mr. Gooding and Mr. Blaber agreed that cargo underwriters writing open covers do not expect to be given information about the underlying commercial arrangements between the insured and those with whom he is doing business. Apart from anything else, since such arrangements are liable to change during the life of the policy, information of that kind would not normally assist the underwriter's assessment of the risks represented by the cover as a whole. Underwriters writing this class of business are presumed to be aware of all the normal incidents of the trade in question and the insured is therefore only bound to disclose those aspects of his business, if any, that are extraordinary in the sense that they fall outside the range of circumstances that a prudent underwriter would foresee as liable to arise. I would accept, however, that a combination of circumstances, none of which could be described in isolation as extraordinary, could in the same way give rise to a duty of disclosure. Insofar as Mr. Blaber was inclined to suggest that any relationship that could have an adverse effect on the risk should be disclosed, I am unable to accept it. Mr. Gooding's evidence to the contrary is in my view more consistent with the authorities to which I referred earlier.

130. Mr. Kendrick submitted that some or all of these features to which I have referred materially increased the risk of loss and that individually or in combination they represented such a departure from the ordinary state of affairs that they could not reasonably have been within the contemplation of a prudent underwriter familiar with the business of oil trading.

131. Cover on oil in floating storage at Fujairah was not provided under the general terms of condition 1.18 and to that extent it can be said not to fall within the general scope of the open cover. On the other hand, it is not right to treat it as if it were an entirely separate risk divorced from the policy of which it formed part. In substance condition 1.29, as its terms recognise, extended the operation of condition 1.18 to one particular location. Because that location was subject to marine perils as well as the perils ordinarily associated with storage, Glencore was probably under a duty to disclose matters that were relevant to that aspect of the risk since Alpina could not be presumed to have taken those into account when assessing the storage risk generally despite the fact that condition 1.18 covered storage in barges. However, I do not think that Glencore was under a duty to disclose matters relating only to the usual risks of storage since Alpina must be taken to have assessed them already when considering condition 1.18.

132. It is convenient to mention briefly at this stage one factor relied on by Mr. Kendrick as indicating that the business was unusual, namely, the size of the loss. It was certainly very substantial, amounting to approximately 2 million tons of oil in all, but it must be remembered that a very large overall deficiency may represent a large number of relatively small losses. Indeed, very many large scale frauds have been committed in exactly that way. The fact that such a large loss occurred is not of itself any indication that the relationship between the parties was out of the ordinary or that the business was being conducted in a way that was beyond the contemplation of a prudent underwriter. At most it may suggest that some of the safeguards against fraudulent practice that might have been expected to exist were not present, but that is a question that must be considered on its own merits.

133. One aspect of the way in which oil was stored at Fujairah that can usefully be touched on at this stage since it has a bearing on many of the arguments put forward by the parties is the practice of commingling oil in store in a common bulk. It was generally accepted that the commingling in storage of oil belonging to different owners is a recognised feature of the oil trade. Mr. Kendrick submitted that it is less common than Mr. Sumption sought to suggest because most storage contractors offer facilities for segregated storage and most depositors prefer to have their goods held in that way, but in my view that is beside the point. Under a policy of this kind the underwriter accepts the risk of any recognised practices of the trade and in the present case both of those who were called as expert witnesses of the oil trade, Mr. Rambaud and Miss Bossley, accepted that commingled storage was not unusual. It must therefore be assumed to be within the contemplation of any underwriter who grants an open cover on storage risks worldwide and indeed Mr. Panzeri, and probably Mr. Krause as well, was aware of it.

134. As a general rule where goods belonging to different owners are commingled so that the original parcels cannot be identified the owners of those parcels become tenants in common of the bulk in proportions which reflect their respective contributions. Each time an owner draws from or adds to the stock the interests of the contributors are adjusted by law so as to reflect their contributions to the resulting bulk. If there is a loss, all those who were interested in the bulk at the time of the loss bear a rateable proportion of it and their interests in the remaining bulk are adjusted accordingly. These principles, which apply both under English law and the law of Fujairah, are discussed in my judgment in Phase 1.

135. What constitutes the commingled bulk for these purposes is essentially a question of fact, but the very fact that individual contributors to the bulk become owners in common of the whole in proportion to their contributions has certain consequences for the keeping of records and the verification of the amounts held to their order. Foremost among these is the fact that an inspection of the tanks or compartments containing the bulk will not by itself enable the inspector to ascertain how much is being held to the order of his own client. In order to determine that he needs to know what amounts have been contributed to (and drawn from) the bulk by others. That is essentially an accounting exercise which requires access to the records kept by the operator. Of course, if the whole of the bulk falls short of the quantity that should be held for his client, the inspector will know that something has gone wrong, but even then he will not be able to determine how much is still held for his client without knowing what has been drawn and by whom and that again depends on obtaining access to records which the operator may not readily disclose if they relate to the affairs of commercial competitors. Moreover, if there has been some wrongful dealing with the bulk, calculating residual entitlements can be a complex task, as this litigation has demonstrated.

     (a) The volume of oil traded between the parties

136. The first aspect of the relationship that Alpina says should have been disclosed is the extent of the business that was being conducted by Glencore with MTI. The amount of oil being shipped by Glencore into Fujairah was undoubtedly substantial: well over 7 million tons passed through the facility during the 15 months to the end of June 1997 and the amount expected to pass through in the 1997-98 year was estimated at 7 million tons. An analysis of movements during the period from June 1996 to February 1998 made by Alpina during the Phase 2 trial shows that as at June 1997 the amount in store at any one time had varied between about 125,000 tons (in the period from June 1996 to October 1996) and 400,000 tons (from November 1996 to June 1997) with a value of between US$15 million and US$48 million. (It rose substantially during the period between August 1997 and February 1998, but there is no reason to think that Glencore could have foreseen that in June 1997 when the cover was renewed.) None of the witnesses suggested that these were abnormally large quantities to hold in store by the standards of the trade and they fall well short of the 13 million barrels (about 2 million tons) of crude oil with a value of about US$200 million that Glencore held in store at Saldanha Bay, South Africa for a long period in 1998. Even the total amount of the throughput was not a particularly large proportion of the total quantity declared to the cover in the course of those two years. I am unable to accept that the quantity of oil passing through the facility was particularly unusual in itself and it must be borne in mind that the risk of accumulations from time to time in excess of the limit of cover was expressly accepted by Alpina in condition 1.20 under which the limits were to be deemed first loss.

     (b) The sale of oil to MTI as end-user

137. Next Mr. Kendrick submitted that the relationship between Glencore and MTI was essentially one of seller and buyer to which the arrangements for storage were merely an adjunct. That was sufficient in his submission to take it outside the ambit of the kind of arrangements that an underwriter could be expected to have in contemplation and enhanced the risk of loss.

138. The commercial purpose of the JV1 arrangements was more than merely the sale of oil by Glencore to MTI. It was a joint venture in the true sense and Glencore took an active role in marketing the finished products as well as in obtaining the raw materials needed to produce them. The storage of the raw materials was an integral part of the arrangement, as it would be for any continuous operation of that kind, and the volume of stock held from time to time depended primarily on the demand for finished products. Since a significant part of Glencore's contribution to the joint venture was the supply of raw materials which remained its property until transferred to MTI, it was not simply performing the role of a financier, but the fact remains that MTI could not have financed the purchase of such large stocks from its own resources. Quite apart from all that, however, I do not think that it means very much to say that the storage was simply an adjunct to the sale of the oil to MTI. Storage is rarely, if ever, an end in itself. It is simply one step on the way to the sale or consumption of the goods. The real thrust of Mr. Kendrick's point was that the goods were being delivered into the possession of the intended end-user.

139. Mr. Kendrick submitted that for an oil trader to enter into an arrangement under which goods are stored with the intended end user is unusual, but Mr. Rambaud did not think so, having had experience of transactions in various parts of the world, albeit involving straightforward sales of much smaller quantities. Moreover, it is interesting to note that both Mobil and Caltex entered into broadly similar arrangements with MTI. For her part Miss Bossley described such an arrangement as "atypical but not unheard of" and she recognised that the ability to recognise market niches and to construct innovative deals is one of the prime functions of a trading company. Her main concern was to emphasise the need to recognise the risks involved in such an arrangement and to guard against them in an appropriate manner.

140. Mr. Kendrick, supported by Mr. Blaber, suggested a variety of reasons why the risk of loss was greater where goods are placed in the control of an end-user in MTI's position. I accept that there is an enhanced risk of loss where goods are stored with a person who is to make use of them in the course of his own business, either as raw materials for the production of finished goods or simply as stock for resale to third parties. That is not because the person in control of the goods is more likely to be dishonest than one who is merely storing them for redelivery, but because if he is tempted to be dishonest, it is easier for him to dispose of the goods through his existing outlets. However, that of itself is not enough to give rise to a duty of disclosure if the nature of the arrangement does not fall outside the range that is reasonably to be contemplated by an underwriter writing this class of business (including, in this case, storage at refineries under condition 1.18). In my view Mr. Blaber did not really deal adequately with the question whether the JV1 and JV2 arrangements were so unusual that an underwriter writing an open cover ought to be advised of them. In this context he tended to give the impression that he was approaching the question as if he were considering an individual facultative cover rather than an open cover for a commodity trader carrying on a worldwide business. I also felt, not for the first time, that his knowledge of the way in which commodity traders do business was less well developed than it might have been. In the light of the evidence given by Miss Bossley and Mr. Rambaud I accept that arrangements of the kind embodied in JV1 and JV2 are uncommon, but not that they are so unusual as to take them outside the wide range of possibilities that Alpina must be presumed to have had in mind.

     (c) The absence of a regular inspection and recording regime

141. Three aspects of the regime at Fujairah came in for particular criticism by Alpina: the failure from April 1995 onwards to employ independent inspectors to monitor discharge of cargoes carried into the facility, the lack of formal holding certificates and the absence of regular inspections to determine the quantities held by MTI.

142. In the initial stages of their relationship Glencore and MTI did instruct independent surveyors to monitor the discharge of all oil cargoes into the facility, but, as I have already mentioned, in April 1995 they agreed to dispense with them (save in relation to certain shipments from Iran by vessels not chartered to MTI) except in cases where there was a significant difference between the ullage of the vessel's tanks on arrival and the bill of lading figures. There were several reasons for doing that. First, most of the vessels carrying cargoes into Fujairah were operated by MTI under time charter and the cargoes were carried on terms that rendered MTI ultimately liable for losses in the course of the voyage. It did not add greatly therefore to its existing liability towards Glencore for MTI to treat the bills of lading as conclusive evidence of the quantity discharged into store. Secondly, by dispensing with discharge inspectors the parties avoided the costs associated with their appointment. Moreover, most of the cargoes brought into Fujairah came from further up the Gulf so the voyage to Fujairah was quite short and the risk of loss in transit was therefore relatively small. In addition, many cargoes were carried in the same group of vessels which plied continuously between Fujairah and ports further up the Gulf, so that if there was a short discharge on one voyage, the quantity left on board was likely to be discharged on the next.

143. The primary function of discharge inspectors is to verify the quantity discharged into the custody of the receiver; in many cases they will also monitor the quality of the cargo being discharged to ascertain whether there has been contamination on board the vessel. Discrepancies between ship's figures and shore figures at the port of loading are not unusual and even in the absence of a casualty of some kind there may be minor losses due to the ordinary incidents of carriage such as evaporation and clingage. Independent verification of the quantity actually discharged is therefore usually of as much interest to the receiver as it is to the ship or the holder of the bill of lading if he is not himself the receiver. However, although inspectors cannot monitor discharging operations that do not take place, they are not usually appointed to ensure that a vessel which is ordered to discharge at a particular port actually does so. There are likely to be easier and cheaper ways of achieving that end.

144. Mr. Kendrick submitted that by failing to insist on the attendance of independent inspectors to monitor the discharge of every vessel ordered to Fujairah Glencore had created a risk that cargoes might be diverted by MTI to other destinations without being discharged at all, or might be discharged in whole or in part into other vessels for immediate export without ever being received into store, as indeed happened. Furthermore, he submitted that the effect of agreeing to treat bill of lading figures as binding was not only to transfer any losses (actual or apparent) occurring during the voyage from the carrier to the receiver (including those falling within the 0.5% deductible applicable to transit cover), but to submerge them into a much larger loss such as that in respect of which the present claim is made.

145. Once again it is necessary to consider separately the nature of the arrangement and the extent to which it enhanced the risk of loss. Not surprisingly, both Mr. Rambaud and Miss Bossley considered it normal to appoint independent discharge inspectors when a cargo of oil is put into store and I am satisfied that that is the case. However, Mr. Rambaud was certainly not surprised at the idea that parties in the position of Glencore and MTI should agree to be bound by bill of lading figures and both he and Miss Bossley recognised that it favoured Glencore. Miss Bossley was at pains to emphasise at various points in her evidence that it was necessary to view the picture as a whole and that is a matter to which I shall return, but she did not go so far as to say that this particular aspect of the arrangements between Glencore and MTI was extraordinary and I do not think it was. Two major international oil companies that were doing business with MTI at that time also agreed to dispense with discharge inspectors in circumstances where cargoes were carried into Fujairah on ships time-chartered to MTI. That is not conclusive, of course, but it does provide further support for Mr. Rambaud's view that it is not uncommon in certain circumstances to accept bill of lading figures as binding. I am not persuaded that an arrangement of this kind is so unusual that it ought to be disclosed in connection with a policy of the kind under consideration.

146. That being so, it does not matter whether this particular aspect of the relationship enhanced the risk of loss. I think it did, but only to a very limited degree. Mr. Gooding, in common with Mr. Rambaud, found it difficult to contemplate the appointment of discharge inspectors otherwise than for the purpose of verifying the quantity and quality of cargo discharged. Neither considered that inspectors would be appointed simply to verify the fact that discharge had occurred, although that would necessarily be a consequence of their attendance. Mr. Blaber considered it standard practice for underwriters to insist on a discharge survey, as indeed did Mr. Gooding, at least where insurance is provided on "guaranteed outturn" terms. Insofar as his concern was addressed to the need to establish the true quantity discharged, I accept what he said, but I am unable to accept that a prudent underwriter would want the insured to appoint a discharge inspector simply to ensure that the cargo was discharged and to guard against fraudulent diversion. With hindsight one can see that the appointment of discharge inspectors would probably have prevented the diversion of complete cargoes and perhaps also the transhipment and immediate export of part cargoes and to that extent the agreement to dispense with them did facilitate the loss. However, I am not persuaded that that is a consideration that would have presented itself to the mind of a prudent underwriter at the time. In my view this particular limb of the argument is largely the result of hindsight.

147. The suggestion that underwriters' position might be prejudiced by the aggregation of small shortages is one that I am unable to accept either. Since most of the vessels carrying oil into Fujairah for Glencore were regularly operating on the same route, the scope for shortage was much reduced. In any event, in order for a claim to be made under the policy it is necessary for Glencore to establish not only that there has been a loss but the amount of that loss. The agreement to dispense with inspectors may render the bills of lading conclusive evidence as against MTI of the quantity delivered into storage, but it does not render them conclusive evidence against Alpina. Moreover, I think Mr. Sumption was right in saying that in order for it to be fully effective as between Glencore and MTI it was implicit in the agreement that any shortfall was to be made good by the immediate transfer of the corresponding quantity from MTI to Glencore and likewise that any excess over bill of lading quantity was to be immediately transferred from Glencore to MTI.

148. Finally Mr. Kendrick submitted that Glencore gave Alpina the clear impression that independent discharge inspectors would be instructed whenever possible and that having done so it was under a duty to disclose the existence of the agreement dispensing with them. I am unable to accept that. In the course of the original negotiations in 1995 condition 1.3 of the expiring wording, which provided that independent surveyors' reports, or, where circumstances demanded, reports from other professionals, should be accepted as conclusive evidence of shipped and delivered quantities, came under review. Mr. Edwards did not like those provisions. His own preference was for a term requiring Glencore to obtain an independent surveyor's report in support of a claim unless prevented from doing so by circumstances beyond its control, in which case evidence from other prescribed sources would be acceptable. In the discussions which followed Glencore through Lloyd Thompson made it clear that although it wanted to retain the conclusive evidence provisions, it was not willing to accept an obligation to produce an independent surveyor's report in support of a shortage claim even where it was possible to do so and the clause remained in its original form. In cross-examination Mr. Edwards recognised that that was the effect of the wording, although he appears to have gained the impression from those discussions that Glencore would appoint independent inspectors whenever possible. Neither Mr. Panzeri nor Mr. Krause thought that, however, and it is not clear what led Mr. Edwards to his understanding. Whatever it was, however, it has not been suggested that it was sufficient to give rise to a representation of any kind. I do not think that Alpina can create some heightened duty of disclosure out of statements too vague to amount to a representation, but in any event it can hardly be said that the possibility that Glencore would not instruct independent inspectors at discharge ports was not made clear.

     (d) Absence of procedures for the verification of quantities in store

149. Alpina say that Glencore failed to disclose the absence of two important procedural safeguards that would ordinarily enable a depositor to monitor the amount of oil held to his order, namely, the issue of holding certificates and the use of independent inspectors to verify the quantity of oil in store.

150. A holding certificate is a document issued by the operator of a storage facility acknowledging that he holds goods to the order of another person, usually the depositor. Such documents are often (though not invariably) issued by operators of oil storage facilities and have an obvious value as a means of evidencing the existence of the deposit. They may also evidence the terms of storage and act as documents of title. The value of such documents is obvious in the case of isolated arrangements for storage where there may be no other record of the deposit or the terms on which it is made, but they may be required in other circumstances as well, particularly where the depositor contemplates a transfer of the oil in store to a third party.

151. The storage agreement covering five cargoes of Iranian straight run fuel oil made between Glencore and MTI in May 1994 did provide for the masters of the storage vessels to confirm by telex to Glencore that they were holding the goods to its order and Mr. Kendrick submitted that there would have been no difficulty in obtaining similar certificates at a later stage. It is true that the May 1994 storage agreement formed the basis on which the parties' relationship subsequently developed, but no holding certificates of the kind contemplated by that agreement were issued after May 1994. That is hardly surprising, however, since the original contract was for certain specific cargoes to be held in segregated storage. Once the parties' relationship developed into new areas and they embarked on commingled storage, certificates of that kind could not have been given. I accept that it would have been possible for MTI to issue regular statements of the quantities of oil held for Glencore, but that would simply have imposed an additional administrative burden from which nothing of any value would have been gained. Not only did the parties have a record of the cargoes discharged into storage in the form of the bills of lading and discharge telexes, but from about June 1994 Mr. Hawkins for Glencore and Miss Gene for MTI kept independent stock records in the form of matching spreadsheets which they exchanged at frequent intervals and used to reconcile their positions. In addition the parties met at roughly quarterly intervals to agree their figures and to establish profits and losses under joint venture arrangements.

152. Mr. Kendrick submitted that the periodic issuing of holding certificates provides three main benefits: it is a safeguard against erroneous misappropriation; it provides the holder with a clear acknowledgment of his entitlement; and it operates in conjunction with independent inspections as a deterrent against theft and as an aid to identifying it if it occurs. These are all perfectly valid points, but they all flow from the same thing: a statement from the storage contractor of what he currently holds for the account of the depositor which the depositor can reconcile with his own records and, if necessary, produce as evidence if a dispute arises. However, a holding certificate is not the only form in which a statement which is sufficient to achieve these ends can be made. Any document unequivocally recording the quantities held by the storage contractor for the account of the depositor will do. Following the agreement to dispense with discharge inspectors MTI sent Glencore telexes confirming the discharge of each vessel which, together with the bill of lading, was a sufficient and contractually binding record of the quantity and grade of cargo received for Glencore's account.

153. At this point it is relevant to mention another aspect of the arrangements between Glencore and MTI which bears on this question, namely, the maintenance and periodic reconciliation of stock records. Before the introduction of JV2 in June 1996 Miss Gene and Mr. Hawkins exchanged inventory records at intervals of no more than three to four weeks and used them as the basis for the regular reconciliation of stocks. There were also quarterly reconciliations of profits and losses which necessarily involved reconciliations of stock movements and formal stock reconciliations took place at the end of 1995 and 1996 to determine the quantities to be carried into the following years. These exchanges of information and regular reconciliations enabled Glencore to know what MTI professed to be holding to its order and while they continued I do not think that the mere absence of holding certificates adversely affected the risk where the parties had put in place other methods of stock control. In this case I cannot see that storage certificates would have added anything of value.

154. JV1 was not entirely superseded by JV2 and the parties continued to operate the established procedures in relation to the relatively small number of cargoes that were handled under it. Mr. Kendrick rightly reminded me, however, that under JV2 the exchange of information dwindled almost to nothing. Quite how quickly that occurred is unclear and Glencore could only be expected to disclose at renewal in June 1997 the position as it stood at that time. There is no evidence that there was a deliberate change in the existing procedures and having regard to the way in which the operational side of the business functioned, I think it unlikely that there was. I doubt, therefore, that the previous practices came to an end overnight, but it is impossible to say quite how and when they died out. There was certainly a formal year-end reconciliation at the end of 1996 and the reconciliation process in January 1998 began with Miss Gene sending Mr. Hawkins an inventory covering only the second half of 1997, so it seems likely that there had already been a reconciliation in respect of the first half of that year. However, there appear to have been no subsequent reconciliations before the end of the year.

155. The utility of holding certificates as a tool for preventing theft should not be over-stated. As all the witnesses recognised, such certificates depend for their value on the honesty of the operator. If the goods are held in segregated storage it is often possible to verify the certificate by independent inspection, but where they are commingled with goods held for other depositors or by the operator in his own right that is not possible without access to the operator's records, including movement records. The same is true, of course of other forms of record keeping such as the spreadsheets kept by Mr. Hawkins and Miss Gene in the present case. Although they could provide a safeguard against error, they could not prevent loss by dishonest misappropriation.

156. Mr. Gooding and Mr. Blaber agreed that a cargo underwriter would not expect to be told about the nature of the documentation issued at any particular storage location, but Mr. Blaber said that he would not have agreed to write this risk unless holding certificates verified by independent inspection were issued for each cargo put into store. Again, I think his views were coloured by approaching the question as if storage cover at Fujairah were being written as a single facultative risk rather than as one element of an open cover. I can see that to place goods into storage without obtaining any documentary evidence from the recipient of the quantity or quality of the goods deposited might be said to exceed the limits of what underwriters would contemplate, but that is not this case. I prefer Mr. Gooding's evidence on this issue. The prudent underwriter will no doubt assume that where goods are put into store the insured will obtain some evidence of the deposit simply in order to safeguard his own position. What that evidence will be will depend on the particular location and the circumstances in which the deposit is made, but that is not a matter with which the underwriter is concerned.

157. For these reasons I am unable to accept that Glencore ought to have disclosed that MTI did not issue holding certificates in respect of cargoes in store at Fujairah.

     (e) The lack of independent inspection

158. The use of independent inspectors to carry out periodic checks on goods held in store is a well-recognised precaution against loss where the goods are segregated from others of a similar kind, although it is not clear that it is a practice universally adopted in the absence of some specific risk of loss. It was accepted by both Mr. Rambaud and Miss Bossley, however, that in the case of commingled storage where the depositor's interest in the bulk is relatively small physical inspection is of little value for the reasons touched on earlier. I am not persuaded that it is general practice to carry out inspections under those circumstances unless there is some specific reason for doing so. Miss Bossley considered, however, that where the interest of one depositor is so large that he owns almost the whole of the bulk the position can be assimilated to segregated storage so that checks by independent inspectors serve a useful purpose. I agree that the larger the interest of a depositor, the greater the chances that an inspection will reveal a shortage that affects his interest, but it will not be possible for even a large depositor to calculate the size of his interest without knowing what others have contributed to and drawn from the bulk. Moreover, the position may change from day to day as further deposits and drawings are made.

159. Mr. Kendrick submitted that it would have been normal practice for Glencore as the largest, or in some cases the sole, depositor of each grade of oil to have appointed independent inspectors to carry out spot checks on a regular basis, both as a deterrent to theft and as a means of identifying it if it occurred. Again, however, there is an element of hindsight attaching to this argument. Of course, it is now possible to see that Glencore did hold a preponderant interest in most grades of oil stored at Fujairah and the sole interest in some grades such as cutter stock, but it is not clear that anyone at Glencore was aware of that at the time. In any event, there is a danger of approaching the question from the wrong end. Alpina's complaint is not that Glencore failed to take proper care to protect its property but that it failed to disclose material aspects of its relationship with MTI when renewing the cover in 1996 and 1997. Alpina has not sought to show that Glencore was aware at that time that during the policy year it was likely to have the sole, or even the preponderant, interest in the bulk of any grade of oil held at Fujairah for a significant period of time and that in that knowledge it would not seek to appoint inspectors to carry out spot checks. Mr. Kendrick's argument applies, of course, to all cases in which oil is held in commingled storage and Mr. Blaber said that it was normal practice even in the case of commingled storage for depositors to send inspectors to verify their holdings. However, it became clear that what he had in mind was gauging of the tanks before further oil was pumped into them, which is a very different matter. In the circumstances I am unable to accept that Glencore's failure to instruct independent inspectors to measure the quantity of oil held in storage significantly affected the risk or that it was something that a reasonable underwriter could not be expected to have in contemplation when writing this kind of cover. Accordingly, Glencore was not in breach of duty in failing to disclose it.

     (f) MTI's choice of storage vessels

160. The agreement between Glencore and MTI left MTI free to choose which vessels to employ as floating storage units; it also gave MTI complete operational control over the disposition of the oil within the facility. Mr. Kendrick submitted that that was a matter that ought to have been disclosed to Alpina.

161. Miss Bossley did not consider it unusual that MTI should be given the right to decide how oil was disposed around the facility as a whole or that masters of incoming vessels were told to discharge in accordance with MTI's instructions. She did say, however, that she found it surprising that Glencore did not inform itself of the identity of the vessels comprising the facility at all times, primarily because of the risk of an unauthorised ship-to-ship transfer from an arriving vessel to an exporting vessel. She was also concerned that Glencore might run a greater risk of incurring liability for pollution if it did not have the opportunity to approve vessels chartered for floating storage. Mr. Rambaud recognised that in some cases a potential depositor might need to check the characteristics of the vessels making up a floating storage facility for insurance or finance purposes, but subject to that he did not think that a depositor would be interested in how his oil was disposed around a facility of this kind.

162. If oil is put into commingled storage it will be possible to identify the bulk of which it then forms part. In the case of land storage that could be a single tank or a group of tanks, perhaps varying from time to time in size and number as demand for storage space varies. In the case of floating storage the bulk might be held in a single tank, a group of tanks on one vessel or a group of tanks on several vessels, again depending on the demand for space. What constitutes the bulk in any given case is essentially a question of fact. The precise destination of his original parcel, however, cannot matter to the depositor because once it has joined the bulk it loses its identity and he acquires a proportionate interest in the bulk. All this is inherent in commingled storage and I can therefore see no basis on which an underwriter of an open cover which includes storage risks could expect the insured to have a right to determine into which tank his particular parcel is pumped. In any case this is essentially an operational matter over which the depositor could not expect to exercise control.

163. The suggestion that a depositor should be in a position to approve the vessels making up a floating storage facility before deciding to place oil in storage raises rather different questions. One can see why marine underwriters in particular might wish to approve the ships being used to store the goods, being familiar with the need to have regard to a vessel's age and other characteristics when assessing marine perils, but the usual way of dealing with this aspect of the risk in an open cover is by inserting suitable terms in the policy to keep the risk within acceptable limits. There is no evidence at all that underwriters writing this class of business expect the insured to assess vessels applying their own standards rather than those which the underwriters themselves would adopt. In any event, it makes no sense as a practical proposition in the case of commingled storage to suggest that the insured should approve all the vessels comprising the facility, whatever standards he might apply. For example, if the bulk were distributed across several vessels, it would be impossible for a potential depositor to stipulate that his oil should not be put on board one of those vessels without obliging the operator to exclude it from service for that grade altogether. The only alternative would be for the depositor to avoid using that facility. An even greater difficulty would arise if the operator proposed to enlarge the capacity of the facility by bringing in an additional vessel to which an existing depositor objected. Giving depositors control of that kind over the choice of vessels making up a floating storage facility would be commercially unworkable. In my view it would be obvious to an underwriter familiar with the trade that an individual depositor placing oil in commingled floating storage could not expect to have any control over the choice of vessels used to make up the facility any more than he could expect to control the choice of tanks used for commingled storage in a land based tank farm. Accordingly it is not a matter that Glencore was required to disclose.

     (g) The relationship as a whole

164. For the reasons given above I do not think that any of the individual matters of which Alpina complain were so unusual that they fell outside the range of possibilities that a prudent underwriter writing this policy could be expected to have taken into account. Mr. Kendrick submitted, however, that taken together these various aspects of the way in which Glencore did business with MTI involved such an unusual combination of features that they ought to have been drawn to Alpina's attention: see C.T.I. v Oceanus [1984] 1 Lloyd's Rep. 476 per Parker L.J. at page 516.

165. The real thrust of Mr. Kendrick's argument is that viewed overall the nature of this enterprise was so unusual that it could not be regarded as falling within the range of an ordinary oil trader's business. The oil industry, as is well known, is dominated by the multi-national companies whose interests include the whole range of activities from the production of crude oil to the retailing of refined products. Traders such as Glencore therefore depend for their existence on exploiting new opportunities as they appear. Many of these are short-term and those which are not usually involve some elements of a kind that the major oil companies do not consider attractive. In the present case it was MTI's need for financial support in obtaining cargoes for blending as well as Glencore's established position as a trader of oil and products that provided Glencore with the opportunity to become involved in the JV1 arrangements. It is inevitable that ventures of that kind will often involve relatively small enterprises whose management is readily accessible and who will wish to keep administrative procedures to a minimum. One might expect, therefore, that if operations are conducted successfully over a period of time a relationship of trust will develop which will encourage both parties to reduce administrative burdens to a minimum. That is exactly what happened in the present case.

166. Again, when considering this submission it is necessary to remind oneself that the question is not whether Glencore could or even should have done more to protect its position, but whether in March 1996 or June 1997 the nature of its arrangements with MTI were so unusual that their characteristics ought to have been disclosed to Alpina because they fell outside the range of ventures that a prudent underwriter could be expected to have had in mind.

167. Miss Bossley considered that, although viewed individually those features of the arrangement which she criticised were understandable, their cumulative effect was such that the arrangements as a whole failed to provide the protections against loss that she would have expected a normal trading company to require. This may be a fair criticism, but it does not really address the question whether, taken as a whole, the arrangements between Glencore and MTI fell outside the range of what might be expected. Mr. Kendrick submitted that they did and that the deficiencies in Glencore's procedures were highlighted by its inability to ascertain with reasonable ease the quantity of oil held to its order at any given moment. Mr. Sumption submitted that the argument was ill-founded in the present case, however, because the various individual safeguards whose absence Mr. Kendrick relied on were not different ways of achieving the same result (so that the presence of one would compensate for the absence of another), but were directed to very different ends. Thus, discharge inspectors are instructed to identify short delivery; holding certificates are issued as proof of receipt or to reflect a subsequent transfer of the goods and to enable the holder to deal with them; tank inspections are directed to verifying the physical presence of the goods; the maintaining and reconciliation of inventories is a means of correcting errors in the recording of movements of oil; and the vetting of ships is aimed at avoiding ordinary marine perils. Accordingly, the fact that one of these had been dispensed with does not necessarily strengthen the case for retaining one or more of the others.

168. I think Mr. Sumption's argument has a good deal of force in a case where, as here, none of the characteristics of the arrangements relied on by Mr. Kendrick can be regarded as extraordinary and where none of them can therefore be said to increase the risk of loss in a way that the insurers could not be expected to have had in mind. If the underwriter is to be taken to have been aware that individual safeguards might be dispensed with, he can also be taken to have been aware that any or all of them may have been dispensed with in any given instance. Quite apart from that, however, some of the individual factors are in my view more relevant to this argument than others. For example, I do not think that much significance can be attached to the inability of Glencore to vet ships or to the parties' decision to dispense with discharge inspectors. Of more potential significance, however, are the lack of holding certificates, the absence of periodic tank inspections and the failure to carry out frequent reconciliations, but in truth none of these operates in combination with any of the others in such a way as to enhance the risk of loss significantly. It is quite true that following the collapse of MTI it has proved difficult for Glencore to identify with precision the amount of oil held to its order at any given time, but the difficulty, as I understand it, arises more from certain fundamental aspects of the arrangements which are not the subject of complaint (principally commingling in storage and the release of oil on the ITT date) than from any of the deficiencies Alpina has relied on. Thus the problem lies not so much in identifying what quantities or grades of oil Glencore delivered into store at Fujairah, but in ascertaining exactly how particular parcels were subsequently disposed of in the production of blended products and whether MTI was entitled to make use of them at the time. That requires a detailed understanding of, among other things, MTI's internal processes which is only available (if at all) from its records of oil movements. It also requires in some instances information about the circumstances affecting the interests of other parties with whom MTI had been trading. However, none of these has anything to do with those aspects of the relationship of which Alpina says it was not warned.

169. For these reasons I am unable to accept that Glencore was bound to disclose the nature of its relationship with MTI.

     (h) Inducement

170. In these circumstances the question of inducement does not arise and I do not think that anything is to be gained by debating at length what effect these matters would have had on Mr. Panzeri's mind if they had been drawn to his attention. However, it is perhaps fair to note that although he said in chief that they would have led him to impose different terms, he did not really maintain that position in cross-examination. The practice of Alpina, which may reflect that of Swiss insurers generally, was to engage in a far greater degree of direct contact with the insured than would be normal in the London market. He described it as a 'triangle' philosophy. I think it is likely, therefore, that if Glencore's relationship with MTI had been brought up in the discussions with the brokers Mr. Panzeri would have wanted to know more about it so that Alpina could discuss directly with Glencore potential pitfalls and the ways in which they could be avoided. However, I am not persuaded that at the end of the day he would have insisted on imposing any different terms in the face of objection by Glencore.

     3. Blending

171. Alpina contended that Glencore failed to disclose that oil to which it had title or which was at its risk might be blended while in storage with MTI and positively misrepresented that cargoes discharged into the custody of MTI at Fujairah were not subject to any risk from blending.

172. When Mr. Vanderson asked in April 1995 for cover to be extended to oil in floating storage at Fujairah he referred to the previous cover with Lloyd's and attached to his fax copies of the relevant endorsements from which it could be seen that cover had been given in accordance with the terms of what later became condition 1.18 (then condition 1.19). He then continued

     "This is purely a storage risk exposure and there are no blending risks involved."

     Alpina submitted that that amounted to a representation that oil in store at Fujairah was not subject to any blending operations and that in the absence of any later statement to the contrary it was impliedly repeated when the cover was renewed in 1996 and 1997.

173. In my judgment at Phase 2 I made a number of findings about blending operations at Fujairah. They included a finding that during 1995 and the first half of 1996 while JV1 was in operation MTI produced parcels of blended products with Glencore's approval using stocks held to its order. In some cases Glencore sold blended cargoes to third parties on behalf of the joint venture and in many cases MTI sold parcels of bunkers on behalf of the joint venture. In some instances MTI produced blended parcels for Glencore using Glencore's own components to enable it to make sales to a third party. In some cases blended parcels were produced for sale by Glencore to MTI. These were recorded by Mr. Hawkins in his spreadsheet as drawings from the stocks of the relevant components to produce the blended product and a sale of the blended product by the joint venture to MTI. I found that in these cases Mr. Hawkins was aware that the blends had already been produced by the time he received the relevant advice from Miss Gene and I rejected the suggestion that these were in reality sales of the components rather than the resulting blend. I also found, however, that MTI had always obtained Glencore's authority to make the blend in question. At a later stage during the operation of JV2 MTI began drawing from Glencore's stocks to produce blended products for its own use without Glencore's consent, although Glencore was not aware of that.

174. The critical question it seems to me is what was meant by "blending risks" in Mr. Vanderson's fax. Mr. Edwards, the underwriter at Alpina who initially dealt with the matter, said that as a result of an earlier telephone conversation with Mr. Divine he had understood that part of the message as confirming that no blending of oil was being carried out at Fujairah. However, Mr. Edwards was about to travel abroad when Mr. Vanderson's fax came in, so he passed it to Mr. Krause, who took it to Mr. Panzeri for a decision. It is not clear whether Mr. Krause formed any independent view of the matter at the time, but in any event it was Mr. Panzeri who had to make the underwriting decision and his evidence that he understood the risk to be a pure storage risk without any blending was not seriously challenged.

175. In the event, when Alpina agreed to extend cover to oil in floating storage at Fujairah it did so expressly in accordance with condition 1.18, as Lloyd's underwriters had done before it. Condition 1.18 covers physical loss of or damage to oil in the course of storage and blending, although losses caused by faulty blending are expressly excluded. Mr. Sumption submitted that since Alpina agreed to accept the risks of blending, Glencore was under no duty to disclose the fact that blending might take place. For that reason alone, he submitted, the case based on non-disclosure must fail. He also submitted, however, that the facts were not material and would not have made any difference to Mr. Panzeri's decision to grant cover on the terms of the indorsement.

176. If Mr. Vanderson had said nothing either way in his telex about blending at Fujairah, I do not think that it could possibly have been said that Glencore had failed to disclose any material facts. One would then simply have had a request for cover on the terms of condition 1.18, including blending, which Alpina could have considered on its merits. Insofar as the argument is put on the basis of non-disclosure, therefore, it must fail. The argument based on misrepresentation raises different issues, however, because if a positive statement was made that there was no blending of Glencore's oil, that might make an underwriter less concerned about extending the relatively broad form of cover provided under condition 1.18 to that particular location.

177. Mr. Sumption submitted that on a fair reading of Mr. Vanderson's fax no statement of that kind was made. He argued that the words in question merely emphasised that Glencore was not seeking cover against the risk of faulty blending as opposed to the risk of physical loss in storage. I find that difficult to accept, however, because there was no need to say anything about blending in view of the terms of condition 1.18. In the context in which it was written Mr. Vanderson's comment only makes sense as a positive statement that Glencore did not have reason to believe that there would be any blending of oil held by MTI to its order at Fujairah. That was not the case, however, either in April 1995 or March 1996, or indeed in June 1997. In April 1995 Glencore did expect to lift cargoes of blended products from the joint venture from time to time for the purposes of sale to third parties and knew that these would be derived from oil held to its order. By March 1996 the practice of producing blended products for sale to MTI which I described in paragraphs 107-115 of the Phase 2 judgment had become established and was no doubt expected to continue, although no doubt as occasional rather than frequent events. By June 1997 JV1 had been largely superseded by JV2, but from time to time MTI still continued to produce blended cargoes for Glencore to enable it to supply its own buyers, mainly in Pakistan.

178. Mr. Sumption submitted that even if there was a misrepresentation in relation to blending, it was not material in the light of the fact that Alpina did actually agree to cover losses sustained in the course of blending. I can see the attraction in that argument, but for the reasons touched on earlier, I think a distinction is to be drawn between saying nothing and making a positive statement that no blending was expected to be carried out. Mr. Gooding was of the view that since losses caused by faulty blending were excluded, the statement that no blending was expected to take place was immaterial. Mr. Blaber thought that the statement was material because of the concern which he thought a prudent underwriter would have about the consequences of carrying out blending operations in the facility where the oil was being stored, in particular as a result of the additional handling involved.

179. I found many of Mr. Blaber's grounds of concern about blending operations unconvincing. For example, I fail to see how, as he suggested, the risk of a loss to the insured is increased by the fact that oil belonging to other depositors is drawn from a commingled bulk for blending rather than for any other reason. However, it is unnecessary to discuss any of them in detail because Alpina's complaint relates only to the blending of oil in which Glencore retained an interest. The question, therefore, is whether a statement that none of Glencore's oil was expected to be blended at Fujairah was material. The test of materiality in relation to misrepresentation is the same as for non-disclosure, namely, whether the statement in question would influence the judgment of a prudent underwriter in fixing the premium or deciding whether to accept the risk. I am unable to say that the prudent underwriter would have entirely disregarded a statement of this kind when deciding whether to extend the cover, because I think it possible that the increased handling involved in the blending process might be seen as enhancing the risk of physical loss.

180. I turn, therefore, to the question of inducement. There was a tendency on Mr. Kendrick's part to approach the question of inducement by considering how Alpina would have responded if in his fax of 3^rd April 1995 Mr. Vanderson had given a description of blending operations at Fujairah. That may be appropriate when the issue before the court is one of non-disclosure, but is not appropriate in the case of misrepresentation because the court is then concerned only with the effect the false statement had on the underwriter's mind. In the present case I think there is real doubt whether Mr. Panzeri or Mr. Krause had Mr. Vanderson's fax in mind at all when they considered the renewal of the contract in 1996. Neither of them had reminded himself in any detail of the correspondence passing between Alpina and the brokers in March and April 1995 and neither of them said that he had in mind the terms of Mr. Vanderson's fax. But quite apart from that, having heard Mr. Panzeri give evidence I am far from persuaded that the statement made any difference to his perception of the risk when he first decided to extend the cover. The fact is that he was willing to follow Lloyd's underwriters by extending cover on the terms of condition 1.18 at a commensurate premium. I am confident that if this statement had not been made and nothing at all had been said about the likelihood of blending his decision would have been exactly the same. Nothing would have been said to Alpina's co-insurers, of course, who did not expect to be informed of this kind of extension to the cover in any event. In these circumstances the insurers are not entitled to avoid the contract on this ground.

181. For the reasons given above I am satisfied that none of the grounds on which Alpina sought to avoid the open cover have been established. It follows, therefore, that Alpina and its co-insurers are not entitled to avoid the open cover for either the 1996-97 or the 1997-98 year.

     V. Coverage issues

182. A number of issues relating to the construction and application of the policy terms fall to be decided at this stage in the litigation; others will have to await a later phase. The main issues that arise for determination now relate to the location of loss, the application of the policy to cargoes that were diverted without being discharged at Fujairah or which were discharged directly into exporting vessels, the misappropriation of oil that was subsequently the subject of an ITT contract, the shipment of various cargoes from Fujairah between 9^th and 13^th February 1998 following the collapse of the Metro group and Glencore's claim in respect of refined products that it says ought to have been held to its order by MOC.

     1. The location of loss

183. A number of points relating to the construction of condition 1.29 can conveniently be dealt with at this stage. The first concerns the location of loss for the purposes of the open cover. Glencore accepted that any claim for the loss of oil held in the floating storage facility must be brought within the terms of condition 1.29. Specific findings about the manner in which individual losses occurred cannot be made at this stage, but it is likely that they all occurred in one of two ways: either oil was drawn from the bulk for blending and was lost when it was incorporated into the new product, or oil was drawn from the bulk for sale and delivery to third parties and was lost when delivery was made.

184. The cover provided under condition 1.29 extends only to oil held on the Metrotank and the Mount Athos and to oil held by way of incidental storage in shore tanks or a final supplying vessel pending delivery to the Metrotank. (The Mount Athos was not used as a storage vessel after 1995 and can be disregarded for these purposes.) During the latter part of 1997 and early 1998 a number of other vessels were used to provide additional storage capacity, including the Violet, the Nafkratis, the Athenian Splendour and the Sea Giant (see paragraph 144 of the Phase 2 judgment), and one consequence of commingled storage in circumstances where the bulk comprised the whole quantity of any given grade is that Glencore was almost certainly interested in the contents of all those vessels. On that basis an unauthorised disposal of oil drawn from any of them will have resulted in a loss to Glencore, but Alpina is only liable for Glencore's proportion of any misappropriation from the Metrotank or an incidental storage vessel or land tank. From time to time MTI may have had stocks of its own on which it was entitled to draw and may also have been entitled to draw from Glencore's stock as a result of the issue of an ITT contract. Though I think it unlikely, I do not think that one can at this stage exclude the possibility that MTI may have been entitled to draw from stocks held for other depositors. I use the word "unauthorised", therefore, to mean any drawing not justified in one or other of those ways.

     2. "Final supplying vessels"

185. Next there is the meaning of the expression "final supplying vessels". Since for the purposes of the policy the floating storage facility comprised only the Metrotank and the Mount Athos, the expression "final supplying vessels" must be construed as referring only to vessels carrying oil intended for discharge into the Metrotank. That much was not in dispute.

186. Mr. Sumption submitted that it was irrelevant that the oil had previously been stored in a vessel not falling within condition 1.29, so that a vessel transferring oil from one of the other dedicated storage vessels to the Metrotank would be a "final supplying vessel" for these purposes, but that does not seem to me to reflect the true meaning and intention of the words used. The primary purpose of condition 1.29 was to provide cover in respect of oil held in floating storage on board one of the named vessels. That cover was then extended to oil held in incidental storage pending delivery into one of the named storage vessels. The purpose of that must have been to ensure that cargoes coming into the facility were covered while temporarily on board another vessel awaiting transfer to the Metrotank as their final destination. Cargoes discharged into other dedicated storage vessels reached their final destination for the purposes of storage at that point. A vessel used to transfer cargo from one of the other dedicated storage vessels to the Metrotank was not performing the function of incidental storage between the ocean carrier and the Metrotank; in effect it was performing a carrying voyage between different parts of the facility and its function was more analogous to that of an ocean carrier delivering cargo to the facility. In my view that is not what condition 1.29 has in mind.

     3. Incidental storage

187. There was also disagreement between the parties about the meaning of the phrase "incidental storage" in condition 1.29. The extension of cover to oil on board vessels in which incoming cargoes were held temporarily pending delivery to the permanent storage vessels was first granted by an indorsement dated 11^th January 1996 which included the following passage:

     ". . . . Assured advise for clarification purposes that where it is not possible for interest to remain on board the delivering vessel prior to ultimate delivery it may be discharged onto either MT "EVOIKOS, "FAY" or SHERVAN" for incidental (i.e. usually no more than one week) storage (where Assured may acquire interest) prior to final delivery to the dedicated floating storage vessels;

     which noted and agreed that the latter three vessels be included under the coverage afforded by endorsement dated 18^th October, 1995 in their capacity as incidental floating storage vessels."

188. The expression "usually no more than one week" was not reproduced when the endorsements relating to Fujairah were replaced by condition 1.29 in the 1996 policy, but Mr. Kendrick submitted that the expression "incidental storage" must be construed as limited to one week because the parties had, in effect, defined its meaning in the original indorsement. I am unable to accept that argument for a number of reasons. First, the indorsement itself falls into two parts: the first contains a description of the circumstances which have given rise to the request for the extension of cover; the second contains the substantive agreement. The expression contained in the paragraph containing the information is "usually no more than one week". That obviously suggests that in some instances oil may remain on board one of the incidental storage vessels for longer than a week. In those circumstances I do not see how it can be suggested that the parties were somehow agreeing to treat the concept of incidental storage as limited to one week and it is to be noted that the insurers agreed to extend cover to the three named vessels in their capacity as incidental floating storage vessels without limiting the period of cover in respect of any particular cargo. It is at least doubtful, therefore, whether the cover granted by the original indorsement was limited in duration as Mr. Kendrick suggested.

189. Secondly, the various endorsements relating to floating storage at Fujairah were superseded by condition 1.29. The condition is drafted in language similar to that of the earlier endorsements, but there are differences and one can see that in certain respects the cover it affords is wider. I can see no basis for reading a limitation into the clause by reference to what had gone before, if only because there is no reason to assume that the parties were not content to extend the scope of the clause in that respect. In my judgment cover in respect of cargoes on board final supplying vessels continued until they were discharged into the Metrotank or until a decision was taken by MTI to discharge them into one of the other dedicated storage vessels, at which point they ceased to be final supplying vessels within the meaning of condition 1.29.

     4. Oil on board the 'Metrotank'

190. The Metrotank had been specially modified and equipped to carry out blending operations and it is therefore likely that a large proportion (Mr. Sumption suggested it might be as much as 70%) of the blending carried out by MTI occurred on board that vessel. It is also likely that in very many cases the constituents from which the blends were produced were held on board for some period of time before they were used. The question therefore arises whether that oil was insured at the time of loss regardless of the fact that it may have spent most of its time at Fujairah on board one of the other dedicated storage vessels.

191. Condition 1.29 provided cover to oil on board the Metrotank in accordance with condition 1.18 which extended not just to oil in storage but to oil in the course of blending as well, though it excluded loss and damage caused by faulty blending. I do not think it could be said, therefore, that the movement of oil between tanks on the Metrotank in preparation for blending somehow took it outside the terms of cover. Accordingly, if oil was stored on the Metrotank, it did not cease to be covered at the point when it was moved into a working tank in preparation for blending. The position is less clear, however, in the case of parcels of oil brought onto the Metrotank from one of the other dedicated storage vessels for the sole purposes of blending. Mr. Kendrick submitted that they fell outside the cover provided by condition 1.29 because the relevant misappropriation occurred when the parcel was drawn from the original storage vessel.

192. I find this argument difficult to accept. Although there may well have been a misappropriation of some kind when MTI drew oil from a dedicated storage vessel with a view to blending it, the insurance is against the risk of loss and I do not think that a loss occurred at that point within the meaning of the policy. The oil was still in MTI's possession, it retained its original character, was undamaged and was not irretrievably lost to Glencore because MTI could still have changed its mind and retained it in store. Leaving aside losses by way of destruction or leakage, a loss occurred only when the oil was blended, thereby losing its identity, or when it was delivered by MTI to a third party. The only basis on which it might be said that oil on board the Metrotank was not covered is that condition 1.29 only extends cover to oil held in storage and that oil transferred to the vessel for the sole purpose of blending was no longer being held in store. The difficulty with that argument, however, is that it is hard to draw a satisfactory line between short-term storage and holding for the purposes of blending. In my view all the oil that entered the facility remained in store somewhere until it was disposed of in one of the ways mentioned earlier. Whether Glencore's interest was insured depends on where the oil was at any given time. In my view any oil that was brought on board the Metrotank fell within the scope of the cover provided by condition 1.29.

     5. Diverted cargoes

193. On a number of occasions oil belonging to Glencore that had been carried into Fujairah for discharge into the floating storage facility was diverted by MTI to other destinations. In some cases whole cargoes were diverted without being discharged; in others part of the incoming cargo was transhipped into another vessel and carried on to a new destination. In some of the latter cases the incoming vessel discharged cargo directly into the exporting vessel; in others the incoming vessel and the exporting vessel were both moored alongside the Metrotank and the cargo was transhipped using her pumps and lines.

194. Two issues arose in relation to these cargoes: whether they had reached their final destination within the meaning of condition 1.15 of the policy before the diversion occurred; and whether the diversion gave rise to a loss in those cases where after the diversion MTI still retained in store enough oil of the grade in question to satisfy the entitlements of all other interested parties, including, in the case of Glencore, its entitlement in respect of the quantity diverted.

195. Condition 1.15 provided as follows:

     "Cover to attach from the time the Assured becomes at risk or assumes interest and continues . . . . until finally delivered to final destination . . . ."

196. Mr. Sumption submitted that in all of these cases there had been a loss under condition 1.15 of the policy for the simple reason that MTI had misappropriated goods belonging to Glencore before they had entered the storage facility. Mr. Kendrick, on the other hand, submitted that MTI was entitled to dispose of cargoes brought into Fujairah for storage because of the wide scope of its authority to manage the facility as a whole.

197. The practice was for Glencore to instruct the master of the carrying vessel to proceed to Fujairah, give notice of readiness and discharge in accordance with MTI's instructions. Since MTI had operational control of the facility, that is not surprising. Mr. Kendrick submitted, however, that in reality all these cargoes had reached their final destination when the vessels arrived at Fujairah and gave notice of readiness to discharge. Thereafter they were fully under the control of MTI with Glencore's consent and had entered the storage facility. By the same token they had been "finally delivered to final destination" within the meaning of condition 1.15.

198. At first sight it seems obvious that the cargoes which were diverted without being discharged had not been delivered to their final destination and were misappropriated by MTI. The mere fact that the vessel had given notice of readiness and had been placed under the operational control of MTI was not sufficient to terminate the carrying voyage; nor for that matter was it sufficient to terminate the insured transit since by virtue of the American Institute Cargo Clauses which were incorporated in the policy insured that continued until delivery to the final warehouse or other place of storage. Orders to the carrying vessels to discharge in accordance with MTI's instructions amounted to nothing more than orders to discharge into the storage facility at the particular location indicated by MTI.

199. Mr. Kendrick submitted, however, that it was open to MTI to transform an incoming vessel into a temporary storage vessel, thereby completing the delivery of the cargo into the facility in accordance with Glencore's instructions. The effect of that, he submitted, was that the transit ended before there had been any dealing with the cargo. No doubt one advantage of a floating storage facility is that its capacity can be increased relatively easily by incorporating additional tonnage, but I do not think that is what in fact occurred in any of these cases. In all of them MTI went to some lengths to disguise from Glencore the fact that the vessel had not discharged and the vessel never was in any practical sense incorporated into the facility.

200. As an alternative argument Mr. Kendrick submitted that if MTI was already holding enough oil of the grade in question for its own account to enable it to make the delivery from its own stocks, there was no misappropriation of any kind. All that MTI was doing was giving directions for discharging and for the handling of oil within the facility that were within the scope of its operational discretion. Mr. Sumption submitted that this might be a relevant consideration in cases where the oil had actually entered the storage facility but was not relevant if it was diverted before it had done so, as was the case here. He also submitted that this argument was not open to Alpina because it had already been decided in Phase 2 that all the diverted cargoes had been converted by MTI.

201. I do not think that it is quite correct to say that this matter was finally determined in Phase 2. At the Phase 2 trial MTI argued that under both JV1 and JV2 it was entitled to dispose of oil shipped into Fujairah by Glencore from the moment of its arrival. In that context it became necessary to consider the circumstances surrounding the diversion of cargoes and the issue of discharge telexes to see to what extent they shed any light on the matter. The picture that emerged was less stark than might have been expected (see paragraph 125) and although I was quite satisfied that MTI had sent Glencore false discharge telexes on a number of occasions, I was not persuaded that it did so because in every case it was aware that it had no right to dispose of those cargoes in question (paragraph 126). I did hold, however, that MTI had no general right to dispose of cargoes as soon as they reached Fujairah and it follows that in principle the diversion of any cargo, either in whole or part, before it had entered the storage facility amounted to a conversion. It was not necessary, however, for me to decide whether Glencore had agreed to any particular disposal, either expressly or by implication, and I made no findings in relation to specific cargoes other than those in paragraph 125 relating to the cargoes on the Epic and the Shoko. In these circumstances I do not think that Alpina is precluded from contending that specific cargoes were disposed of by agreement with Glencore.

202. The fact that MTI had control over the conduct of operations at Fujairah does not mean that all the steps that it took in relation to the movement of cargoes can be assumed to have been carried out in furtherance of its agreements with Glencore and the other depositors. However, when MTI was holding sufficient stock for its own account to enable it to make a delivery, I can see that in some circumstances operational considerations might make it attractive to treat an incoming cargo as part of the existing stock before it had been physically discharged into one of the dedicated or incidental storage vessels. However, I think that Mr. Sumption is right in saying that even under those circumstances the diversion of cargo would amount to a misappropriation unless it could be shown that Glencore had consented to it. Each of these incidents may turn on its own facts, so it is impossible to reach any final conclusion at this stage, but as far as I am aware Glencore was not specifically asked by MTI to approve any such course of action. In each case, however, Glencore was credited in inventory with the full amount of the incoming cargo and it may therefore be possible to infer that MTI intended to, and did, transfer to Glencore an amount of oil from its own stock equivalent to that which had been diverted. A transfer of that kind could not be made without Glencore's consent, of course, and would depend on MTI's having sufficient stock at its own disposal to support it, but by giving MTI full operational control over the facility and the handling of incoming cargoes I think Glencore impliedly assented to any transfers of title necessary to reflect operational movements of that kind. Accordingly, subject to further consideration of the particular circumstances surrounding each of these transactions, it may be possible to infer that there was a transfer by MTI to Glencore of title to an equivalent volume of oil in store and therefore no loss at that point.

203. I do not think that the use of the Metrotank to assist the transhipment of cargoes between incoming vessels and exporting vessels raises any question of principle. If, as appears to have happened on some occasions, the Metrotank's pumps and lines were used to transfer cargo between two vessels moored alongside, I do not think that the cargo can properly be said to have entered the storage facility at all. The same would apply if the cargo had been pumped through one of the vessel's tanks. If, on the other hand, cargo was put into one of the tanks even for a limited period pending other operations, the position might be different. That is a question which is better considered at a later stage when there may be more information before the court about the circumstances in which individual movements occurred.

     6. Oil misappropriated prior to ITT – an uninsured credit risk?

204. In its points of defence Alpina has pleaded that any loss suffered by Glencore was caused by MTI's failure to pay for the oil in store at Fujairah, not by any loss of the oil itself. Put another way, it said that Glencore has suffered the consequences of an uninsured credit risk.

205. In the event this argument was not pursued in its full width, quite rightly in my view. Title to oil at Fujairah remained in Glencore until it was transferred to MTI under or pursuant to a contract of sale in the form of an ITT contract. Until that occurred Glencore enjoyed the whole of the legal and beneficial interest in the oil and could therefore insure it for its full value. The argument did survive, however, in relation to certain quantities of oil which were disposed of by MTI before any ITT contract had been entered into in respect of them but which Glencore subsequently agreed to sell in ignorance of the fact that the oil had already been taken. At a later date Glencore obtained judgment against MTI for the price of the oil due under the various ITT contracts, but the judgment has not been satisfied.

206. Mr. Kendrick submitted that in these circumstances Glencore could not recover substantial damages in conversion from MTI and cannot recover from Alpina in respect of a loss of the goods themselves. However, as Mr. Sumption pointed out, in each case insured property was lost, either through being incorporated into a blend or through being delivered to a third party, and in each case the loss was caused by an insured peril, namely, misappropriation on the part of MTI. He submitted that at that point Glencore became entitled to recover the value of the goods from the insurers. If the disposal of the goods was a wrongful act when it was carried out, it could not thereafter cease to be one unless Glencore subsequently ratified it or otherwise agreed not to treat it as wrongful, neither of which had occurred in this case. He accepted that a loss by an insured peril may be made good by subsequent events so as to release the insurer from liability, but he submitted that unless the loss is made good the insurer remains liable for it. In the present case payment by MTI under the ITT contract would have made good the loss, but its failure to pay for the goods did not cause the loss.

207. Two points need to be mentioned by way of introduction. The first is that although the argument proceeds on the assumption that a quantity of oil previously disposed of by MTI can be related to a subsequent ITT contract, it may in fact be difficult to relate one to the other, especially while sufficient stocks remain to satisfy new contracts. That is simply a consequence of the fact that each ITT contract (except one that covered the final parcel of any grade) was for the sale of part of an undivided bulk. The position is therefore not quite as simple as that which would arise if specific goods were stolen and later sold by the owner to the thief. However, no attempt has been made at this stage to grapple with issues relating to the identity of the goods covered by any particular ITT contract which will have to be resolved, if necessary, at a later trial. The second point is that it is important to bear in mind that the insurance is against the risk of physical loss and damage, not against conduct that constitutes a conversion in law.

208. In support of his argument Mr. Kendrick relied heavily on the decision of the Court of Appeal in Hiort v London and North Western Railway Co. (1879) 4 Ex. D. 188. In that case the plaintiffs stored grain in bulk with the defendant and issued delivery orders for the release of parcels as and when necessary. Grimmett, an agent of the plaintiffs, obtained delivery of 60 quarters from the defendants against his undertaking to forward a delivery order from the plaintiffs in due course. Another merchant, Tarpler, agreed to buy 60 quarters of grain from the plaintiffs who gave him a delivery order for that quantity which he then indorsed in favour of Grimmett. Grimmett gave the delivery to the defendants in fulfilment of his earlier undertaking. Tarpler was unable to pay the plaintiffs for the grain he had bought and Grimmett failed to account to them for the grain he had obtained. The plaintiffs made a claim against the defendants for conversion of the goods delivered to Grimmett. The Court of Appeal held by a majority (Bramwell and Thesiger L.JJ.) that they were entitled to recover nominal damages only.

209. The two Lords Justices representing the majority reached their conclusions by rather different routes which makes it difficult to extract a single clear principle from the decision. Bramwell L.J. accepted that there had been a conversion of the goods when the defendants delivered them to third parties at Grimmett's request without the production of a delivery order and did not think that it could be purged by what subsequently occurred. However, he regarded Grimmett's action in handing over the delivery order to the defendants as equivalent to returning the goods so that the plaintiffs suffered no loss as a result of the original conversion. The plaintiffs were entitled to recover the price from Tarpler because the goods had been delivered to Grimmett in circumstances that involved the satisfaction of the delivery order.

210. Thesiger L.J. also held that the original delivery at Grimmett's request was a wrongful act in respect of which a right of action vested in the plaintiffs and did not cease to be a wrongful act as a result of anything that subsequently occurred. However, he considered that no damages flowed from it because if there had been no earlier misdelivery the defendants would nonetheless have been bound to deliver the goods to Grimmett's order to satisfy the delivery order indorsed to him by Tarpler. He therefore concluded that the damage was not caused by the original misdelivery but by the sale to Tarpler and his failure to pay for the goods. He also considered that although the plaintiffs had the property in and right to possession of the goods at the time of the misdelivery, by the time they brought the action they had transferred all their rights to Tarpler. They could not recover the full value of the goods because they could no longer transfer title to the defendants on receiving satisfaction of a judgment.

211. Baggallay L.J. dissented on the grounds that since the defendants were bound to deliver the grain to Grimmett but had already done so, the plaintiffs had suffered no loss of any kind.

212. The decision has been referred to in a number of later cases, mainly in support of the proposition that no loss flows from a conversion to the extent that the property in question has been recovered. It was referred to most recently by Lord Nicholls in paragraph 63 of the speeches in Kuwait Airways Corpn v Iraqi Airways Co. (Nos 4 and 5) [2002] 2 AC 883, but only in support of the proposition that a plaintiff who successfully sues in conversion cannot invariably expect to recover the full value of the goods. I do not think that his observations are directed to the issue I have to decide.

213. The majority in Hiort v London and North Western Railway Co. were clearly of the view that although subsequent events did not expunge the defendant's wrongful act, it had in fact caused the plaintiffs no loss. The proposition that the claimant cannot recover more than nominal damages if the goods are returned to him undamaged is one with which I have no difficulty; it is in my view no more than an example of damage being made good by subsequent events. Whether that has occurred is essentially a question of fact. Although in that case the goods themselves had not been returned to the plaintiffs, Bramwell L.J. obviously thought that Grimmett's act in handing over the delivery order to the defendant amounted to the same thing.

214. The primary ground on which Thesiger L.J. based his judgment might be seen as supporting a general proposition that a person suffers no damage if he loses property that would have been lost from another cause at a later date, but I do not think that is what he meant. He clearly had in mind the particular circumstances of the case before him in which, as the second limb of his reasoning shows, the plaintiffs had transferred their rights over the goods to a third party to whose order the goods themselves had been delivered. Having sold the goods to Tarpler, the plaintiffs no longer retained an interest of a kind that would ordinarily be reflected in a judgment for the value of the goods; nor could they pass title to the defendants on satisfaction of any judgment. The damage they suffered was therefore limited to the loss of use during the period of their ownership.

215. In Hiort v London and North Western Railway Co. the court was concerned with a claim in conversion, whereas in the present case I am concerned with a claim under a contract of insurance against physical loss of or damage to goods. The concept of loss is therefore one that needs to be kept firmly in mind. In the present case when MTI disposed of oil without authority there was a loss by an insured peril and I agree with Mr. Sumption that that loss is recoverable from the insurers unless it has subsequently been made good. Hiort v London and North Western Railway Co. suggests that in some circumstances a loss can be made good otherwise than by the recovery of the property itself or payment of its value, but whether it has been made good must depend on the particular circumstances of the case. In that case the effect of Grimmett's handing the delivery order to the defendant was regarded as equivalent to returning the goods to them, and thereby returning them to the plaintiffs. It is not difficult to see why in those circumstances the plaintiffs could not recover their value from the defendant.

216. Mr. Kendrick submitted that the subsequent issue of an ITT contract by Glencore in the present case was the direct equivalent of Grimmett's surrender of the delivery order, but I find it difficult to accept that. A delivery order entitles the holder to obtain delivery of the goods and for many purposes represents the goods themselves. One can see, therefore, why Bramwell L.J. regarded Grimmett's surrender of the delivery order to the defendant as in the nature of a return of the goods that had previously been wrongfully delivered to his order. In the present case, by contrast, issuing an ITT contract did not involve any redelivery of oil already taken; it simply involved an agreement to sell the oil in question and the release of that oil to MTI. However, the fact that the oil had already been taken meant that the contract was incapable of being performed in accordance with its terms unless Glencore chose to treat the earlier misappropriation as good delivery under it.

217. Mr. Kendrick submitted that by bringing proceedings against MTI for the price of the oil sold to MTI under the ITT contracts and obtaining judgment against it Glencore had done exactly that and could not now treat the original taking as a wrongful act. In support of that submission he relied on the decisions in Bradley and Cohn Ltd v Ramsay and Co. (1912) 106 L.T. 771 and Verschures Creameries Ltd v Hull and Netherlands Steamship Co. Ltd [1921] 2 K.B. 608.

218. In Bradley v Ramsay the plaintiffs allowed a jeweller, B, to have some gems on approval with a view to a sale at not less than £750. B sold the gems to the defendant for £300. The plaintiffs brought an action against B in which they alleged that he had agreed to pay £750 for the gems or return them and on the basis of that pleading they recovered judgment by consent for £750. B failed to satisfy the judgment and the plaintiffs then began proceedings against the defendant in detinue. Phillimore J. and the Court of Appeal held that the judgment against B amounted to an affirmation of a contract of sale to him and that the plaintiffs were therefore precluded from recovering against the defendants.

219. In Verschures Creameries Ltd v Hull and Netherlands Steamship Co a forwarding agent acting for the plaintiffs delivered goods to one of their customers contrary to their instructions. The plaintiffs then invoiced the goods to the customer, brought proceedings for the price of goods sold and delivered and recovered judgment against him. The customer failed to satisfy the judgment and was adjudicated bankrupt. The plaintiffs then brought proceedings against the forwarding agents claiming damages for conversion by misdelivery. The Court of Appeal (Bankes, Scrutton and Atkin L.JJ.) held that by suing the customer for the price of goods sold and delivered the plaintiffs had elected to treat the act of the forwarding agents in delivering the goods to him as lawful and could not in a later action against the agents treat it as unlawful.

220. The Verschures Creameries case was considered in one of the leading cases on election, United Australia Ltd v Barclays Bank Ltd [1941] A.C. 1. In that case a cheque payable to the plaintiffs had been converted by M and collected on its behalf by the defendant. The plaintiff sued M in quasi-contract (i.e. seeking a restitutionary remedy) but discontinued the action before suing the defendant in tort. The defendant, relying on the Verschures Creameries case, argued that since the restitutionary claim was based on a (fictitious) contract the plaintiff had elected to treat the payment to M as lawful and could therefore no longer treat it as wrongful for the purposes of an action against it in conversion. Both Lord Atkin (at pages 31-32) and Lord Porter (at pages 51-52) regarded Verschures Creameries as a plain case of ratification by the plaintiffs of the forwarding agents' act in delivering the goods which necessarily prevented them from alleging that there had been a breach of authority. The House held that alternative remedies were available to the claimant for conversion of the cheque which did not involve an election between inconsistent rights and that in such cases nothing short of a judgment and satisfaction was sufficient to bar a second action. These principles were re-stated by the Privy Council in Tang Man Sit v Capacious Investments Ltd [1996] 1 A.C. 514.

221. In the present case the essential facts surrounding each of the relevant transactions are these: MTI disposed of a parcel of oil without authority; subsequently Glencore, in ignorance of the misappropriation, issued an ITT contract the effect of which was to release the oil to MTI; MTI failed to open a letter of credit and thus failed to pay for the oil; having discovered that the oil to which the ITT contract related had already been disposed of, Glencore sued MTI for the price and recovered judgment, but the judgment remains unsatisfied.

222. In my view once it had become aware that oil covered by an ITT contract had been taken and disposed of before the contract had been entered into Glencore had a choice: it could either treat the taking as wrongful and claim damages for conversion or it could treat the taking as constituting delivery under the contract and seek to recover the price. Unlike the position in United Australia v Barclays Bank, where the same wrongful act supported a remedy in restitution or in tort, a claim under a contract of sale to recover the price of the goods depends on the seller's having passed the property in the goods to the buyer, or at any rate being willing and able to do so. The seller cannot obtain both judgment for the price and judgment for damages for conversion, because they depend on inconsistent rights, as the court held in Bradley v Ramsay. The distinction is not academic because damages for conversion would normally be measured by the value of the goods which might differ significantly from the price due under the contract of sale. By suing it for the price Glencore elected to treat MTI's previous appropriation as amounting to delivery under the ITT contract instead of an unlawful dealing with the goods. I therefore think that Mr. Kendrick was right in saying that Glencore would be unable to recover damages from MTI in respect of the loss of the goods on the grounds that in taking them it had acted unlawfully.

223. However, the fact remains that the oil was wrongfully taken in the first place and that Glencore has not been paid for it or otherwise compensated in respect of it. Moreover, as I have already pointed out, the issue in the present case is not whether Glencore would be entitled to recover damages from MTI in conversion, but whether it is entitled to recover under a policy of insurance in respect of a loss of goods. Mr. Kendrick submitted that it is not because it is to be treated for all purposes as if the damage caused by the original misappropriation has been made good.

224. This argument proceeds on the footing that by agreeing to sell the goods to MTI Glencore's loss was made good, but in fact it was not. What Glencore obtained by entering into the ITT contracts was a right to obtain the price of the goods in place of a right to damages for their misappropriation, but it did not obtain the price itself, any more than it recovered damages for the misappropriation. It clearly could not have recovered both the price of the goods from MTI and their full value from the insurers; in such a situation either payment of the price makes good the loss pro tanto or the insured must hold it for the benefit of the insurers, so either way he cannot obtain more than an indemnity. By selling the goods to MTI Glencore certainly lost any right it had to recover them and also lost the right to recover substantial damages for conversion and if it thereby prejudiced Alpina's ability to recover from MTI under its right of subrogation its right to a full indemnity would no doubt be affected. However, it is wrong to say that Glencore is seeking to recover from Alpina in respect of a "credit risk" loss, that is, the risk that MTI might not pay the price, rather than the physical loss of the goods. Once a loss by misappropriation had occurred the risk that MTI would be unable to make it good, whether by paying damages or by paying the price of the goods themselves, already existed and was one that rested with the insurers. For these reasons I do not think that the fact that Glencore is now unable to recover damages for conversion from MTI means that the original loss has been made good or that it is unable to recover from Alpina in respect of that loss.

     7. Parcels of fuel oil despatched after 7^th February 1998

225. Between 7^th February 1998 when Glencore first learnt that oil was missing from Fujairah and 13^th February 1998 various parcels of fuel oil were released from the floating storage facility. These comprised (a) a cargo of low sulphur fuel oil shipped on the Horizon XII on 8^th and 9^th February under a contract of sale between MTI and Fal Oil Co. Ltd; (b) a cargo of fuel oil shipped on board the Athenian Horizon between 8^th and 10^th February; and (c) various parcels of bunker fuel loaded on a number of bunker barges on various days up to and including 13^th February. Mr. Rainey Q.C. submitted that between 7^th and 13^th February 1998 Glencore was in a position to exercise a high degree of control over MTI's operations at Fujairah and could have prevented all movements of oil if it had chosen to do so. He submitted that insofar as oil was shipped out of Fujairah or supplied to third parties in the form of bunkers with Glencore's consent during that period there was no misappropriation by MTI and no loss within the meaning of the policy.

226. It seems highly likely that all these parcels were blended at Fujairah by MTI using constituents held in the floating storage facility. The cargo shipped on the Horizon XII was produced to a particular specification required by the Pakistani market and was blended in the vessel's tanks by transhipment of components from the Metrotank. The Athenian Horizon left Fujairah carrying a cargo of 380 cst. fuel oil. That was one of the most common grades of blended bunker fuel oil produced by MTI and it is likely, therefore, to have been blended at Fujairah, although it is not possible to reach any final decision at this stage. The evidence shows that the parcels loaded into the bunker barges were likewise probably all blended fuel oil, but it is not clear whether they were blended on board the Metrotank or shipped as components and blended in the barges' tanks. A decision on that question (if it is necessary) will have to await a later stage.

227. The significance of blending lies in the fact that it destroys the identity of the components from which the blend is produced and therefore gives rise to an immediate loss of those components. Mr. Rainey submitted that where components were delivered to a vessel and blended in its tanks the relevant misappropriation occurred when the oil was delivered to the vessel rather than when it was blended, in other words, that there was a taking (and therefore a loss) immediately prior to the blending. He may be right about that, but I do not think it makes any difference in the present case. Insofar as any of these cargoes were blended up before Glencore became aware of what was going on, the operation will have been unauthorised and will have given rise to a loss under the cover at that point. Insofar as the cargoes were blended into the vessels at a time when Glencore knew what was being done and chose not to intervene, then it may be appropriate to draw the inference that Glencore consented to both the taking and the blending (which occurred almost simultaneously) with the result that no insured loss can be said to have occurred. It is to this question that I now turn.

228. In my judgment at Phase 2 I made findings about the steps that Glencore took to protect its position after it became clear that the Metro group was no longer able to meet its commitments. On 7^th February 1998 various representatives of Glencore met Mr. Kilakos in Athens to discuss the position generally and to find out exactly how much oil was still being held by MTI. At that meeting Mr. Kilakos and his son John signed an agreement giving Glencore a substantial measure of control over the affairs of MTI and MOC. It gave Glencore the right to monitor and control all the facilities operated by the two companies. It also prohibited MTI from undertaking any trading without Glencore's permission and MTI was specifically required to discontinue bunkering operations unless they were specifically approved, supervised and controlled by Glencore.

229. At a meeting in London the next day Glencore's senior management discussed the best way of dealing with the situation that confronted them. In the event they decided to allow MTI to continue bunkering operations at Fujairah. They also decided to allow it to continue making exports from Fujairah provided Glencore was able to control the proceeds of sale. Mr. Kilakos had hoped that Glencore would support the Metro group through the crisis and initially Glencore was willing to see whether the businesses could be saved. It was not anxious, therefore, to bring their operations to a standstill while there was still a chance that they might be rescued. In order to implement these decisions Glencore sent Mrs. Freeman to Athens to oversee the administration and Mr. Garrett to Fujairah to oversee shipping operations.

230. My findings as to the way in which the situation developed thereafter are set out in paragraphs 249 to 252 of my judgment in Phase 2, the material parts of which I repeat here for convenience:

     "249. On 8^th February there was a meeting at Glencore's offices in London to take stock of the position. Following that meeting Glencore sent messages to the masters of the storage vessels informing them that the oil on board had been transferred to it and seeking confirmation that they would follow Glencore's instructions in relation to its disposal. One such message was sent to the master of the Metrotank who replied early the next day that ship-to-ship transfers were currently going on with the Horizon XII and the Athenian Horizon. It is likely that this information was distributed to a number of people in Glencore's office, but no steps were taken at any stage to prevent the shipment from being completed.

     250. On 9^th February Mr. John Garrett arrived in Fujairah. He had been sent out by Glencore to monitor shipments of oil and to arrange for the amount of oil held in the floating storage facility to be measured. He was asked by his superior, Mr. Bloss, to take charge of the shipping documents for the cargo on the Horizon XII and he took delivery of them from MTI on 10^th February. He delivered them to another employee of Glencore, Mr. Jan de Laat, for carriage to London. Beyond that, however, he played no part in the loading of the vessel.

     251. At about the same time as Mr. Garrett was sent out to Fujairah another of Glencore's employees, Mrs Freeman, was sent to Athens to monitor operations in MTI's office there. She arrived there during the morning of 9^th February. On a copy of a telex from BTCL dated 6^th February advising MTI of the opening of a letter of credit covering 25,000 tons +/- 5% of low sulphur fuel oil she noted "Horizon 12 loading now", from which it seems clear that someone in MTI's office had informed her that the shipment was taking place.

     252. The arrangements for this shipment must have been made well before the meeting of 7^th February, but I am unable to accept that this is a case in which the cargo was delivered to the defendants without Glencore's approval. Although at the time of the meeting on 7^th February Glencore had little idea of the extent of MTI's commitments, all those involved must have been aware, as indeed Mr. Heuzé recognised, that its current operations would continue unless steps were taken to interrupt them. One way of doing that would have been to send immediate instructions to the loading master at Fujairah and to the masters of the storage vessels to cease all operations pending further instructions from Glencore. Steps of that kind could have been taken, but for understandable reasons Glencore preferred to allow MTI's operations to continue while monitoring and controlling what went on. It was advised of the shipment on the Horizon XII and in due course took control of the shipping documents which would ordinarily have given it control of the cargo. . . . . . . . . . . ."

231. On 10^th February Glencore sent a message to the master of the Metrotank instructing him to ensure that no cargo was received, transferred or loaded onto another vessel without its express permission and asking him to advise it of any current activities. He replied the same day confirming that he would comply with those instructions and informing Glencore that he was currently discharging fuel into the bunker barge Tritonas. It appears that no movements of oil or products took place after 9^th February without Glencore's approval and that all operations were monitored and approved by Mr. Garrett or surveyors acting under his control.

232. The issue to which this ultimately goes is whether Glencore consented to the disposal by MTI of the cargoes in question in such a way as to prevent it from making claims for losses under policy. I emphasise that because the issue to which the findings in the Phase 2 judgment were principally directed was whether Glencore could recover damages in conversion from Fal Oil on the grounds that it had taken delivery of the cargo loaded on the Horizon XII without authority. Having said that, the insurers are just as much bound by the findings made in Phase 2 as any other party. Mr. Sumption sought to persuade me otherwise, relying on the recent decision of the Court of Appeal in Kennecott Utah Copper Corpn v Minet Ltd [2003] EWCA (Civ) 905; 2003 LRIR 503, but the two cases are not comparable. In that case the issues before the court on the second occasion were quite different from those which had been before it on the first. Accordingly, the judge's remarks on the earlier occasion were plainly obiter and not binding. Here, by contrast, the litigation has been conducted from the outset on the basis that all the parties to the many actions that are being managed together are to be bound by the decisions of fact and law made at each stage of the proceedings and are entitled to be heard before any judgment is given by which their interests may be affected. Alpina did, of course, appear and take an active part in the Phase 2 trial.

233. A loss occurs under a policy of this kind when goods covered by it are lost or damaged by a fortuitous event. Destruction or damage is not the result of a fortuitous event, however, if it is caused by a third party with the prior knowledge and approval of the insured. Mr. Sumption submitted that in the early days of February 1998 Glencore was still uncertain about its rights because it knew, or thought it likely, that others were also interested in the oil still held by MTI and therefore could not be expected to intervene to prevent cargo leaving Fujairah. I accept that Glencore realised that others beside itself might be interested in the oil remaining at Fujairah, but that did not stop it asserting its own title vigorously and I am unable to accept that it was inhibited in its dealings with MTI by any reservations of that kind. Mr. Sumption also submitted that the failure to intervene to prevent an unlawful act does not involve consent to it, but in my view what inference, if any, is be drawn from a failure to intervene depends on the circumstances of the case. In the case of the Horizon XII and Athenian Horizon it is not a matter of imputing to Glencore consent to the disposal of the oil because of its failure to intervene; rather, its failure to intervene reflected a willingness on its part to allow these vessels to be loaded. For this reason alone I do not think that the delivery of the cargoes to these two vessels gave rise to a loss under the policy.

234. Quite apart from that, however, if the cargoes of these two vessels had been blended prior to the commencement of loading, the loss was complete at that point and any subsequent consent to their delivery made no difference. If, as seems possible, the cargo was wholly or partly blended on board from components supplied from the Metrotank, the position is more complicated. Mr. Kendrick accepted that any delivery prior to 7^th February would give rise to a loss under the policy, but he submitted that any delivery thereafter would not do so, having occurred with Glencore's consent. I think his approach was correct in principle because once loading was under way a loss had already occurred which could not then be avoided. However, adequate allowance must be made for the time needed to assess the situation and reach a decision. In my view Glencore could not reasonably have been expected to implement any decision before noon on 9^th February, having been informed for the first time only that morning that both vessels were in the course of loading. The position at that time remains unclear and the precise extent of any recoverable loss will have to be determined at a later stage.

235. The position in relation to the bunker barges differs only to this extent: Glencore was well aware that the supply of bunkers to vessels at Fujairah was a routine part of MTI's business. It was therefore possible for it to reach a decision as early as 8^th February, as indeed it did, whether to allow MTI to continue bunkering operations or to intervene to stop them. If it had wished to intervene, it could have done so early on 9^th February. Accordingly bunkers supplied on or after 9^th February are to be treated as having been supplied with Glencore's approval. That approval would necessarily extend in my view to any blending required to produce fuel of the required grade and specification.

     8. Cheleken crude cargoes

236. Included in Glencore's claim is the value of three cargoes of Cheleken crude oil shipped to Fujairah on board the Zanjan, the Fassa and the Neka between September and November 1997. Alpina contended that none of these cargoes was lost because they all became MOC's property on passing the vessel's permanent flange connection on discharge at Fujairah.

237. The original contract for the sale of Cheleken crude oil by Glencore to MOC was made on 8^th March 1996. It provided for the sale of 20-40,000 metric tons of oil at seller's option for delivery in March, April and May 1996. The price was calculated by reference to a formula based on certain Platts quotations. Payment was to be made by letter of credit 30 days after the vessel gave notice of readiness. The contract provided for title and risk to pass when the cargo passed the ship's permanent flange connection at Fujairah. At a later date the parties agreed to extend the contract to cover deliveries in 1997.

238. Towards the end of June 1997 Mr. Kilakos asked Mr. Heuzé whether Glencore could extend the period of credit allowed to MOC. Mr. Heuzé was rather cautious in his response. On 1^st July he sent Mr. Kilakos a fax in which he said that Glencore would look into the matter but that it would require a holding certificate and perhaps other documentation as security. He undertook to respond later that day. There is no direct confirmation of Glencore's agreement to Mr. Kilakos' request, but I am satisfied that it must have been forthcoming because on 4^th July Miss Gene confirmed MOC's understanding that all deliveries of Cheleken crude arriving at Fujairah on and after 9^th June would be allowed 90 days credit. She asked for confirmation by return, but appears not to have received it until 17^th July when Mr. Hawkins sent a telex confirming that and two other amendments. Mr. Hawkins said that although he could not remember the details of their conversation, Mr. Heuzé must have told him about the changes to the crude oil sale contract he had agreed with Mr. Kilakos when he asked him to draft the confirming telex.

239. The telex message which Mr. Hawkins sent to Miss Gene on 17^th July confirmed two changes to the contract: the addition of condensate as an alternative grade and the extension of payment terms from 30 days to 90 days for deliveries made during June, July and August. There was no mention, however, of the need for a stock transfer certificate and the message confirming those amendments concluded with the words "Otherwise unchanged".

240. It had been Mr. Hawkins' practice to enter crude oil cargoes shipped into Fujairah pursuant to this contract in his JV1 spreadsheet as stock held for the account of the Glencore in order to record their arrival but to make an immediate matching entry showing a transfer to MOC. He did that, he said, to reflect the fact that they were already committed to MOC and were not available to Glencore for disposal elsewhere. One can see from his spreadsheet that Mr. Hawkins made entries in respect of the cargoes on the Zanjan and the Fassa in that way. The cargo on the Neka was shown as entering the inventory, but for some reason no corresponding entry was made at that time recording it as leaving the inventory. Mr. Hawkins sent his pricing calculations for all three cargoes to Miss Gene for her agreement and when she received them Miss Gene set in motion the arrangements for opening letters of credit. However, no letters of credit were opened by MOC and Glencore did not issue any stock transfer certificates.

241. On 13^th January 1998 Miss Gene sent Mr. Hawkins a copy of her spreadsheet covering the period from June to the end of December 1997 to enable him to begin the process of reconciling their stock records. That showed the three Cheleken crude cargoes as being held for Glencore's account. Mr. Hawkins passed the sheet to a member of the accounts department, Mr. Tawana, who compared it with the records kept by Mr. Hawkins. He saw the discrepancy in the treatment of the crude oil cargoes and noted in manuscript on his copy of Miss Gene's spreadsheet that at least two, and perhaps all three, of these cargoes (the document is not very clear) had been sold to MOC.

242. Shortly afterwards someone at Glencore prepared a draft response to Miss Gene identifying the various discrepancies that had been identified between the two spreadsheets, including the different treatment of the three crude cargoes. The draft pointed out that Miss Gene had failed to deduct those cargoes from Glencore's inventory. It is not clear who was responsible for producing that draft. It may have been Mr. Hawkins, but I think it more likely that Mr. Tawana produced it for the purposes of a discussion with Mr. Hawkins, and in any event I am quite satisfied that the two of them did discuss it.

243. On 5^th February at the request of Mr. Heuzé Mr. Hawkins sent a copy of his JV1 spreadsheet to Mr. Kilakos. By that time he had added an entry showing the cargo on the Neka as having been transferred to MOC, thus bringing it into line with the entries for the Zanjan and the Fassa. At the same time he sent an additional sheet headed "Metro reconciliation" which had been prepared by the finance department identifying cargoes that were due to be sold under ITT contracts once letters of credit had been opened. The three Cheleken crude cargoes were included under that head although they were not subject to the established ITT procedures by which other parcels of oil were sold to MTI.

244. At Glencore's internal meeting on 8^th February 1998, which was attended by both Mr. Heuzé and Mr. Hawkins, Mr. Gibson made a note to the effect that 755,000 tons of fuel oil and crude oil had been sold by ITT to MTI of which 50,000 tons was crude. That must have included the three Cheleken cargoes. On 10^th February Mr. Hawkins sent a copy of his JV1 spreadsheet, still showing those cargoes as having left Glencore's inventory, to the insurance department, but a few days later, probably during the week-end of 14^th/15^th February when he was again in the office, he altered the spreadsheet yet again by removing the entries recording the transfer of the three Cheleken crude cargoes to MOC.

245. Mr. Hawkins was asked to explain this somewhat curious sequence of events. He said that, although he could not remember the precise terms of their conversation, Mr. Heuzé had told him that he had agreed to extend the period of credit under the crude oil sale contract from 30 to 90 days as requested by Mr. Kilakos on the basis that Glencore would issue a stock transfer certificate to MOC when the letter of credit was opened. His failure to include that provision in the telex of 17^th July was the result of an oversight. He said that Miss Gene had told him early in September when the Zanjan was about to load that MOC did not want the crude oil parcel she was carrying after all and had effectively refused to accept it. She had subsequently adopted the same position in relation to the following two parcels of crude oil on the Fassa and the Neka. He said that he expected some solution to be negotiated between Mr. Heuzé and Mr. Kilakos and had issued pricing calculations in order to push the matter forward, but in the event the position had not been resolved by the time of the collapse. His original failure to record the Neka cargo as leaving the inventory was an oversight. His reversal of the entries after the collapse of Metro simply reflected his understanding that the oil had not been sold to MOC and therefore was still held to Glencore's order.

246. Three principal issues arise from all this: did Mr. Heuzé and Mr. Kilakos agree to vary the terms of the crude oil sale contract in the manner described by Mr. Hawkins in relation to the cargoes on the Zanjan, the Fassa and the Neka? did title pass to MOC when the oil was discharged at Fujairah? did the parties agree to cancel the contracts for the sale of these three cargoes of crude?

247. Mr. Rainey submitted that Mr. Hawkins' evidence on this part of the case was wholly unreliable and in some respects deliberately false. He submitted that his account of the conversations with Miss Gene relating to these cargoes was not consistent with the documents and that his attempts to explain away the evidence of the spreadsheets was unconvincing. He submitted that no reliance could be placed on what Mr. Hawkins said he had been told by Mr. Heuzé about his agreement with Mr. Kilakos which was in any event at odds with the telex of 17^th July which he had himself drafted on Mr. Heuzé's instructions. In these circumstances he submitted that the terms of the contract as to the passing property had not been varied and that in the absence of an agreement to cancel the sales of these cargoes (of which there was no evidence) property had passed to MOC on discharge.

248. I agree with Mr. Rainey that Mr. Hawkins' evidence on this part of the case was not impressive, but I do not accept he was being deliberately untruthful. Much of the difficulty in my view arises from the fact that he had no clear recollection of the events in question. He wanted to explain how matters had developed and that led him to reconstruct what he thought must have occurred basing himself on his own rather hazy recollections, on his knowledge of how Mr. Heuzé and Mr. Kilakos generally handled things and, to a lesser extent, on the documents. I am unable to reject his evidence outright because in some respects it is consistent with the contemporaneous documents, but I do think it is necessary to approach it with a good deal of caution.

249. It is convenient to begin with the terms of the contract itself and with the question whether there was an agreement between Mr. Heuzé and Mr. Kilakos to vary them in the way Glencore contend. If the matter turned solely on the evidence of Mr. Hawkins, I should have found it difficult to accept that there had been any such agreement, but subsequent events tend to support his account. Except in the case of one small parcel, the parties did change the way in which they handled the sale of crude oil cargoes, in particular, by introducing the use of stock transfer certificates similar to those used to implement ITT contracts. Mr. Hawkins prepared pricing calculations for each cargo for submission to Miss Gene who, having satisfied herself of their accuracy, gave instructions for the opening of a letter of credit providing for payment against, inter alia, a stock transfer certificate. The stock transfer certificates issued by Glencore certified that title to a certain quantity of oil "in tank seller's storage vessel TT METROTANK off Fujairah" was transferred to MOC under the agreement of 17^th July 1997.

250. Strictly speaking the stock transfer certificates were inaccurate, of course, because none of these crude oil cargoes were discharged into the Metrotank, but by the latter part of 1997 certificates in that form had been in use in connection with ITT contracts for a long time and the reference to the Metrotank was used by both parties as the equivalent of "in store Fujairah". The overwhelming likelihood is that Mr. Hawkins simply continued to issue documents in the established form without thinking too carefully about the wording. What is important, however, is the fact that stock transfer certificates of any kind came to be issued under this contract. I can see no reason for this change of practice other than to give effect to an agreement that title should no longer pass on discharge of the cargo but on the issue of a stock transfer certificate. Mr. Rainey submitted that it was MOC who introduced a requirement for stock transfer certificates because that was what its bank had demanded, but there is really no evidence to support that conclusion. The extension of the period of credit from 30 to 90 days significantly increased Glencore's risk and it is not surprising that it should have wished to retain title to the oil until it had obtained the security of a letter of credit. I accept that it is certainly surprising that Mr. Hawkins failed to include an express reference to this new procedure in his telex of 17^th July, especially if one accepts his evidence that he would have shown the telex to Mr. Heuzé before he sent it since it is not the sort of thing that I would expect him to overlook. However, Mr. Heuzé travelled a great deal and Mr. Hawkins certainly did not have a reliable recollection of showing him this particular document. I think it very likely that that part of Mr. Hawkins' evidence reflects the procedure he usually adopted, but that his suggestion that he showed this particular telex to Mr. Heuzé owed more to an attempt to shift, or at least share, the blame for what was on any view a serious oversight.

251. Mr. Rainey also relied on the fact that Minton, Treharne & Davies Ltd ("MTD"), who were engaged by Glencore to conduct an investigation into the loss at Fujairah, did not mention this variation when describing the terms of the crude oil sale contract in their preliminary report issued in August 1998. It is apparent from the report itself, however, that MTD were not seeking to do more than describe in general terms the nature of what had originally been an exchange agreement between Glencore and MOC which had its origin in two separate term contracts of which the contract for the sale of Russian crude oil was one. Although MTD referred to three variations of this contract, they did not deal in detail with their terms and did not, for example, record that the period of credit was extended from 30 days to 90 days. That is scarcely surprising since it was irrelevant to the point they were seeking to address. The general description of the contract terms could be, and I think probably was, obtained from the documents in MTD's possession. Mr. Minton said in evidence that he was given the impression by Mr. Hawkins that as from July 1997 Glencore and MOC had agreed to adopt a procedure similar to that which applied in relation to sales to MTI. It is possible that Mr. Minton was confused about exactly how or when he obtained that information; I rather doubt it, but I do not think that much turns on it either way. Provisions dealing with the passing of title are often of the first importance, of course, as Mr. Minton recognised, but they were not of particular importance for the purposes of a description of the exchange arrangements and I see no reason to think that anyone paid particular attention to them at that time. Accordingly, I am unable to attach a great deal of weight to the fact that MTD did not refer in their report to any requirement for stock transfer certificates to be issued.

252. It was suggested to Mr. Hawkins in cross-examination that the treatment of these cargoes in his inventory spreadsheet, and indeed the way they were treated in the reconciliation process, indicate that Glencore regarded them as having been sold to MOC, but I do not think that is the case for reasons which will become apparent a little later. In all the circumstances, therefore, I am satisfied that there was an agreement between Mr. Heuzé and Mr. Kilakos varying the terms of the contract to provide that property would pass only on the issue of a stock transfer certificate.

253. Mr. Schaff Q.C. submitted on behalf of Glencore that the agreement in its revised form only applied to cargoes that reached Fujairah before the end of August 1997 and that there was therefore no contract governing deliveries after that date. However, I do not think that is correct and it does not appear to have been the parties' understanding of the position at the time. There is evidence in the form of a telex sent by Glencore to MOC on 30^th May 1997 that the original contract had been extended to cover deliveries in 1997 and there is no reason to think that it was to expire before the end of the year. It is true that the agreement to extend the period of credit was originally expressed to apply only to cargoes arriving at Fujairah up to the end of August, but if that lapsed the parties would simply be back to the original terms. In the ordinary way perhaps that is what would have happened, but it was characteristic of Glencore and Metro to continue operating under existing arrangements until one or other of them sought a change and it would have been very surprising in my view if MOC had meekly accepted a return to 30 days credit without raising the matter. In fact, one can see that the next two cargoes were handled on the same terms. The Zanjan arrived at Fujairah on 8^th September carrying three different grades of cargo. She discharged her cargo of crude oil on 14^th September and the other grades between 14^th and 20^th September. On 24^th September Mr. Hawkins sent a pricing calculation for the crude oil to Miss Gene which included a charge for 60 days interest at 7.25% in accordance with the July agreement. He later followed the same procedure in relation to the cargo on the Fassa which arrived on 10^th October and discharged a parcel of crude oil between 11^th and 12^th October. The pricing calculation was sent out on 27^th October. At the end of October, therefore, the parties were still operating under the terms as varied in July.

254. The Neka arrived at Fujairah on 9^th November and completed discharge four days later. For some reason Mr. Hawkins did not send a pricing calculation to Miss Gene until 18^th December and when he did so he did not include a charge for interest; instead he said that the due date for payment had already passed. Mr. Hawkins said that he did not receive specific instructions in relation to any of these three cargoes so it remains unclear why he treated the Neka differently from the other two. However, in the absence of any suggestion that there had been further discussion between the parties, I think that the only sensible conclusion is that the arrangements made in July continued to govern the sale of all these crude oil cargoes.

255. The Cheleken crude cargoes originated in Turkmenistan but were shipped across the Caspian Sea and then carried by road across Iran before being shipped to Fujairah. Glencore was bound under its transportation agreement with Iranian Maritime Services to load the cargo in Iran for shipment to Fujairah and MOC was likewise bound to purchase all cargoes (within the limits set by the contract) sent down to Fujairah in that way. Nonetheless, I do not find it inherently incredible that Miss Gene may have indicated to Mr. Hawkins that MOC did not want to take any more of these cargoes, or that Mr. Kilakos might seek to negotiate some alternative arrangements on the back of MOC's failure to honour its contractual obligations. However, it has to be said that the documents do not lend any support to Mr. Hawkins' account that MOC refused to accept these cargoes, much less that there was any agreement to cancel them. One can see that he sent pricing calculations to Miss Gene in respect of each cargo and requests to open letters of credit without even a passing reference to any such refusal and documents disclosed by Metro confirm that for her part Miss Gene set in motion the procedures leading to payment. Moreover, Mr. Hawkins accepted that when he was discussing the year-end reconciliation with Mr. Tawana he did not mention that MOC had rejected the cargoes or say anything to contradict Mr. Tawana's understanding that the oil had been sold. Mr. Hawkins was certainly vague about the content of his conversations with Miss Gene on this subject, but if any agreement had been reached between Mr. Heuzé and Mr. Kilakos to cancel the sale of these three cargoes it would almost certainly have been documented in some form. It would also have been raised in the course of the discussions that went on in Athens and London on 7^th and 8^th February 1998, but nothing was said on either of those occasions about any such agreement. In the light of all this evidence I am satisfied that no such agreement was made.

256. The way Mr. Hawkins treated these cargoes in his inventory spreadsheet is curious, but is not in my view evidence that he understood that property had passed to MOC on discharge at Fujairah. In my judgment at Phase 2 I made a number of detailed findings about the way in which Mr. Hawkins kept his JV1 spreadsheet. In summary, he treated oil as released to MTI, and therefore as leaving Glencore's inventory, on the ITT date, that is, the deemed stock transfer date for payment purposes contained in each ITT contract. The crude oil cargoes could easily be fitted into that general scheme because under the terms originally agreed property passed on delivery to MOC at Fujairah. It is not surprising, therefore, that Mr. Hawkins entered them in his spreadsheet on arrival and immediately made a corresponding entry reflecting their transfer to MOC. Indeed, Miss Gene appears to have done the same.

257. Once the terms of the contract of sale had been changed that way of recording crude oil cargoes became inappropriate. The cargoes were not intended to be placed at MOC's disposal until a stock transfer certificate had been issued and they were not subject to the course of dealing that had grown up in relation to fuel oil and other products sold to MTI by ITT under JV1 and JV2. However, once stock transfer certificates and extended credit terms had been introduced Mr. Hawkins and the finance department appear to have regarded them in the same light as ITT contracts, as one can see from the reconciliation sheet sent to Mr. Kilakos on 5^th February. Miss Gene, on the other hand, was more accurate, entering them in her spreadsheet as cargoes which still formed part of Glencore's inventory. It is difficult to see why she should have recorded them in that way if the understanding of those at MOC was that title had already passed. The most one can extract from Mr. Hawkins' spreadsheet, therefore, is that as at 8^th February he regarded the oil as "sold" in the sense of being subject to a contract of sale and that before he sent a copy of that document to Mr. Kilakos he made the entry in relation to the Neka which he ought logically to have made much earlier, but had overlooked. One cannot, however, treat these entries as evidence that Glencore thought that property had already passed to MOC, much less as being determinative of the position. Similarly, Mr. Hawkins' later reversal of the entries on 14^th or 15^th February to show the oil still held for Glencore's account could not alter the true position.

258. In the light of all the evidence I am satisfied that title to these three cargoes did not pass to MOC and that its use of them was a misappropriation.

     9. The MOC Refinery claim

259. On 6^th November 1996 Glencore entered into a contract with MOC for the processing of feedstock in its refinery at Fujairah in return for the payment of a processing fee and royalty. Under the contract Glencore was to deliver feedstock to the refinery at agreed rates and MOC was to make products available a day later to Glencore or its nominee, either by delivery f.o.b. refinery or by in-tank transfer. Title to the feedstock and the product remained in Glencore throughout, but MOC bore the risk of loss or damage. The agreement provided for both feedstock and products to be held in commingled storage.

260. In this action Glencore is seeking to recover in respect of a loss caused by the misappropriation by MOC of products produced under the processing agreement. There is a dispute as to what, if any, products were actually produced from feedstocks delivered by Glencore and what, if any, quantities remained in stock on 31^st May 1998 which is the date by reference to which the claim is made. However, these issues call for a detailed analysis of MOC's processing operations over the relevant period and cannot be determined at this trial. There are, however, some issues of principle arising out of the construction of condition 1.18 which can conveniently be decided at this stage. They fall under the following broad heads: (a) whether it was possible for a declaration to be made with retrospective effect; (b) if so, whether a declaration with retrospective effect could be made after a loss had occurred; (c) whether a provisional declaration made on 29^th April 1998 was effective to attach the risk to the cover; (d) if not, whether a declaration made on 7^th December 1998 was effective for that purpose.

     (a) Declaration with retrospective effect

261. The first two questions are closely related and can conveniently be considered together. Condition 1.18 extends cover to oil in storage "if required". Mr. Sumption accepted, therefore, that Glencore could choose whether to insure any particular parcel of oil in storage and, if it wished to do so, had to communicate a requirement for cover by making an appropriate declaration. However, he submitted that the policy did not require declarations under condition 1.18 to be made before the attachment of the risk or even before loss. He submitted that the policy itself contemplates the making of declarations in arrears (i.e. with retrospective effect) and that it would not be commercially effective if declarations had to be made before the risk could attach or a loss had occurred.

262. Although the policy does not stipulate in terms how declarations are to be made, the clause dealing with the brokers' claims settling authority refers expressly to "the Assured's customary monthly bordereaux" which is a clear indication that the parties contemplated that Glencore would send monthly bordereaux containing the necessary information, as indeed is established market practice. The policy draws no specific distinction for these purposes between transit risks and storage risks and in the case of transit risks the basis of valuation clause expressly contemplates that declarations will be made in arrears. By the time the policy was renewed in June 1997 there already existed an established practice of making declarations monthly in arrears, both in relation to transit risks and storage risks. Most risks were declared two to three months after attachment, but in some cases the interval was longer. Mr. Panzeri said that he had expected declarations to be made rather more promptly than that and from time to time Alpina complained that there was too great a delay, but at no stage did it suggest that storage risks had to be declared and cover requested before they could attach.

263. Under a contract of this kind making declarations in arrears does not give rise to any difficulty in relation to transits because the insurance is obligatory on both sides. Accordingly, even if the insured fails to declare a shipment it is covered: see Glencore International A.G. v Ryan (The 'Beursgracht') [2001] EWCA Civ 2051; [2002] 1 Lloyd's Rep. 574. Declarations simply provide the insurer with information about the risks that have already attached to the cover in accordance with its terms. Under this cover, however, the position in relation to storage risks is different because in respect of them the insurance is facultative on the part of the insured, but obligatory on the part of the insurers. Without a declaration of some kind, therefore, there is no means of telling whether the insured wishes to invoke his right to obtain cover. I think Mr. Rainey Q.C. was right, therefore, in saying that in that situation the declaration performs a quite different function.

264. Mr. Rainey submitted that in principle under a facultative/obligatory cover on goods such as this the insured must communicate his intention to attach the particular goods to the cover before the insurers can come on risk. That is certainly a good starting point, but it is necessary to have regard to the particular terms of the contract, the nature of the business insured and any established course of dealing between the parties. In the present case it would have been difficult in practical terms for Glencore to have declared storage risks in advance. Transit risks were declared some months in arrears because, as Alpina recognised, it took time for an organisation such as Glencore to collate the information necessary to enable that to be done, but in most cases the storage risk arose when the relevant transit risk ended and it would have been equally difficult therefore for Glencore to collate the necessary information to enable it to declare goods for storage cover more promptly. Moreover, if declarations with retrospective effect could never be made, the storage cover would have lost much of its commercial value in a situation where risks were routinely being declared three months or more in arrears because in many cases the goods are likely to remain in store for a much shorter period than that. That appears to have been recognised by Alpina which consistently accepted declarations of storage risks in arrears in the same way as transit risks without ever suggesting that cover attached only from the date the declaration was received. Mr. Rainey submitted that those declarations did not make it clear whether cover was being sought retrospectively or only prospectively, but the natural inference in the absence of any indication to the contrary is that Glencore was seeking cover from the moment the goods were put into store rather than from some later and largely arbitrary date and expected to be charged a premium on that basis (as indeed it was). In those circumstances, whatever might otherwise have been the position, I do not think that Alpina can now say that declarations of storage risks were effective only from the date at which it received the relevant declarations and that the goods were therefore covered only from that time.

265. Whether Glencore was entitled to declare storage risks to the cover with retrospective effect after a loss had occurred raises more difficult questions. Mr. Rainey submitted that it could not, relying on the general principle that a contract of insurance is a contract to indemnify the insured against the risk of loss, not to indemnify him against a loss that has already occurred. However, there are exceptions to that principle, one of the best known being the insurance of a ship "lost or not lost", the policy in such cases being effective even in the case of an existing loss provided it is not known to the insured.

266. Once one accepts that a declaration to the cover could be made with retrospective effect it is difficult to deny the possibility that in some circumstances, at any rate, an effective declaration could be made even after a loss had occurred. For example, if Glencore declared goods for storage risks a month in arrears and they were damaged by fire the day before the bordereaux reached Alpina, it would be not be possible for the cover to attach retrospectively from the time the goods entered storage but for the loss not to be covered. Either the risk had effectively attached or it had not. One solution would be to say that a retrospective declaration could not be made at all in respect of goods that had suffered loss or damage of any kind, but that would leave Alpina in the happy position of being able to charge premiums on a retrospective basis in respect of goods that had not suffered a loss while being in no danger of incurring liability in respect of any that had. In my view, therefore, Alpina's agreement to accept declarations of storage risks with retrospective effect necessarily entitled Glencore to make declarations after a loss had occurred, at least in those cases in which it was unaware of the loss at the time the declaration was made.

267. However, that still leaves the question whether Glencore was entitled to make an effective declaration to the cover of goods in storage after a loss was known to have occurred. Mr. Schaff submitted that it could. He pointed out that if, for example, oil in store which Glencore intended to declare in the following month were lost or damaged before the bordereaux were delivered to Alpina, it could make no practical difference that by the time they were actually delivered the loss was already known to have occurred.

268. I can see the commercial force of that argument in a case where it can be shown that there was a fixed intention to declare the risk in question to the cover, but it does face very substantial difficulties. The element of risk is fundamental to a contract of insurance and for that reason it is essential where the cover is facultative/obligatory in nature that an unequivocal step of some kind attaching the risk to the cover has been taken before the insured knows of the loss. If, whatever may have been its original intention, Glencore remained free to change its mind at any time before making a formal declaration, it would have been able to declare a risk when a loss had been suffered and free not to declare it if there had been no loss.

269. Mr. Schaff's answer to these points was that market practice and the course of dealing between Glencore and Alpina were such that no distinction was to be drawn between declarations of transit risks and declarations of storage risks. On a practical level that is correct: the two were treated in the same way. That does not advance the matter, however, unless it is possible to infer from the way in which declarations were handled that Alpina had agreed that Glencore was entitled to declare storage risks with retrospective effect, even after a loss was known to have occurred (and on what terms and under what circumstances), or that the parties had agreed to assimilate storage risks to transit risks for all purposes so as to render the contract obligatory on both sides. Neither market practice nor the course of dealing in relation to declarations, however, supports either of these conclusions. I accept that Glencore did in fact think that storage risks were obligatory on both sides and that in practice it did declare all storage risks to the cover, but there is no evidence that Alpina was aware of that or that it ever had occasion to consider the question. There is nothing in its conduct, therefore, that could be said to demonstrate an agreement to accept declarations after a known loss and it is equally impossible to find in the parties' conduct any agreement on the prerequisites, other than the making of a declaration, for effectively declaring goods to the cover with retrospective effect in those circumstances. Equally, I can find nothing that would support the conclusion that the parties agreed to treat the contract as obligatory on both sides.

270. Mr. Schaff recognised that if Glencore was entitled to declare a risk after a loss was known to have occurred it would be in a position to select adversely to Alpina unless that right were restricted in some way. He submitted that the principles of utmost good faith would prevent it from doing so and that in effect Glencore could only declare to the cover risks which it had always had a fixed intention to declare. At first sight that may appear to be a plausible solution, but it runs into similar difficulties. It would still be necessary to determine whether some overt act was necessary to evidence the formation of such an intention (rather like the underwriter's notation signifying an intention to cede risks in Baker v Black Sea and Baltic General Insurance Co. Ltd [1995] LRLR 261) and whether an intention to declare the risk must itself have been formed before the insured knew of the loss. (If so, there is the added difficulty that even at the beginning of February 1998 Glencore's insurance department was not aware that the processing agreement with MOC had come into operation and that products were being produced which might need to be insured prior to shipment.) These questions could only be answered by referring to the agreement between the parties, but as I have already pointed out, it is not possible to identify any agreement on matters of this kind.

271. In these circumstances I am unable to accept that Glencore was entitled to declare to the cover with retrospective effect goods which it already knew had been subject to a loss.

     (b) The declaration on 29^th April 1998

272. On 29^th April 1998 Glencore made a declaration in the following terms:

     "We have declared the quantity of feedstock delivered to Fujairah under a process agreement with Metro Oil Corp (MOC). Although MOC began processing condensate in December 1997 for declaration purposes, we have declared the condensate only and not the yields. This is because of the situation at Fujairah whereby we are unable to establish when, and in what quantities, feedstock was actually processed by MOC. This information may in due course become known to us, hereby please treat as a provisional declaration."

273. Mr. Rainey submitted that for a declaration to be effective to attach goods to the cover it must at least identify them with reasonable precision, state the date from which cover is intended to attach and identify the place of storage. It must also make it clear that the goods are being declared to the cover. Accordingly, he submitted, this attempt to make a declaration in relation to unspecified quantities of unidentified products was ineffective. In principle I think that is right. I can sympathise with Glencore's difficulty in identifying the products in the present case, but if, as I think, a declaration was required in order for the risk to attach, I do not think that this provisional declaration was adequate for that purpose. The problem lies not in the absence of a clear intention to insure the goods but in the failure to identify them adequately. It is impossible to tell from this declaration what grades and quantities of oil are involved or the dates at which cover is to attach. One cannot tell from this what is the subject matter of the insurance and for that reason the declaration cannot in my view be effective.

     (c) The declaration on 7^th December 1998

274. On 7^th December 1998 Glencore made a further declaration in respect of products manufactured for it by MOC under the processing agreement. That declaration did contain all the necessary information, but by then the policy had been cancelled by agreement with effect from 31^st May 1998. Alpina contended that by December it was too late for a valid declaration to be made.

275. The policy for the 1997-98 year ran from 1^st July 1997 to 30^th June 1998. However, on 8^th May 1998 Alpina wrote to Glencore giving notice of its intention not to take any further steps in relation to the policy and to suspend dealing with all claims until it had reviewed its position. From that point relations between the parties deteriorated to the point where on 27^th May Mr. Gibson wrote to Mr. Panzeri in the following terms:

     "It occurs to us that in the circumstances it would be sensible if we brought forward the cancellation date of the policy . . . . . . You have made it quite clear that you do not wish to engage in further business with our Group, therefore we suggest that the policy is cancelled at midnight on 31^st May 1998. If you agree to this suggestion we will continue to declare all shipments that commence loading before this time, however, storage exposures will of course cease on 31^st May."

276. On 28^th May Alpina responded accepting Glencore's suggestion

     "that there should be an early termination of the open cover with effect from midnight on 31^st May 1998."

277. Despite Mr. Gibson's use of the word "cancellation", I do not think there is any doubt that what both Glencore and Alpina had in mind was the termination of the policy a month before it would otherwise have expired. Mr. Rainey submitted that once the policy had come to an end in that way the open offer to accept further risks also came to an end and could not therefore be accepted by a declaration made in December.

278. It is clear from the terms of Mr. Gibson's letter that he envisaged sending declarations to Alpina in respect of vessels loading before midnight on 31^st May and, although he did not say as much, the logic of his position was that Glencore would also continue to send declarations in respect of goods entering storage before that date. He recognised, however, that cover for storage risks would cease at midnight on 31^st May. All this must in my view have been apparent to Alpina who must be taken to have agreed to it when they accepted Mr. Gibson's suggestion.

279. If, as I think, Glencore was entitled to declare storage risks to the cover with retrospective effect, I can see no reason why that aspect of the contract should have been affected by advancing the date of termination. Since such risks were typically being declared three months or more in arrears, any suggestion that no further effective storage declarations could be made after midnight on 31^st May would have had serious implications for Glencore which are not touched on in Mr. Gibson's letter. I think it is clear that the effect of the parties' agreement was that no risks incepting after midnight on 31^st May could attach to the cover, whether automatically or by way of declaration, but that risks incepting before that date would be covered in the normal way. Accordingly, declarations could be made after 31^st May in respect of storage risks incepting prior to that date providing Glencore was not already aware of the loss, but cover would terminate at midnight on 31^st May in accordance with the parties' agreement. In the present case however, this is entirely academic since by December 1998 Glencore had been aware for some time of the loss of the products in question.

280. Finally it is necessary to mention in this connection the errors and omissions clause in the policy. This read as follows:

     "Assured shall not be prejudiced by any unintentional delay or omission in the reporting hereunder or any unintentional error in the amount or description of the interest, vessel or voyage, or if the subject matter of the insurance be shipped by any other vessel, if notice be given to Underwriters as soon as practicable after said facts become known to the Assured and deficiency of premium, if any, made good."

281. Mr. Schaff suggested that any failure to declare products at the appropriate time was unintentional, having been the result of an oversight, and could therefore be corrected by a declaration made at a later date under this clause. Mr. Rainey submitted, however, that when it refers to "delay in reporting" the clause is concerned only with a failure to communicate an existing decision or state of fact and does not enable the insured to take steps which have substantive effect, for example, by making a declaration that would attach goods to the cover. That would apply with even greater force, of course, in relation to a declaration which would have the effect of attaching goods to the cover after a loss was known to have occurred. In support of that proposition he referred me to a passage in the judgment of Steyn L.J. in Pan Atlantic Insurance Co. Ltd v Pine Top Insurance Co. Ltd [1993] 1 Lloyd's Rep. 496 at pages 502-503 and to a German work, Reinsurance Principles and Practice, by a group of authors under the direction of Dr. Klaus Gerathewohl.

282. Errors and omissions clauses vary in their terms and although some assistance can be obtained from the way the courts have viewed clauses of this kind on other occasions and from discussions of their nature and purpose to be found in published works, the question ultimately comes down to one of the construction of this particular clause. In my view Mr. Rainey was right in saying that the clause in the present case is by its very terms limited to mistakes and delays in reporting matters of an essentially administrative nature and does not extend to decisions by the insured that have the effect of attaching a risk to the cover. I do not, therefore, think that the clause can be relied on to enable Glencore to declare a risk to the cover after a loss has occurred and accordingly I am unable to accept that it enables the declaration made in December 1998 to be treated as effective for that purpose.

     (d) The 'Maersk Visual'

283. In July 1997 MOC entered into a contract with Shell International Trading and Shipping Co. Ltd ("Shell") for the sale of 65-85,000 metric tons of naphtha a month from October 1997 to September 1999 inclusive f.o.b. refinery Fujairah. In order to meet its obligations MOC entered into a contract with Glencore on 22^nd December 1997 on substantially back to back terms, the intention being to lift the cargoes out of product available to Glencore under the processing agreement.

284. Glencore terminated its contract with MOC by a telex sent from Switzerland at 15.32 hours local time (18.32 hours Fujairah time) on 18^th February 1998. At 15.42 hours local time Glencore also sent a telex to Shell in which it informed it of the termination of its contract with MOC and gave notice that MOC had no right to sell naphtha held to its order and could not pass a good title. At 15.00 hours local time on 18^th February the Maersk Visual had begun loading a cargo of naphtha for Shell from the MOC refinery. In its telex notifying Shell of the termination of its contract with MOC Glencore offered to sell the cargo to Shell on the same terms as it had agreed to buy it from MOC. The vessel completed loading at 0930 hours on 19^th February having taken on board 21,896.324 metric tons of cargo.

285. Only one issue in relation to this shipment arises for determination at this stage, namely, whether, as Alpina contends, none of the cargo was misappropriated because Glencore consented to its shipment. Glencore has accepted that there was no misappropriation of that part of the cargo that was loaded prior to the termination of the contract because until that time MOC was authorised to ship naphtha held for its account under the contract that existed between them. The issue is therefore limited to that part of the cargo that was shipped after Glencore had given notice of termination.

286. Mr. Rainey submitted that as from 7^th February Glencore had practical control over the whole of Metro's operations at Fujairah, including the refinery loading jetties operated by MOC. Mr. Garrett was on hand to direct events and, if Glencore had so wished, could have ensured that loading operations were stopped shortly after its contract with MOC was terminated. Instead Glencore decided to allow loading to be completed with a view to obtaining control over the price due from Shell.

287. The issues surrounding this shipment are similar to those arising out of the shipments on the Horizon XII and the Athenian Horizon. The important question in my view is whether Glencore deliberately allowed the cargo to be delivered to Shell when it could have intervened effectively to prevent its being loaded. Glencore became aware on 17^th February that a vessel was coming in next day to load and it appointed SGS to look after its interests. At that stage there was no reason to do otherwise since the contract with MOC remained in existence. However, I have little doubt that Glencore could, or at any rate thought it could, exercise control over loading operations and the fact that it made no attempt of any kind to do so strongly suggests that it was content to allow the cargo to be shipped. It was clearly quite willing for it to be delivered to Shell because it offered to sell it on the same terms as Shell had agreed with MOC. The only matter of any importance was to ensure that the price was not paid to MOC. The conclusion that Glencore was willing for the shipment to be made is further supported by various internal documents that were brought into being soon afterwards. As in the case of the Horizon XII and the Athenian Horizon, I am satisfied that in this case the cargo was shipped with the full knowledge and approval of Glencore and that there was therefore no loss under the policy.

     IV. Quantum

     1. The policy limit

288. The insurers' liability under the policy was subject to an overall limit expressed in the following terms:

     "USD 80,000,000 (or equivalent in any other currencies) any one vessel, aircraft, postal sending, conveyance, or any one loss any one location."

289. It was not in dispute that the floating storage facility constituted one location for these purposes, but there was dispute between the parties as to the number of losses that had occurred in this case.

290. Mr. Sumption submitted that a distinction was to be drawn between diverted cargoes, that is, cargoes that were sent on from Fujairah to other destinations without being discharged, or which were transhipped and sent on without entering the facility, and oil that was discharged into the facility from which MTI later drew without authority. He submitted that cover was available in the full sum of US$80 million in respect of each diverted cargo because they were separate losses in transit in respect of which the limit applied to each vessel individually. In the case of misappropriation of oil in storage he submitted that, in the absence of any words of aggregation, a loss occurred each time MTI misappropriated part of a bulk in which Glencore had an interest, either by delivering it to a third party or by using it to produce a blended product. Accordingly, there were as many losses as there were misappropriations. Mr. Kendrick submitted that in the present case there was only one loss represented by the quantity of oil that MTI was unable to redeliver to Glencore when called upon to do so in February 1998 because it all flowed from one cause, namely, a continuous course of dishonest behaviour on the part of MTI affecting an undifferentiated bulk.

291. Although there is at first sight some attraction in the suggestion that there was only one loss flowing from one continuous course of action, I do not think it is correct. The first point that should be made is that there was not in fact one undifferentiated bulk, but several bulks, each comprising the whole of the stock of the grade in question for the time being. From a brief glance at Mr. Hawkins' spreadsheet it appears that during the second half of 1997 eight different grades of oil were being held in store for Glencore at one time or another. This makes it impossible to contend that there was only one loss because there was only one stock of oil which was depleted by MTI's misappropriations.

292. There is no provision for aggregation in this policy other than whatever can be spelled out of the simple word "loss". This is a policy against physical loss and damage to goods, so in the ordinary way a loss within the meaning of the policy occurs whenever the goods insured are damaged, destroyed or lost to the insured. Thus, several unrelated fires affecting goods in storage would give rise to several losses, as would several unrelated thefts. The position may be more complicated if several losses are related - as, for example, where an arsonist sets fire to two adjacent tanks in the course of a single attack - and no doubt a certain amount of common sense has to be applied when deciding how many losses have occurred in any given case. Thus, if thieves enter a warehouse containing bagged goods which they remove using a number of different vehicles, pausing from time to time to bring up a new vehicle, it is difficult to see how that could be regarded as more than one loss.

293. In both these examples the unifying factor is that the loss occurred on one occasion in the course of a single enterprise. However, the fact that similar goods are stolen on several occasions from the same location by the same person using the same method does not in my view entitle the insurers to treat them all as a single loss under this policy. In Pennsylvania Co v Mumford [1920] 2 K.B. 537 the claimant bank obtained insurance against

     ". . . . . all losses which they might, during the twelve months, discover that they had sustained . . . . . by reason of any bonds . . . . . or other similar securities . . . . . being . . . . . made away with by . . . . . theft . . . . . whether by the officers, clerks and servants of the assured or any other person."

294. During the currency of the policy the bank discovered that one of its servants had, on 41 separate occasions during the relevant period, fraudulently made away with quantities of securities deposited by customers. His practice had been to pretend to the vault clerk that he was authorised to receive the securities on behalf of the owner and to forge the necessary receipt. The court held that the insurers were not liable for reasons that are not relevant to the present issue, but a question had arisen as to the extent of the insurers' liability under the policy on which all three members of the court, Lord Sterndale M.R., Warrington and Scrutton L.JJ. thought it appropriate to express an opinion.

295. The policy contained a proviso that

     "the total liability of each of the undersigned in respect of any one loss under this guarantee is limited to the amount underwritten by him irrespective of the total value of the securities comprised in such loss . . . . . "

     The insurers submitted that "any one loss" meant any one discovery of losses giving rise to a claim, but all three members of the court rejected that argument in favour of the view that there were in that case 41 separate losses, although they had all been discovered at the same time. The argument that there had been only one loss was based on the particular wording of the policy, but all three members of the court considered that a loss occurred each time a group of securities were stolen and the judgments clearly support Mr. Sumption's submission that in the present case there was a separate loss each time MTI misappropriated oil held in store.

296. In Philadelphia National Bank v Price (1938) 60 Lloyd's Law Rep. 257 the defendant insured the claimant bank against losses up to US$300,000 resulting from making advances against forged documents. The policy contained the following proviso:

     "Notwithstanding anything to the contrary herein contained this insurance is only to pay claims for the excess of 200,000 dollars, ultimate net loss, by each and every loss or occurrence."

     The bank also held a second policy in the same terms to covering losses up to US$175,000 in excess of US$25,000.

297. The bank agreed to make advances to a customer, Brown, against the security of invoices for coal shipped to customers which were to be assigned by Brown to the bank accompanied by promissory notes covering the repayment of the advance. Each invoice was to stand as security for the repayment of the notes which accompanied it and also for all past and future advances. Cheques sent to Brown in payment of the invoices were to be delivered to the bank and provided the means of paying off the advance. Brown defrauded the bank by submitting false invoices to obtain advances and subsequently went bankrupt. The total loss to the bank was something in excess of US$400,000 but none of the individual advances exceeded US$25,000. It was therefore necessary to decide whether the bank had suffered one loss for which the insurers were liable (subject to the deductible), or a series of losses each of which fell below the limit of the deductible.

298. The Court of Appeal held that there had been a series of losses and that the bank's claim therefore failed. Sir Wilfrid Greene M.R. pointed out that each time Brown deposited a set of documents he obtained a piece of credit so that the bank was making a series of advances. If any advance could not be recovered the bank suffered a loss in that amount. He considered that to say that the bank had suffered a single loss of US$400,000 was to approach the matter from the wrong end, despite the fact that each invoice was to stand as security for all unpaid advances. The other members of the court agreed, MacKinnon L.J. observing at page 268 that the fact that the individual loans were merged in one general account merely served to disguise the real nature of the business.

299. It can be seen that, although the facts of that case are far removed from those of the present, the two nonetheless have certain significant features in common. In particular, there was in each case a single course of conduct repeated at frequent intervals over a period of time leading to what could be regarded as a single overall loss. In the present case, as in Philadelphia National Bank v Price, the fact that individual misappropriations resulted in a progressive reduction of the total amount held in store is apt to disguise the fact that there were in fact several takings and several losses.

300. Mr. Kendrick sought to draw support from two authorities, Haydon v Lo [1997] 1 WLR 198 (P.C.) and Bristol and West of England Bank v Midland Railway Co. [1891 2 QB 653. Haydon v Lo concerned a contract of professional indemnity insurance by which the insurers agreed to indemnify a firm of solicitors against losses up to a maximum of H.K.$5 million in respect of "any one claim hereunder". By a separate policy of insurance the firm obtained cover in the same terms for liabilities up to a maximum of H.K.$75 million in excess of H.K.$5 million. In each case cover was against any claims made against the firm during the period of insurance.

301. By means of a series of thefts committed on 43 occasions a clerk employed by the firm stole a total a total of H.K.$50 million from one of the firm's clients, and using a forged power of attorney he stole on 8 separate occasions various parcels of shares valued as H.K.$11 million from another client. The first client brought an action against the firm to recover its loss. The second client brought proceedings against the companies in which the shares were held for rectification of the share registers and joined the firm as an additional defendant. The firm brought proceedings against the insurers to determine whether these events gave rise to one claim by each client or multiple claims. The excess insurers argued that each of the 43 thefts constituted a separate claim; in effect, that a "claim" was to be construed as equivalent to a cause of action. The Privy Council rejected that argument. Lord Lloyd of Berwick, giving the opinion of the Board, said that the issue was to be determined by reference to the underlying facts rather than the way in which the claim was formulated by the third party. However, he considered that

     ". . . . it seems unnatural to say, on the facts of the [first] case, that there were 43 separate claims. The reality is that there was only one demand, namely, the demand made by [the client] on Lo & Lo. Although the nature of the demand cannot be decisive, it at least provides a useful starting point in a claims made policy, such as this was."

302. I do not find this decision of great assistance in the present case. As Lord Lloyd pointed out, there is a significant distinction between an all risks policy on goods and a "claims made" professional indemnity policy of the kind with which their Lordships were concerned. A claim is not the same as a cause of action.

303. The distinction between a cause of action and a loss lies at the heart of Mr. Kendrick's second authority, Bristol and West of England Bank v Midland Railway Co. In that case documents of title relating to goods shipped from abroad were deposited with a bank as security for the price of the goods with instructions to the bank to sell the goods if the consignee failed to accept a bill of exchange or pay it a maturity. The goods arrived in this country and were deposited with the defendants to hold them to the order of the shipping company. The consignee accepted the bill of exchange and paid the shipping charges, but before the bill of exchange became due he persuaded the defendant to deliver them to him without the production of a delivery order from the shipping company. When the bill of exchange became due he asked the claimant bank to advance him the money and take up the documents. The bank did so and subsequently obtained delivery orders from the shipping company. The consignee failed to pay the bill of exchange at maturity and when the bank applied to the defendant for the goods it found they had gone.

304. The court held that although the defendant had misdelivered the goods before the documents came into the bank's possession, the bank was a pledgee of the goods and could therefore maintain an action simply on the basis of the defendant's failure to deliver them up on demand. Mr. Kendrick submitted that the same situation arose in this case: Glencore could maintain an action against MTI in respect of the whole of the loss as a result of its failure to deliver up in full the stocks of oil which it ought to have been holding to Glencore's order. I do not doubt that under English law Glencore could maintain an action against MTI on that basis, but I do not think that assists in answering the present question. The insurance in this case is against physical loss or damage to the goods. The fact that a failure to redeliver the missing quantity may give rise to a single cause of action in conversion (if the claimant chooses to pursue a claim in that way) does not mean that there has been only one loss within the meaning of the policy. In my view each time MTI drew oil from the bulk and disposed of it without authority a loss occurred. The number and size of those losses will have to be determined on another occasion.

     2. The contingency policy premium

305. I have already referred to the letter written by Alpina to Glencore on 8^th May 1998 in which it said that it was investigating the possibility of avoiding the open cover and announced that until it had reached a decision it would take no further steps in relation to it. The relevant part of that letter read as follows:

     "Pending completion of our investigations we reserve all our rights in relation to the marine cargo open cover which incepted on 1^st July 1997 (including, for the avoidance of doubt, any right to avoid the open cover for misrepresentation and/or non-disclosure). Until such time as we and our co-insurers under the open cover have determined the nature of our rights we intend to take no further steps in relation to the open cover and shall, in particular, suspend dealing with any claims made under the open cover."

306. As a precaution Glencore immediately obtained replacement cover in the London market on similar terms for the remainder of the policy period, that is, from 8^th May to 30^th June 1998. In the present action Glencore is seeking to recover from Alpina the cost of obtaining that replacement cover as damages flowing from its refusal to perform the contract.

307. If an insurer purports to avoid a contract of insurance without justification part way through its term the insured can treat it as discharged and seek alternative cover elsewhere. If he does so he will be entitled to recover by way of damages any additional cost of obtaining that replacement cover. In its letter of 8^th May, however, Alpina did not say that it was avoiding the policy and indeed it was not until much later in the year that it purported to do so and by that time the parties had already agreed to treat the policy as terminating on 31^st May rather than 30^th June.

308. The starting point in the present case, therefore, is the letter of 8^th May. Although in that letter Alpina did not purport to avoid the policy, it did announce its intention to suspend further performance until it had assessed its position. I agree with Mr. Kendrick that Alpina was entitled to a reasonable opportunity to assess what was on any view a large and complex claim and to investigate the background to it, but it does not follow that it was entitled to suspend further performance of the contract while it did so. Once an insurer is in possession of information that may justify his avoiding the policy he will undoubtedly be wise to reserve his position clearly in order not to give the insured grounds for saying that he has affirmed it and he does not repudiate the policy by taking that course even though by doing so he may cast doubt on its validity. However, Alpina went further than simply reserving its position by stating in terms that it did not intend to take any further steps in relation to the open cover and would suspend dealing with claims until further notice. Repudiation of a contract is a serious matter which should not be lightly inferred, but, unless the contract otherwise provides, neither party is entitled to withhold performance while he decides what course to take and a party who announces his intention of doing so runs the risk of repudiating the contract.

309. Alpina's refusal to continue dealing with claims for an indefinite period was in my view sufficiently serious to amount to a renunciation of the contract, but it did not of itself discharge it. That depended on Glencore's response, but instead of electing to treat the contract as discharged Glencore chose to terminate it by agreement. It may well be, as Mr. Sumption submitted, that both parties recognised that the contract had become commercially unworkable, but that is beside the point. A claimant can only recover damages representing the loss of his bargain if he has been deprived of the benefit of the contract by the defendant's actionable wrong and the renunciation of a contract is not an actionable wrong unless and until the other party chooses to treat it as such. Glencore did not do that and accordingly, even though its response to Alpina's letter was reasonable in commercial terms, this limb of its claim must fail.

     3. The basis of valuation

310. The policy contained a 'basis of valuation' clause which provided as follows:

     " 1. Shipments which are sold by the Assured prior to attachment of risk are valued at:

     1.1 The sum declared if such declaration is made prior to known or reported loss but in no event shall such declaration be less than the Assured's Sale Price.

     1.2 In the event of no declaration having been made prior to loss, valued at Assured's Sale Price plus, if applicable, additional charges.

     2. Shipments which are not sold at the time of attachment of risk are valued at:

     2.1 The sum declared if such declaration is made prior to loss;

     2.2 If no declaration is made prior to loss valued at the higher of either the Assured's cost plus expenses plus 10% or replacement cost.

     3. Shipments insured on instruction of third parties:

     . . . . . . . . . . . .

     4. Interests in store which are not sold at time of attachment are valued at "replacement cost"."

311. Glencore is seeking to recover in respect of oil misappropriated from the floating storage facility by reference to the declared value under paragraph 2.1 of the clause. It seeks to recover in respect of the three Cheleken crude oil cargoes by reference to 110% of the cost under paragraph 2.2 of the clause. Mr. Kendrick submitted that the policy provides separate cover for transit and storage risks and that this clause reflects that approach by dealing separately with goods in transit and goods in store. In the present case, he submitted, oil misappropriated from the floating storage facility can only fall within paragraph 4 and is to be valued at replacement cost. Mr. Sumption submitted that "interests in store" in paragraph 4 refers only to goods that had not previously been the subject of insurance in transit. Goods that had been insured for transit before being discharged into store are therefore to be valued by reference to paragraph 1 or 2, as the case may be.

312. The insurance is against all risks of loss and damage and therefore, although goods in store are not exposed to the same range of perils as goods afloat, the nature of cover is the same in each case. A cargo that is insured in transit before being placed in store is successively a 'shipment' and an 'interest in store'. Mr. Sumption submitted that in such cases the cover attaches once only and the basis of valuation is that which applies when it attaches, but the clause draws a clear distinction between 'shipments' (i.e. goods in transit) and 'interests in store' (i.e. goods in storage). Goods declared for transit under condition 1.15 remain 'shipments' even though they are temporarily in store in the course of the transit, but they cease to be 'shipments' once they reach their final destination. If they are then covered for storage under condition 1.18 they become 'interests in store' from that point onwards. The fact that no additional premium is payable in respect of the first 30 days' storage cover if the goods were previously insured in transit does not mean that the cover continues under condition 1.15, nor is it inconsistent with treating the transit as having come to an end for insurance purposes when the goods are discharged into store if that is their final destination under the contract of carriage. In my view the risk "attaches" for this purpose at the moment when the goods fall within the relevant provisions of the cover, whether that be condition 1.15 or 1.18, and does not attach for all purposes as soon as they come within one or other of those provisions. In the case of shipments, therefore, the risk attaches when the insured becomes at risk or assumes an interest as provided by condition 1.15 without the need for a declaration. In most, if not all cases that occurred at or shortly after the time of shipment. In the case of goods in store the risk attaches with effect from the date of entry into store if the goods are declared for cover in accordance with the policy.

313. In the case of the cargoes that were wrongfully diverted and did not enter the storage facility there was a loss in transit. It therefore becomes necessary to ascertain whether the shipment had been sold prior to the attachment of the risk in order to determine whether paragraph 1 or 2 of the clause applies. It was common ground that when the clause speaks of shipments that have been "sold" it refers to the situation in which the goods are subject to a contract of sale which has not yet been completed by a transfer of title. In most cases it will not be hard to ascertain whether goods were or were not subject to a contract of sale when the risk attached, but the arrangements between Glencore and MTI in this case give rise to more difficult issues. Mr. Sumption submitted that oil in transit had been "sold" within the meaning of the clause prior to the attachment of risk because Glencore had already agreed under the JV2 arrangements to sell it to MTI at a formula price related to the cost of acquisition.

314. Under the JV1 arrangements there was no general agreement between Glencore and MTI for the sale of the oil brought into Fujairah, although that is no doubt what the parties contemplated would happen to much of it. Instead, individual parcels of oil were sold to MTI as and when required by means of individually negotiated ITT contracts. Under JV2 the position was different inasmuch as MTI had agreed to buy all the oil that Glencore shipped into Fujairah pursuant to those arrangements at a price related to the cost of acquisition. However, even under the JV2 arrangements the parties contemplated that they would enter into separate ITT contracts for the sale of individual parcels of oil. In other words, the parties intended that individual ITT contracts, rather than the JV2 arrangements themselves, should be the means by which title to the oil would be transferred from Glencore to MTI.

315. It follows that if the word "sold" in the basis of valuation clause is only apt to apply to situations in which the goods are subject to an immediate contract of sale, none of the diverted cargoes had been sold at the time of the attachment of the risk and that they all fall within paragraph 2 of the basis of valuation clause. However, it is necessary to consider a little more closely what the parties were seeking to achieve by this clause. I think that Mr. Sumption was right in submitting that the reason for drawing a distinction between shipments that are sold and those that are unsold at the time the risk attaches is that where the goods have been sold the insured has secured a means of disposing of them and the price he has obtained for them represents their value to him regardless of any intervening movement in the market. By contrast, where the goods remain unsold their value to the insured is represented by their value in the market at large. Although in the present case the JV2 arrangements did not of themselves involve a binding agreement to buy and sell specific parcels of oil, they did embody an agreement which, if performed in accordance with its terms, would inevitably lead to the conclusion of a series of individual contracts under which the oil would be sold to MTI at a pre-determined price. In these circumstances I think that oil shipped to Fujairah by Glencore pursuant to the JV2 arrangements is to be regarded as having been "sold" to MTI within the meaning of the basis of the clause from the time that Glencore obtained an interest in it. Accordingly, the wrongfully diverted cargoes fall to be valued in accordance with paragraph 1 of the clause.

316. The position is more difficult in relation to oil which did enter the storage facility because although paragraph 4 makes specific provision for the valuation of interests in store which have not been sold at the time of attachment, it makes no provision one way or the other for the valuation of interests which have been sold. This could arise in a number of different ways, perhaps the most likely being as a result of a sale of oil afloat for delivery ex tank at the port of destination. If paragraph 4 had dealt with goods in store in such a way as to apply indifferently to those sold and those not sold, one would have to accept that the same basis of valuation applied in each case, but that is not the case here. There is simply a lacuna in the clause. The preceding paragraphs make it abundantly clear that the parties intended goods sold at the time of the attachment of the risk to be valued at the declared value or sale price as appropriate and in those circumstances I think that it is necessary to imply an additional term in relation to goods in store corresponding to paragraph 1.1 of the clause, both to give effect to the parties' intentions and to give business efficacy to the contract. In view of the fact that declarations with retrospective effect could only be made in the absence of a known loss, it would not ordinarily be necessary to imply a term corresponding to paragraph 1.2, but insofar as declarations were made and accepted after a loss was known to have occurred I think that the parties must have intended that the goods should be valued in accordance with principles contained in paragraph 1.2.

317. Glencore became aware of the loss on 7^th February 1998. It follows that in my judgment oil that had been declared for storage prior to that date is to be valued in accordance with the principles set out in paragraph 1.1. As far as cargoes declared after that date are concerned, the question is not so much how they should be valued as whether cover attached to them at all, given that an effective declaration could not be made after a known loss. The question is not made any easier to answer by the fact that although Glencore and MTI kept records of the quantities of oil in storage by reference to the cargoes shipped into Fujairah, those cargoes did not, in the main, retain their identity after they entered the facility but became merged in a single bulk. As a result, although Glencore knew on 7^th February that large quantities of oil were missing, it could not relate the missing quantities to any particular incoming cargoes. These aspects of the matter were not fully explored in argument and I therefore prefer to leave for decision at a later stage the question whether oil declared after that date was covered, and if so, the basis on which it should be valued.

318. Different considerations apply to the three cargoes of Cheleken crude oil. Glencore originally sought to recover from Alpina on the basis set out in paragraph 2.2 of the clause, but Mr. Sumption accepted that a claim could not be sustained on that basis. He accepted that these cargoes had been sold at the time the risk attached and that therefore the correct basis of valuation was the declared value in accordance with paragraph 1.1, given that all three cargoes had been declared before Glencore became aware of the loss on 7^th February. The fact that they were all declared prior to 7^th February was not in issue, but there was a separate issue between the parties as to whether it was necessary for Glencore to make a further declaration in order to continue cover beyond the first 30 days and, if so, whether they remained insured at the time of loss. Those questions were not fully explored in the course of argument and will remain for determination at a later date.

     4. The deductible

319. The open cover provided cover against leakage and shortage subject to a deductible of 0.5%. Mr. Kendrick submitted that the effect of the agreement to treat bill of lading quantities as binding was to allow undetected shortages to build up which only became apparent when MTI collapsed. He submitted that Glencore was not entitled to make a claim in relation to that shortage without applying the deductible to every shipment discharged into the facility without a discharge survey after April 1995. I am unable to accept that submission, however, because, for the reasons I have already given, I am satisfied that the effect of the parties' agreement was to transfer oil between MTI and Glencore as necessary to ensure that Glencore's interest in the bulk following discharge reflected the full bill of lading quantity, neither more nor less. If MTI did not own enough oil of the grade in question to make up a shortage at the time of discharge, the position would undoubtedly be more complicated since, apart from anything else, it would be necessary to consider whether the necessary transfer could have been and was made at a later date. Insofar as Glencore's interest in the bulk was made up to reflect the quantity shown in the bill of lading it will not need to give credit for the deductible. However, I do not think it would be helpful to embark on a closer examination of the problem in the absence of more detailed information about the movement of cargoes in and out of the facility. That will have to await a later stage in the litigation.

     VII. Summary

320. For the reasons given in this judgment I have reached the conclusion that Alpina is not entitled to avoid the open cover but is liable to indemnify Glencore in respect of such losses as may be proved at a later stage in this litigation to have occurred, taking into account the various decisions I have made on the meaning and effect of the cover.